90nm Growing Pains

[MacRumors]

IDG News (http://maccentral.macworld.com/news/2004/04/21/ibm/index.php?redirect=1082545613000) details the challenges IBM has faced in ramping up 90nm processor production in their East Fishkill plant. The poor yields were cited by Apple as the reason for the shipping delays on the Xserve G5 last quarter.

The rest of the industry has also seen delays in 90nm Production. AMD is expected to reach commercial production (http://www.xbitlabs.com/news/cpu/display/20040420082153.html) of 90nm chips late this year ("September-October-November"), while Intel appears to be in production at this time.

Apple's new PowerMac G5 updates will likely use the 90nm part from IBM. Readers finally report shipments of their Xserve G5s, indicating that production is proceeding. PowerMac and "metallic" display updates are still expected shortly.

[spankalee]

how can anyone, in all seriousness, blame apple for IBM's problem of taking too long to get their chips out in the proper yields...


I don't know, but it happens...

Reading the article again, I see that they left out a major point that could be causing IBM problems. IBM is trying to convert to 90nm and SSOI at the same time. I believe Intel and AMD are still using SOI.

It's also interesting that they say the the nVidia 6800 is produced on the same process. Is nVidia having supply problems too?

[appleface]

is the upcoming g5 pb expected to use the 90 nm chip, or will it hold out for the 65 nm chip? what are the expectations?

[oldmacfan]

That Apple and IBM are also working on a Power5 based PPC. It is my oppinion that Apple and IBM thought that the 970fx would be in full tilt use by now and now they face a dilemma with the Power5 based PPC which should start ramping in May, and available by September for new systems. IBM will release their Power5 based products in June and they are already promoting them.

[spullara]

I got an offer from Apple for $500 off a G5 + Cinema HD display that expires June 26th. Sounds like the announcement of the new displays and systems will be at WWDC again.

http://www.apple.com/promo/brilliantsavings/

Sam

[chasingapple]

I know people like to think it is a race, companies ALWAYS view things as a race to market for sure, but for me this explains alot about the new Powerbook G4's, and why we do not already have the 3Ghz Powermacs. In time we will get all of what we want, G5 everything in our Apples.

Food for thought, IBM is working as fast as they can, Apple is waiting and when it is possible the G5 everything will come, give them time and stop being pissed at "APPLE" for things like this, enjoy your Macs (especially those that are ordering current generation Powerbooks and iBooks) and if you need the fastest get yourself a Powermac Dual 2Ghz for now, or wait for WWDC to see if they announce the new Powermacs.

Take a deep breathe, take a look around you, go outside and look around, remember that the world is not all about a computer, and take in some fresh air. Calm yourself and enjoy life. Perhaps if things go well you will see the G5 in a Powerbook later this year :)

[Frobozz]

Well, I have to say that this is surprising. I can't imagine IBM failing like this but I guess it happens to the best of them. I have, however, heard that the 975 PowerPC did not have the same production problems. The reality is that the 970fx, which would be required for Laptops, may be longer in coming than the more advanced (and higher power/heat) 975.

I'm also not surprised that Intel is already producing 90nm chips-- they don't care about heat.

[Frobozz]

I don't know, but it happens...

Reading the article again, I see that they left out a major point that could be causing IBM problems. IBM is trying to convert to 90nm and SSOI at the same time. I believe Intel and AMD are still using SOI.

It's also interesting that they say the the nVidia 6800 is produced on the same process. Is nVidia having supply problems too?

VERY good point. I've heard reports that the SSOI process, and not the 90nm process, is the problem. 975's didn't have the same problem with the Strained Silicon on Insulator process that the laptop capable 970fx did.

I guess it's all rumor now, but it makes sense...

[arn]

I got an offer from Apple for $500 off a G5 + Cinema HD display that expires June 26th. Sounds like the announcement of the new displays and systems will be at WWDC again.

http://www.apple.com/promo/brilliantsavings/

Sam

http://www.macrumors.com/pages/2004/03/20040328064523.shtml

[Frobozz]

A quote from the article:

"The foundry business also produces chips based on different architectures with different requirements for customers such as Nvidia Corp. and Intersil Corp. Nvidia's new GeForce 6800 graphics chip is made alongside the PowerPC 970FX in East Fishkill on the new process technology, the spokesman said."

Very interesting. Some people mentioned delays in GPU's as a reason for G5 PowerMac delays. I wonder if nVidia 6800's are on a 90nm SSOI process that would also be suseptable to this problem?

[sworthy]

I got an offer from Apple for $500 off a G5 + Cinema HD display that expires June 26th. Sounds like the announcement of the new displays and systems will be at WWDC again.

http://www.apple.com/promo/brilliantsavings/

Sam

you're about three weeks late there....

[agreenster]

Man, I cant wait to get my paws on a dual 3ghz Powermac G5.

June isnt that far away. Hopefully (crosses fingers) the ship date wont lag toooo far behind the announce date...

[iReilly]

A quote from the article:

"The foundry business also produces chips based on different architectures with different requirements for customers such as Nvidia Corp. and Intersil Corp. Nvidia's new GeForce 6800 graphics chip is made alongside the PowerPC 970FX in East Fishkill on the new process technology, the spokesman said."

Very interesting. Some people mentioned delays in GPU's as a reason for G5 PowerMac delays. I wonder if nVidia 6800's are on a 90nm SSOI process that would also be suseptable to this problem?

From what I understand that nVidia is so pissed off at IBM that they are taking their business back to Twaiwan. I think I read that on Ars yesterday.

As a side note (I'm a newbie to Apple) why don't they lower their prices on the current G5 line? Do they always wait for an update to lower prices? I'd willing to buy a Rev. A G5 and an "outdated" monitor if they lower the price. I won't pay what every one else did 9 months ago. Funny, I got money to burn on a new rig (I'm already buying a 17"PB for my kid) for myself and I sit and wait for one of two things, Rev B or a lower priced Rev A. Why do I have a feeling that this is going to happen on the same day? :cool:

iReilly
------------
Dell 3.2

[dongmin]

I have, however, heard that the 975 PowerPC did not have the same production problems. The reality is that the 970fx, which would be required for Laptops, may be longer in coming than the more advanced (and higher power/heat) 975.
That Apple and IBM are also working on a Power5 based PPC. It is my oppinion that Apple and IBM thought that the 970fx would be in full tilt use by now and now they face a dilemma with the Power5 based PPC which should start ramping in May, and available by September for new systems. IBM will release their Power5 based products in June and they are already promoting them.
Where do people get this stuff? Can you guys provide a link where anyone, even posers like MOSR, claim that the 975, if such a thing really exists, is ahead of the schedule of the 970fx???

The issue is with the 90 nm process production, not with any one particular chip design. The 975--agai,n if such a thing exists--is a 90nm chip so it will suffer the same production and cooling issues as the 970fx.

[gensor]

2004/04/20 Taiwan: TSMC (2330 TT, NT$59.5), UMC (2303 TT, NT$31.5): TSMC and UMC benefit from poor yield of IBM's 12" fab


IBM facing poor yield of 0.13um and 90nm
Main customers such as nVidia and Qualcomm are switching back to TSMC and UMC
Maintain BUY recommendations on TSMC and UMC

According to the "Economic Daily," nVidia, Qualcomm, Broadcom, Cisco and Xilinx are switching back from IBM foundry to TSMC and UMC due to the lower yield of IBM's 12" fab. IBM's lower yield of 0.13um and 90nm, and reducing IP revenue are causing IBM's 1Q losses in the semiconductor division.

Analyst comment:
1) We believe that IBM's lower 12" yield in 1Q04 will benefit TSMC and UMC going forward. The order rotation from IBM back to TSMC and UMC is expected to begin in 2Q04. We expect TSMC to show 8-10% and UMC 10% QoQ sales growth in 2Q04.

2) We expect UMC to benefit from capacity support from SiS and He-Jian beginning in 2Q04, which to show stronger revenue growth momentum than TSMC. However, we retain our BUY on both TSMC and UMC, due to their leading position in the technology universe, and capacity ramp-up, respectively.

James Huang, SinoPac Securities, Office: (886 2) 2316-5143, Mobile: (886) 936-348-680, james.huang@sinopac.com

[areyouwishing]

I know people like to think it is a race, companies ALWAYS view things as a race to market for sure, but for me this explains alot about the new Powerbook G4's, and why we do not already have the 3Ghz Powermacs. In time we will get all of what we want, G5 everything in our Apples.

Food for thought, IBM is working as fast as they can, Apple is waiting and when it is possible the G5 everything will come, give them time and stop being pissed at "APPLE" for things like this, enjoy your Macs (especially those that are ordering current generation Powerbooks and iBooks) and if you need the fastest get yourself a Powermac Dual 2Ghz for now, or wait for WWDC to see if they announce the new Powermacs.

Take a deep breathe, take a look around you, go outside and look around, remember that the world is not all about a computer, and take in some fresh air. Calm yourself and enjoy life. Perhaps if things go well you will see the G5 in a Powerbook later this year :)

This is probably one of the most refreshing things i have heard on a rumor board in a long time.

People need to realize that Apple probably doesn't even know when they (g5 pbs) are going to come out, so what is the use in the average rumor monger trying to predict something that even Apple or IBM doesn't know?

Apple wants to get out the PowerMacs, they released the new Xserves for probably 2 reasons, 1 because the line needed it, and 2 because they asked IBM, "We know your 90nm isn't working that great right now, but could you handle producing a small amount of 90s just for the Xserve, this way it gets us our server update, and you guys don't have to waste useless chips to figure out your 90nm problems. Our low volume selling Xserve is the perfect solution."

Updates will come when they come, but i wouldn't expect anything before WWDC, Jobs doesn't want to make an ass out of himself and not hit the 3g mark, so EVERYONE CAN BE ASSURED THAT THEY ARE DOING EVERYTHING IN THEIR POWER TO GET THE NEW G5s OUT!

[pgwalsh]

IBM's hurting Apple on this one. It's sad to see so many manufactures switching back to the other fabs. A positive note is that IBM can focus more on Apples chips. A negative is that there's less production and therefor they cannot distribute the cost of building FK. That could mean that the costs of the G5 could go up, but I doubt it as I imagine that's already been negotiated in the contract.

Too bad for Apple that they've had Motoroloa's lack of focus on G4's and now they run into 90nm fab issues with IBM. None of which are Apples fault. At least the entire industry is facing the same issues. Hopefully IBM can increase yields soon.

This could be a good explinaion of the delays. Another explination could be the uste of ATI PICX graphics cards... Not that they were planning on using these cards.

[jwmci]

The PowerPC and the AMD chip are both produced in East Fishkill. AMD designs their chip but uses IBM's foundry.

I didn't know that Intel's 90nm chip was released.

[MacFan25]

Hopefully this was just a one time thing with the 90nm chips, and this won't turn into something like the Motorola thing. At least they are now in production, so faster G5s must not be far off. However Apple may just wait until WWDC, since it's really not that far off now. But, let's hope they'll be released sooner.

[Fukui]

I don't know, but it happens...

Reading the article again, I see that they left out a major point that could be causing IBM problems. IBM is trying to convert to 90nm and SSOI at the same time. I believe Intel and AMD are still using SOI.

Intel is 90nm + SS.
AMD is SOI + 130nm - moving to 90nm + SOI.
IBM is 90nm + SS + SOI.... hard stuff...

[Piker]

At first when I read on all the rumor sites that the PowerBook G5 wouldn't surface until '05, I thought it was just a sales "smokescreen" to try to get people to buy the newly-revised G4 'books. After all, that rumor appears to have spawned from something said by an Apple exec, and it would only make sense in the interest of sales figures. So I kinda brushed it off and kept my spirits up.

As we find out more and more about these serious delays on the 970FX chips, I'm starting to take the "no PBG5 until next year" rumors more seriously. Painful.

-Piker

[Hemingray]

Thank GOD it's not Motorola! :)

[MacsRgr8]

As a side note (I'm a newbie to Apple) why don't they lower their prices on the current G5 line? Do they always wait for an update to lower prices? I'd willing to buy a Rev. A G5 and an "outdated" monitor if they lower the price. I won't pay what every one else did 9 months ago. Funny, I got money to burn on a new rig (I'm already buying a 17"PB for my kid) for myself and I sit and wait for one of two things, Rev B or a lower priced Rev A. Why do I have a feeling that this is going to happen on the same day? :cool:

iReilly
------------
Dell 3.2

I'm still trying to remember when Apple ever lowered the prices of PowerMacs? I mean in between updates.
I too think it would be quite logical for Apple to lower the prices of the almost 1-year-old PowerMacs. I bet alot of purchases are held off because most potential buyers assume an update sometime soon!
So, why not lower the prices of the current range, and sell more of these babies. I somehow believe Apple had hoped for more PowerMac sales last quarter....

[hacksaw]

The PowerPC and the AMD chip are both produced in East Fishkill. AMD designs their chip but uses IBM's foundry.

I didn't know that Intel's 90nm chip was released.


Man where do you get that from?????


http://www.xbitlabs.com/articles/editorial/display/tech-process_8.html

I would suggest reading the entire article, it's a good primer on processor tech.

[Frobozz]

Where do people get this stuff? Can you guys provide a link where anyone, even posers like MOSR, claim that the 975, if such a thing really exists, is ahead of the schedule of the 970fx???

The issue is with the 90 nm process production, not with any one particular chip design. The 975--agai,n if such a thing exists--is a 90nm chip so it will suffer the same production and cooling issues as the 970fx.

The problem is not the etch-size reduction to 90nm. It's the transition to Strained Silicon on Insulator (SSOI) instead of Silicon on Insulator (SOI).

[mgargan1]

The PowerPC and the AMD chip are both produced in East Fishkill. AMD designs their chip but uses IBM's foundry.

I didn't know that Intel's 90nm chip was released.


Intel's 90nm process came out on Feb. 1st, and it truly sucks! Clock for the the northwood (130nm) is faster because it has a 20 stage pipeline, while prescott (90nm) has a 33 stage...

[chubad]

From what I understand that nVidia is so pissed off at IBM that they are taking their business back to Twaiwan. I think I read that on Ars yesterday.

As a side note (I'm a newbie to Apple) why don't they lower their prices on the current G5 line? Do they always wait for an update to lower prices? I'd willing to buy a Rev. A G5 and an "outdated" monitor if they lower the price. I won't pay what every one else did 9 months ago. Funny, I got money to burn on a new rig (I'm already buying a 17"PB for my kid) for myself and I sit and wait for one of two things, Rev B or a lower priced Rev A. Why do I have a feeling that this is going to happen on the same day? :cool:

iReilly
------------
Dell 3.2

Buy a Refurb Dual 2ghz off of the online Apple store. They come with the same warranty and every product I've bought as a refurb has looked absolutly brand new. I wonder if this is a case where Apple can get rid of some new units and if people don't know about the referbs then they pay full price. If they don't have the duals listed check back often. I've seen them come and go in a matter of haors. They are rarely up for more than a few days.

[zkmusa]

Has anyone received their XServe Dual 2.0Ghz G5 yet? My lab placed the order back in January, and it still has yet to show up.

[psycho bob]

It truly is refreshing to hear people realise that computers aren't everything. People rave on about the fact that Intel are moving ahead in the GHz race, the fact is there new 90nm chip is quite literally burning up. It runs hot and doesn't deliver a significant increase in performance because its outdated lengthy pipeline structure causes significant loses for large data throughput and is also very expensive to produce. This is no doubt the main reason for Intel now adopting their new naming strategy and also pressing ahead with a successor to the P4.
The real competition for the G5 is AMD however they are facing many the same problems because there chips are also produced by IBM. AMD haven't yet moved to SSOI either, IBM as they have done so often are pioneering new technology so surely deserve a bit of a break. We will get new machines and when we do people will start complaining that we don't have 4 or 5GHz processors. In the last year or so there have been very few processor breakthroughs so what's the big deal.
Professionals will keep working and making the best of what is available because lets face it a computer is a computer. People who want to show off there latest all singing all dancing bit of kit will keep complaining. The fact is this isn't a case of a company not investing in the technology and time to develop processors based on the same old spec like it was with motorola and indeed Intel who essentially just milked the P4 for all it had to get high clock speeds out of it.
IBM have invested huge sums of money, it is there business they can't and won't fail but problems take time to sort out. When they are sorted we will get all the 90nm chips we could ever want and it should be smooth sailing until we reduce the die sizes again to 65nm.
If come the WWDC we do get 3GHz G5's then apple will have stuck to what they said they would do and for a while at least the Powermac will be the fastest desktop on the planet. AMD aren't due to hit these levels until the end of the year start of next and although Intel have faster chips in pure GHz terms the technology they use means it will not be enough to keep them ahead until they completely redesign. Even if apple fail to reach this target we all know it won't be through lack of trying.
Quite frankly what would the point be of releasing a new G5 now with a top clock speed of say 2.5GHz (not going to happen) only to then release 3GHz machines a month or so later it just isn't practical or cost effective. Just be patient and if you can't be then go and by a G5 now. I mean dual 2GHz, that's one slow mac isn't it :rolleyes:

[hacksaw]

It truly is refreshing to hear people realise that computers aren't everything. People rave on about the fact that Intel are moving ahead in the GHz race, the fact is there new 90nm chip is quite literally burning up. It runs hot and doesn't deliver a significant increase in performance because its outdated lengthy pipeline structure causes significant loses for large data throughput and is also very expensive to produce. This is no doubt the main reason for Intel now adopting their new naming strategy and also pressing ahead with a successor to the P4.
The real competition for the G5 is AMD however they are facing many the same problems because there chips are also produced by IBM. AMD haven't yet moved to SSOI either, IBM as they have done so often are pioneering new technology so surely deserve a bit of a break. We will get new machines and when we do people will start complaining that we don't have 4 or 5GHz processors. In the last year or so there have been very few processor breakthroughs so what's the big deal.
Professionals will keep working and making the best of what is available because lets face it a computer is a computer. People who want to show off there latest all singing all dancing bit of kit will keep complaining. The fact is this isn't a case of a company not investing in the technology and time to develop processors based on the same old spec like it was with motorola and indeed Intel who essentially just milked the P4 for all it had to get high clock speeds out of it.
IBM have invested huge sums of money, it is there business they can't and won't fail but problems take time to sort out. When they are sorted we will get all the 90nm chips we could ever want and it should be smooth sailing until we reduce the die sizes again to 65nm.
If come the WWDC we do get 3GHz G5's then apple will have stuck to what they said they would do and for a while at least the Powermac will be the fastest desktop on the planet. AMD aren't due to hit these levels until the end of the year start of next and although Intel have faster chips in pure GHz terms the technology they use means it will not be enough to keep them ahead until they completely redesign. Even if apple fail to reach this target we all know it won't be through lack of trying.
Quite frankly what would the point be of releasing a new G5 now with a top clock speed of say 2.5GHz (not going to happen) only to then release 3GHz machines a month or so later it just isn't practical or cost effective. Just be patient and if you can't be then go and by a G5 now. I mean dual 2GHz, that's one slow mac isn't it :rolleyes:


AMD chips are produced in Fab30, located in the Dresden , Germany

[thetruth]

I got an offer from Apple for $500 off a G5 + Cinema HD display that expires June 26th. Sounds like the announcement of the new displays and systems will be at WWDC again.

http://www.apple.com/promo/brilliantsavings/

Sam

I love how people make suppositions on what they want to believe not what reality is. Apple's promo's like this always end on the last day of the fiscal quarter for Apple. In this case June 26th is that day. Drawing any sort of correllation between the end date of the promo and potential new releases doesn't hold water in this case.

-thetruth

[Mr. Anderson]

Man, I cant wait to get my paws on a dual 3ghz Powermac G5.

June isnt that far away. Hopefully (crosses fingers) the ship date wont lag toooo far behind the announce date...

you really think dual 3 GHz will be available in June? I think that might be a bit optimistic if you ask me. If you're lucky they'll announce them and start taking orders for delivery in 3 months (around September) :D

On that note - I'd also love to have a dual 3 GHz....

D

[mgargan1]

The PowerPC and the AMD chip are both produced in East Fishkill. AMD designs their chip but uses IBM's foundry.

I didn't know that Intel's 90nm chip was released.


Yea, Intel released its Prescott (90nm) process on Feb. 1st, and all the benchmarks show that it blows a big one. It's more hot, and slower clock for clock than a northwood (130nm) P4; however, it does have one thing going for it... 16k of L1 cache, compared to 8k of L1 on northwood, and 1M of L2 vs. 512k of cache on the northwood. Prescott has Strained Silicon, but really that hasn't helped yet, but it will when intel releases faster chips. I think that the P4 is on its last legs, but only time will tell. I really do hope that the 970fx, will be as good as everyone hopes. And when/if Apple does go to the 980 it will be dual core. Imagine a dual core dual processor. 4 totall processors, muuuuch better than that crappy hyperthreading.

But as of right now Intel is the only company mass producing 90nm chips, and they're not even producing too many of them...

[numediaman]

you really think dual 3 GHz will be available in June? I think that might be a bit optimistic if you ask me. If you're lucky they'll announce them and start taking orders for delivery in 3 months (around September) :D

On that note - I'd also love to have a dual 3 GHz....

D

Wouldn't we all.

Right now I'd settle for any upgrade, and some guidance from Apple.

EDIT: any upgrade, that is, that come with a new Power Supply, an 8X DVD, etc.

[psycho bob]

AMD chips are produced in Fab30, located in the Dresden , Germany

Your point being? They are still made by IBM, and if IBM are having probs at one fab they are more than likely going to have them at others. I thought IBM had moved all processor prduction of this type to Fishkill anyway?

[elo]

Several days ago, I mentioned firm information I had regarding PowerBook G5's shipping no earlier than 2005 and was treated pretty rudely by someone else on this board. Even though I recognize that the news from IBM is unpleasant for some, I'm glad that so many people here are willing to get past it.

A couple of points in response to comments here:

1. If you're in the notebook market, now is the perfect time to buy. The new PowerBooks and iBooks are absolutely wonderful machines--the best, I think, that Apple has ever made.

2. Apple can't lower prices significantly on existing G5's because doing so would force Apple to *raise* prices when it releases new models. Nothing kills the buzz of a new release more than having it cost a little more than people expect, even if it's entirely worth it. (Just look at the vocal initial reaction by some to iPod mini pricing, even though the pricing turned out to be entirely correct!) Apple very much wants to avoid that scenario with any upcoming G5 PowerMacs.

Having said the above, there are two exceptions: (1) closeout sales to clear inventory (not needed just yet); and (2) cases where Apple feels that it can bring down the machine's price without having to increase it when subsequent models are introduced. In that case (which can't happen too often), you may see price reductions on current models. In the first case, the sale would usually take the form of a promotion, so that it would be clear to customers that the items hadn't been re-priced.

elo

[hacksaw]

Your point being? They are still made by IBM, and if IBM are having probs at one fab they are more than likely going to have them at others. I thought IBM had moved all processor prduction of this type to Fishkill anyway?


My point being, AMD HAS IT'S OWN FOUNDRY.
AMD only license SOI technology from IBM.
IBM does not own FAB 30 in Germany.
LOOK HERE

http://www.amd.com/us-en/Corporate/AboutAMD/0,,51_52_502_509,00.html


http://www.amd.com/us-en/0,,3715_10023,00.html



http://www.xbitlabs.com/articles/editorial/display/tech-process_8.html

[oldmacfan]

Where do people get this stuff? Can you guys provide a link where anyone, even posers like MOSR, claim that the 975, if such a thing really exists, is ahead of the schedule of the 970fx???

The issue is with the 90 nm process production, not with any one particular chip design. The 975--agai,n if such a thing exists--is a 90nm chip so it will suffer the same production and cooling issues as the 970fx.

Power4 and Power5 are very different. As for my comment, it has been know for quite some time that IBM and Apple were co-developing two lines of chips at the same time. They are doing this so they can make up for lost time. This goes back to last summer if memory serves me.

[uzombie]

Well,
Apple will likely have a G5 powerbook no sooner than the summer of 2005. I am speculating, ofcourse. But I wouldn't be surprised if 1/06 is more of target pending chip-output rumors and IBM manufacturing.

My brother-in-law retired from the Vermont IBM facility (before they moved much of the processes to new Fishkill plant). I'll email him if he knows anyone with info on 90nanometer process.

I am wishing, but a quad G5 processor would really pacify my angst. How about you? Yeah, a 3Ghz G5 DUAL would also help me with my future animation projects...sigh.

[gskiser]

Well,
Apple will likely have a G5 powerbook no sooner than the summer of 2005. I am speculating, ofcourse. But I wouldn't be surprised if 1/06 is more of target pending chip-output rumors and IBM manufacturing.

My brother-in-law retired from the Vermont IBM facility (before they moved much of the processes to new Fishkill plant). I'll email him if he knows anyone with info on 90nanometer process.

I am wishing, but a quad G5 processor would really pacify my angst. How about you? Yeah, a 3Ghz G5 DUAL would also help me with my future animation projects...sigh.

I surely hope you're joking. If Apple cannot get the G5 into the powbook before summer of '05, then they are in serious trouble! The G4 is maxed out (as of Monday) so I don't even see them being able to milk it with another minor G4 speed boost this year. What are they going to do, keep the G4 (an antiquated processor with an embarrasing FSB) in the PB until summer of '05? They might as well get out of the portable business if thats the case because, while the PBG4's are good machines, they're already 1-2 years behind their PC counterparts in terms of processing speed.

Granted Apple is better all around and thats why we keep buying despite the slower G4. However, that can only last so long. If they can't get a G5 into the PB by Jan at the latest, they're in trouble unless they find some G5 alternative (and no, a moto G4 won't cut it). Reality is, even though its IBM's fault, its Apple's problem.

[iLilana]

Man, I cant wait to get my paws on a dual 3ghz Powermac G5.

June isnt that far away. Hopefully (crosses fingers) the ship date wont lag toooo far behind the announce date...

remember that the announcement date and shipping dates are ne'r the same.

[mklos]

I surely hope you're joking. If Apple cannot get the G5 into the powbook before summer of '05, then they are in serious trouble! The G4 is maxed out (as of Monday) so I don't even see them being able to milk it with another minor G4 speed boost this year. What are they going to do, keep the G4 (an antiquated processor with an embarrasing FSB) in the PB until summer of '05? They might as well get out of the portable business if thats the case because, while the PBG4's are good machines, they're already 1-2 years behind their PC counterparts in terms of processing speed.

Granted Apple is better all around and thats why we keep buying despite the slower G4. However, that can only last so long. If they can't get a G5 into the PB by Jan at the latest, they're in trouble unless they find some G5 alternative (and no, a moto G4 won't cut it). Reality is, even though its IBM's fault, its Apple's problem.

The G4 is maxed out as of now! So is the G5! So whats your point? Motorola will increase the speed of the G4 just like IBM will increase the speed of the G5. It all takes time. More than likely Motorola is already working in a faster G4 processor. Apple themselves already said that they aren't getting rid of Motorola's G4 anytime soon. In that same article they also stated that the PowerBook will NOT see a G5 processor anytime soon. People just don't understand that its 100x easier to put a G5 in a desktop/server than it is a laptop. There is no room for a heat sink, no room for hardly any air movement, etc. So its extremely hard to cool it. It also consumes a ton of power so IBM would have to make a low powered version. I believe the PowerMac G5 uses a 600 Watt powersupply. Does that tell you something?

[tex210]

if Apple's portables are so horrible, why are they selling so well?

[singletrack]

The G4 is maxed out as of now! So is the G5! So whats your point? Motorola will increase the speed of the G4 just like IBM will increase the speed of the G5. It all takes time. More than likely Motorola is already working in a faster G4 processor.

This PDF http://www.motorola.com/mot/doc/0/786_MotDoc.pdf is from a presentation Moto did back in June 2003. They reckon the G4 is going all the way to 2Ghz and Dual Core. I'm not sure exactly what they mean with DDRI and DDRII built in - I presume they are adding a memory controller which would cure the current problem with the G4 MPX bus maxing out at 167Mhz. That would be nice. RapidIO isn't going to be too popular with Apple I'd guess after they've plumped for HyperTransport in the G5.

Anyway, that would still make for a pretty damn fine laptop even if it came out in Jan 05. I'd suspect a 2Ghz G4 with built in memory controller would wipe the floor with even a G5 2Ghz bearing in mind the G4's shorter pipelines.

Could you imagine the uproar if Apple did come out with another G4 Powerbook after this one though from the "G5 or I'm buying a Dell" crowd?

[topicolo]

One other company the IBM problems will have an affect on is nVidia. The new Geforce 6800 is supposedly being fabbed at the east fishkill plant and any reduction in yield is really going to hurt the chips availability in June.

[Frobozz]

They might as well get out of the portable business if thats the case because, while the PBG4's are good machines, they're already 1-2 years behind their PC counterparts in terms of processing speed.

Don't over exhaggerate. I don't know about you, but no one I know has a Wintel laptop that's noticably faster than an Apple laptop. It's hard to compare the two, anyway, but I'll say this for sure... I know a lot of people in the high-tech sector that have Compaq and IBM laptops that are S-L-O-W. Brand new, and slow. But that's still comparing apples to oranges.

On an application to application comparison, I have no doubt that some "luggable" PC laptop with desktop processors and a 45 minute battery life could squeeze out 25% more performance. But who cares? Anyone that needs that much more speed should be using a desktop for their tasks. I'm not going to be doing 3D rendering of film quality graphics on my laptop-- maybe preliminary work or test work, but not the final deal. That's for the dual 2 GHz G5's...

I don't think Apple is losing the performance war. They beat the hell out of PC laptops on features, and the speeds are comparable, if not MUCH faster on the Apple side for Video. The price/performance of Apple is a lot better than a PC.

[MadMan]

Has anyone received their XServe Dual 2.0Ghz G5 yet? My lab placed the order back in January, and it still has yet to show up.

Place my order in Jan also. First date was March 10th, then it went to April 29th, where it has stayed... Only wierd thing is that when I checked today, my orginal order has a "canceled" next to it, with a new order below it. No idea when THAT happened, but it still lists Apr 29 as the due date.

:cool:

MM

[alexf]

It also consumes a ton of power so IBM would have to make a low powered version. I believe the PowerMac G5 uses a 600 Watt powersupply. Does that tell you something?

Does the G5 really*consume that much power? For some reason I thought that the G5s were supposed to be more energy efficient than their predecessors...

By the way, does anyone know whether the dual processor G5 consumes significantly more energy than the single?

[evilgEEk]

All this talk about faster processors in laptops...PC laptops are faster than the PB's..etc..etc.. I can't help but notice that not many people bring up the fact that the processor isn't all there is to a laptop.

I'd like to see one PC laptop that can do all that a PB can do, whether it be a G4, G5 or G80. ;)

Apple machines come with so much more than a PC laptop, for starters just look at all the amazing free software! Granted iLife isn't exactly professional grade, it's still pretty sweet. And OS X. :cool:

So even though I agree with most here saying that Apple needs to get the G5 into the PB's (and they will), I'm still going to get a G4 PB this fall. Sure I'd like to see a faster FSB, but I'll be quite pleased with my G4 PB.

Cheers!

:D

[gskiser]

People just don't understand that its 100x easier to put a G5 in a desktop/server than it is a laptop. There is no room for a heat sink, no room for hardly any air movement, etc. So its extremely hard to cool it. It also consumes a ton of power so IBM would have to make a low powered version. I believe the PowerMac G5 uses a 600 Watt powersupply. Does that tell you something?

I know it's not as easy to put it into a laptop. However, your 600 Watt comparison of the G5 tower is not completley fair. We're talking about the 970fx, which is different than what's currently in the tower. From what I've read, the 970fx, when clocked down to 1.6-1.8 actually uses less power than than the 1.5 G4 currently utilized in the PB's.

[Sun Baked]

I know it's not as easy to put it into a laptop. However, your 600 Watt comparison of the G5 tower is not completley fair. We're talking about the 970fx, which is different than what's currently in the tower. From what I've read, the 970fx, when clocked down to 1.6-1.8 actually uses less power than than the 1.5 G4 currently utilized in the PB's.With a 20" LCD monitor the estimated load the dual 2.0 G5 draws through the built-in supply at around 235W at idle and 370-400W running full blast.

Makosuke tested his out as: "just idling quietly, it draws 170W and running all-out it's up around 325W."

The 970FX probably wouldn't be much different at the top end, since Apple is going to probably be pushing the new chips to between 2.5GHz - 3.0GHz.

But at the same speed, they are down quite a bit. But there is also the System Controller sucking down some Watts and creating heat.

A big way around this for the Portables would be to get rid of the 2nd processor port, a channel of DDR memory, and run the darn thing at 1/4 to 1/3 the CPU speed.

Basically the portable won't get the G5 until Apple makes a chipset that runs cooler than the one in the PowerMac.

[gskiser]

if Apple's portables are so horrible, why are they selling so well?

They're really not selling that well. Where are you seeing that the PBG4 sales are so phenomenal? Plus its all relative. I concede that they're decent machines, but decent won't cut in 6 months from now. However, sales are relative. Would you argue that Apple's Powerbook sales would not dratically increase if they introduced a PBG5? Just because PBG4's are selling moderately, doesn't mean there isn't room for improvement. Because we only have a PBG4, you cannot quantify how many sales are going unrealized and being lost out on due to people waiting/wanting a PBG5. They're selling now because thats the only option, and we know the only alternative is to switch to PC, which many of us will not do.

[thatwendigo]

With a 20" LCD monitor the estimated load the dual 2.0 G5 draws through the built-in supply at around 235W at idle and 370-400W running full blast.

Which makes me wonder what they have planned. I've been combining rumors in my head lately, and that 600w PSU seems to be offering quite a bit of extra wattage that isn't being used. As someone else pointed out on the boards, Apple isn't in the habit of providing more power than they plan on using, and so it seems kind of stranges on the surface.

What if the next revisions are going to be playing host to some seriously hot/power intensive cards? That would account for the temperature sensor issue, the delay until a little later on, and the higher power supply, all at once. One of the things I noticed about the graph that's been chucked around here is that most of the PC systems that beat the G5 were running Wildcats, FireGLs, and 9800XTs. What if Apple is about to start getting professional support for the professional cards once more, and also issuing the NV6800 as an option at the high end? It's not that farfetched, when one considers that ATI's top of the line mobile GPU is what's in the new powerbooks.

The 970FX probably wouldn't be much different at the top end, since Apple is going to probably be pushing the new chips to between 2.5GHz - 3.0GHz.

I'd expect the 3.0ghz 970fx to be around 45-60w per chip, or roughly on par with the original 2.0ghz 970. The process shrink halved the heat and power draw, but I'm not entirely certain if it will remain a one-to-one tradeoff.

A big way around this for the Portables would be to get rid of the 2nd processor port, a channel of DDR memory, and run the darn thing at 1/4 to 1/3 the CPU speed.

One G5 is already ridiculously hot for a laptop of the kind of form factor the PowerBook uses, when factoring in the components. I think it goes without saying that there isn't going to be a second processor. Also, making a laptop isn't as simple as just modifying the desktop board. Cutting down in the ways that you mention (other than the processor) would basically kill all advantage the G5 has over the G4. It's memory control system and higher clock are what really make it shine, and removing one of the channels and running it at a paltry 1.0ghz would not only make it an advertising disaster, it would also mean having a system that would have circles run around it by the existing powerbooks.

[thatwendigo]

They're really not selling that well. Where are you seeing that the PBG4 sales are so phenomenal?

You know, the last time I checked, the PowerBooks were the best-selling computer at the Apple Store. They were at #6 overall, but don't even appear in the list a the moment. Oddly, the G5s are back in the list at #11. Huh.

Plus its all relative. I concede that they're decent machines, but decent won't cut in 6 months from now.

PCs are still selling, aren't they? ;)


Would you argue that Apple's Powerbook sales would not dratically increase if they introduced a PBG5?

I wouldn't argue that sales won't increase, but I wouldn't go so far as to say that they would, either. We don't know what's going to happen between now and, well, whenever the PowerBooks get bumped to the next processor. I actually hope they never go to the G5, using the 750vx until the G6 comes along and is a much better-suited processor for the task.

Just because PBG4's are selling moderately, doesn't mean there isn't room for improvement.

There's always room for improvement, or else there would be no reason to ever buy more than one computer, of one type, to accomplish all tasks.

Because we only have a PBG4, you cannot quantify how many sales are going unrealized and being lost out on due to people waiting/wanting a PBG5. They're selling now because thats the only option, and we know the only alternative is to switch to PC, which many of us will not do.

Just out of curiosity... When has it not been the case that the option is to go PC or not? How is that at all relevant?

[Sun Baked]

One G5 is already ridiculously hot for a laptop of the kind of form factor the PowerBook uses, when factoring in the components. I think it goes without saying that there isn't going to be a second processor. Also, making a laptop isn't as simple as just modifying the desktop board. Cutting down in the ways that you mention (other than the processor) would basically kill all advantage the G5 has over the G4. It's memory control system and higher clock are what really make it shine, and removing one of the channels and running it at a paltry 1.0ghz would not only make it an advertising disaster, it would also mean having a system that would have circles run around it by the existing powerbooks.Not really the advantage of the G5 over the G4 is bandwidth (aka GB/s).

The System Controller is running quite hot, and the G5 elastic-Bus is designed to run at a 2, 3, and 4x multiplier. A 1.8GHz G5 in a PowerBook running a 450MHz or 600MHz FSB has better bandwidth than the G4 at 167 MHz -- but the biggest benefit is reduced power/heat at the lower clock cycles.

The single channel DDR memory would mean the current 2 DIMM slots stay and you wouldn't need to toss 2 DDR DIMMs every time you upgrade memory. 4 DIMMs is asking a lot of a portable, and if it comes with DDR2...

And it's not modifying the motherboard, but design of a new chipset for single CPU machines and portables.

[Azmordean]

Don't over exhaggerate. I don't know about you, but no one I know has a Wintel laptop that's noticably faster than an Apple laptop. It's hard to compare the two, anyway, but I'll say this for sure... I know a lot of people in the high-tech sector that have Compaq and IBM laptops that are S-L-O-W. Brand new, and slow. But that's still comparing apples to oranges.

On an application to application comparison, I have no doubt that some "luggable" PC laptop with desktop processors and a 45 minute battery life could squeeze out 25% more performance. But who cares? Anyone that needs that much more speed should be using a desktop for their tasks. I'm not going to be doing 3D rendering of film quality graphics on my laptop-- maybe preliminary work or test work, but not the final deal. That's for the dual 2 GHz G5's...

I don't think Apple is losing the performance war. They beat the hell out of PC laptops on features, and the speeds are comparable, if not MUCH faster on the Apple side for Video. The price/performance of Apple is a lot better than a PC.

I assume you know nothing about PC laptops. First off, this "luggable" nonsense has been false since the Centrino was released. My Centrino machine is the same weight and nearly as thin (like 1.25") as an Apple Powerbook - in short, it is every bit as portable. Second, the screen on my laptop absolutely wipes the floor with the junk Apple puts on its laptops - mine is a 14" screen that does 1400x1050 resolution. You'd have to get a lunch tray 17" powerbook to get that kind of screen space from Apple (and for me, screen space was really the deciding factor). As for speed, the Pentium M (the processor part of Centrino) is clearly faster than a G4 -- I do take your point however that for standard laptop tasks, this may not be noticeable. Price wise, PCs win. As for physical build quality, Apple gets the nod - contrary to popular belief, my DELL is quite solid (and its a lattitude "business model", so its silver and black, not that fisher price blue nonsense they put on inspirons), but it isnt brushed aluminum either.

So, in conclusion, both the Centrinos and Powerbooks have advantages and disadvantages. Either is a solid choice in my opinion. I just had to respond though, because I get the feeling people haven't touched a PC laptop since the centrinos were released - so they love to talk about what junk they are.

PS: I am not a "PC Troll" -- I am writing this on my first mac, a PM G5 1.8 single.

[patriotn11]

:confused: OK, this is it. :confused:

15" loaded :eek:

or

17" loaded :eek:

I don't travel but like to loaf on couch and work.

Advice, I'm buying tomm.


I've stated in past post, I am going to write kids books, so I'll be using 3d drawing and of course mac office.

I love games, dvd watching, music, photo's and I want to make my own website.

PS All this and I'll be a full blown switcher, I will use my XP Pro 3gig fro latest gamig, doom3, halflife 2, farcry, and my FAV Battle Field Vietnam.... :D

[blackfox]

They're really not selling that well. Where are you seeing that the PBG4 sales are so phenomenal? Plus its all relative. I concede that they're decent machines, but decent won't cut in 6 months from now. However, sales are relative.
PB sales are quite respectable. Decent not cutting it in 6 months? Does anything not G5 not qualify? May I remind you that a G5 PB does not exist...how do you or anyone else know that they would be vastly superior than the current G4s? Yes, you can look at the PM, but a portable has completely different design limitations...hold your horses...also, the centrino line of notebooks are not much faster than current PBs, so it is not like we are in risk of being left in the dust...
Would you argue that Apple's Powerbook sales would not dratically increase if they introduced a PBG5? Just because PBG4's are selling moderately, doesn't mean there isn't room for improvement.
Of course and of course...but that argument is so obvious as to be meaningless...
Because we only have a PBG4, you cannot quantify how many sales are going unrealized and being lost out on due to people waiting/wanting a PBG5. They're selling now because thats the only option, and we know the only alternative is to switch to PC, which many of us will not do.
No we can't, but you could take out PBG5 and insert PBG6 Dual-processor...it is the same difference...neither exist...take a look at what we have...an excellent notebook in the latest rev. People do not buy portables for speed primarily, but for portability, design, feature-set and decent speed. Calm down on the G5...it will come, personally I'd keep a G4 for a while if they can keep scaling it...

[tex210]

I actually hope they never go to the G5, using the 750vx until the G6 comes along and is a much better-suited processor for the task.
There's always room for improvement, or else there would be no reason to ever buy more than one computer, of one type, to accomplish all tasks.


I like how you think! It's settled then, G4's in PB until G6. Plenty of time for everyone to work out those kinks. Now that we're divided into two camps... must have a v8 in every model versus the hybrid buyers... I'm wondering, is there still anyone in the g3 corner(such as myself)? Have there been any advancements? Miniaturization?

[JFreak]

the screen on my laptop absolutely wipes the floor with the junk Apple puts on its laptops - mine is a 14" screen that does 1400x1050 resolution. You'd have to get a lunch tray 17" powerbook to get that kind of screen space from Apple (and for me, screen space was really the deciding factor).

you actually said nothing reasonable. you tried to argue about screen quality and screen size, but only succeeded in saying you need more pixels.

there are two factors concerning display quality and they are 1) color accuracy and 2) brightness. you tried to argue that your laptop's screen is better than powerbook's, but failed to mention anything about quality.

then, the screen size... i mean, come on! the screen size is measured in inches [not in pixels], and how much bigger your laptop's 14" is compared to 15" or 17" powerbook screen? (i guess you are not comparing to 12" powerbook, because we're talking about widescreens now, right?) for example i need a widescreen display that is about 15" in size, but i also need to see what's in it when i actually use it. and i mean without straining my eyes too much, because when i begin using it, i will continue using it for 8 hours and cannot afford headaches during that time. were the things any smaller in the screen (than the 15" powerbook's 1280x854) i'd have hard time using the computer - i tried to use a 14" screen that had 1600x1200 pixels, but had headaches in five minutes (the reason could have also been the windows xp, but i think it was the tiny text in the screen).

but you're right - if you need to fit more stuff in the screen simultaneously, then you need more pixels. i'd say you also need bigger screen size, but that's up to you if you for example dream about a 14" screen that has 1920x1200 pixels. imho that makes menu texts practically invisible and i'd take those 2.3 megapixels in 20" screen any day.

(by the way, you mentioned screen resolution and did it incorrectly, too. resolution is measured in dots per inch, not in pixels. saying that screen has 1400x1050 resolution is just as wrong as saying your car has 5.7 litre high speed.)

[thatwendigo]

Not really the advantage of the G5 over the G4 is bandwidth (aka GB/s).

I must not have clearly stated that I knew that, then.

The System Controller is running quite hot, and the G5 elastic-Bus is designed to run at a 2, 3, and 4x multiplier. A 1.8GHz G5 in a PowerBook running a 450MHz or 600MHz FSB has better bandwidth than the G4 at 167 MHz -- but the biggest benefit is reduced power/heat at the lower clock cycles.

You must not have ever read one of my posts where I argue with people who say that the G5 should be out now, now, now. One of my main points in those arguments is that a full-speed G5 ASIC is going to be quite a bit hotter than the G4, and that cutting the FSB down is going to kill a lot of the performance advantage. However, I'm curious where you're getting your numbers on the lower heat at lower clock cycles claim.

The single channel DDR memory would mean the current 2 DIMM slots stay and you wouldn't need to toss 2 DDR DIMMs every time you upgrade memory. 4 DIMMs is asking a lot of a portable, and if it comes with DDR2...

Single-channel DDR would elminate the ability to read and write on the same clock cycle, though, wouldn't it? I've always understood that the reason the processor could do that is that it taps one back going up and the other going down.

And it's not modifying the motherboard, but design of a new chipset for single CPU machines and portables.

I also understand that. It sounded like you were saying Apple only had to modify the dual-processor board.

I assume you know nothing about PC laptops.

Strike One.

First off, this "luggable" nonsense has been false since the Centrino was released.

Strike Two.

My Centrino machine is the same weight and nearly as thin (like 1.25") as an Apple Powerbook - in short, it is every bit as portable. Second, the screen on my laptop absolutely wipes the floor with the junk Apple puts on its laptops - mine is a 14" screen that does 1400x1050 resolution. You'd have to get a lunch tray 17" powerbook to get that kind of screen space from Apple (and for me, screen space was really the deciding factor).

You know, I've got a nearly three year old iBook that's got a screen that's just as sharp as the day it was manufactured, and while I'd like a slightly larger resolution, it works just fine. Often, PC laptops come with cheaper displays that can technically go larger, but which look awful when they do so. My coworker's Dell Inspiron is a wonderful example of a brick of a computer that cost about the same as my iBook, only it runs worse.

Price wise, PCs win.

Dell Inspiron 8600
Pentium-M 1.5ghz
XP Pro with Plus! and Digital Media
512MB RAM 1-DIMM
ATI Radeon 9600 Mobile 128MB
4x CD/DVD-R with RecordNow! and MyDVD Deluxe
80GB HD
Dell Wireless 1450 Internal 802.11
Dell TrueMobile Blutooth
McAffee Security Center with 1yr Enrollment
Dell Jukebox Plus
Dell Picture Stuiod with Photo Album Premium
Cost: $2,686

Apple PowerBook 15"
Motorola G4 1.5ghz
MacOS X 10.3.3
512MB RAM 1-DIMM
ATI Radeon 9700 128MB
SuperDrive
80GB HD
Airport Extreme
Blutooth
iLife
Cost: $2,599

So, in conclusion, both the Centrinos and Powerbooks have advantages and disadvantages. Either is a solid choice in my opinion. I just had to respond though, because I get the feeling people haven't touched a PC laptop since the centrinos were released - so they love to talk about what junk they are.

I have. So what?

[thatwendigo]

I like how you think! It's settled then, G4's in PB until G6. Plenty of time for everyone to work out those kinks. Now that we're divided into two camps... must have a v8 in every model versus the hybrid buyers... I'm wondering, is there still anyone in the g3 corner(such as myself)? Have there been any advancements? Miniaturization?

Uh, hey... The 750vx is a G3. It's a super-optimized design that beats the G4 clock-for-clock, is higher frequency to begin with, has a lower heat profile, a higher system bus, and generally beats the crap out of the current chip. A 2.0ghz 750vx is supposed to run at 11 watts, which is a significant drop over the G4 while also making the G5 look a bit ridiculous as a laptop chip.

So, yes, the G3 has had an amzing advancement sice IBM completely took it over. I wish they'd use it in the PowerBook, perhaps as a dual-chip design.

[Sun Baked]

Single-channel DDR would elminate the ability to read and write on the same clock cycle, though, wouldn't it? I've always understood that the reason the processor could do that is that it taps one back going up and the other going down.The System Controller sees a pair of 64-bit DIMMs as a single 128-bit DIMM, not as two different 64-bit banks of memory.

Going from the current dual channel to the single channel would just cut the bandwidth in half, but this also kills the need to add memory in pairs. And in the limited space of a notebook, pairs of DIMMs are a bigger problem than having memory operate at half the bandwith. Then there is DDR2 coming soon...

The rest of it, think what you will since you know more than I.

Which makes it obvious that you already know all about how the memory works, and there really was no point in explaining it.

[aswitcher]

The G4 is maxed out as of now! So is the G5! So whats your point? Motorola will increase the speed of the G4 just like IBM will increase the speed of the G5. It all takes time. More than likely Motorola is already working in a faster G4 processor. Apple themselves already said that they aren't getting rid of Motorola's G4 anytime soon. In that same article they also stated that the PowerBook will NOT see a G5 processor anytime soon. People just don't understand that its 100x easier to put a G5 in a desktop/server than it is a laptop. There is no room for a heat sink, no room for hardly any air movement, etc. So its extremely hard to cool it. It also consumes a ton of power so IBM would have to make a low powered version. I believe the PowerMac G5 uses a 600 Watt powersupply. Does that tell you something?


Well if this is all true, and it sounds reasonable, then the new G4PB17 looks more tempting...

[klaus]

:confused: OK, this is it. :confused:

15" loaded :eek:

or

17" loaded :eek:

I don't travel but like to loaf on couch and work.

Advice, I'm buying tomm.


I've stated in past post, I am going to write kids books, so I'll be using 3d drawing and of course mac office.

I love games, dvd watching, music, photo's and I want to make my own website.

PS All this and I'll be a full blown switcher, I will use my XP Pro 3gig fro latest gamig, doom3, halflife 2, farcry, and my FAV Battle Field Vietnam.... :D

This isn't a buying advice threat, you'll get more replies if you post a new thread in the appropriate section, don't want to be a nag, but don't think you'll be getting a good answer if you post it here..

[billyboy]

for example i need a widescreen display that is about 15" in size, but i also need to see what's in it when i actually use it. and i mean without straining my eyes too much, because when i begin using it, i will continue using it for 8 hours and cannot afford headaches during that time. were the things any smaller in the screen (than the 15" powerbook's 1280x854) i'd have hard time using the computer - i tried to use a 14" screen that had 1600x1200 pixels, but had headaches in five minutes (the reason could have also been the windows xp, but i think it was the tiny text in the screen).



That is an interesting point actually. Maybe there is a growing pixel myth that has little bearing on reality for most people. I do translation work, and have 4 panes open at once so I can copy drag and paste between documents and dictionaries etc as I need. I have no squinting or trouble deciphering anything on a heap of junk PB screen. On the other hand, my housemate has a 2 year old Dell something or other laptop and he says it still has one of the highest resolution screens. Fair enough, but I dont know if it is "better" than the screen on my PB, because if I look over his shoulder as he is writing his private MSN messages using about half the screen, I cant read a bloody thing because everything is so small.

Maybe I have bad eyes, though. I have recalibrated the PB screen to be a shade or two dimmer than default and it takes a lot of 12 hour days before my eyes feel remotely tired.

[thatwendigo]

Going from the current dual channel to the single channel would just cut the bandwidth in half, but this also kills the need to add memory in pairs. And in the limited space of a notebook, pairs of DIMMs are a bigger problem than having memory operate at half the bandwith. Then there is DDR2 coming soon...

Alright, I get what you're saying, although I disagree on a pair of DIMMS being a bad idea. Yes, it would cut into the commodity approach to just slapping whatever you want into the machine, but every edge we can keep on the G5 while keeping it portable needs to be taken.

I don't know if you're right, though.

Apple (http://www.apple.com/powermac/architecture.html) says:
The new Power Mac G5’s memory controller supports fast 400MHz, 128-bit DDR SDRAM, and enables main memory to address two banks of SDRAM at a time, reading and writing on both the rising and falling edge of each clock cycle. This effectively doubles the bandwidth, enabling the Power Mac G5 to reach a maximum memory throughput of up to 6.4GB per second — an advance that’s especially welcome when you’re working with enormous files. In addition, direct memory access (DMA) works with the point-to-point system controller to give each subsystem — such as PCI cards and graphics processing units — its own 6.4GBps interface to main memory, without siphoning power from your processors.

While I could be reading this wrong, I'd be curious to see what sources you have that say the G5 treats two 64-bit RAM slots as one 128-bit. The implication of the above passage is that it accesses the two slots separately, reading and writing at both edges of the clok cycle.

Which makes it obvious that you already know all about how the memory works, and there really was no point in explaining it.

I'm always open to the idea I could be wrong, and I'm not an electrical engineer. I just know a few of them and tend to read things for myself.

[aswitcher]

That is an interesting point actually. Maybe there is a growing pixel myth that has little bearing on reality for most people. .


From what I recall from a recent article on Apple's screen resolution, Apple have focused on something like 100dpi visible, whilst other manufacturers have bumped up to 125 and higher. What this means is that as long as the icons and fonts are properly scaled to recognise the visible dpi you can view everything in the true visible scale as you can now but because the dpi are higher the icon, text and images are crisper! Thus less eye strain because they are clearer.

But, said this article I can't find, Apple OSX doesn't work as well as XP for this simple scaling, and so Apple may need to wait until they address this in a rewrite (10.4?) before they can really release higher resolution screens without asking everyone to squint or put up with odd scaling problems for somethings.

So, maybe the PBG5 and maybe the new screens have been held back for the same reason???

[thatwendigo]

But, said this article I can't find, Apple OSX doesn't work as well as XP for this simple scaling, and so Apple may need to wait until they address this in a rewrite (10.4?) before they can really release higher resolution screens without asking everyone to squint or put up with odd scaling problems for somethings.

I find that extremely unlikely. In System Preferences, you can adjust the font smoothing by degrees (with recommendations for monitor type). On top of that, all you need for an active demonstration of how silly a claim this happens to be is to go to the Dock preferences. See the Dock Size slider? Play with it. For extra fun, turn magnification on and watch icons smoothly shifting in size.

Sure looks like a "simple scaling" problem to me. :rolleyes:

[aswitcher]

I find that extremely unlikely. In System Preferences, you can adjust the font smoothing by degrees (with recommendations for monitor type). On top of that, all you need for an active demonstration of how silly a claim this happens to be is to go to the Dock preferences. See the Dock Size slider? Play with it. For extra fun, turn magnification on and watch icons smoothly shifting in size.

Sure looks like a "simple scaling" problem to me. :rolleyes:

Ok, I dont have a Mac (YET!) so I cannot verify this claim (nor find the article :( ) but whilst I recall it admitted you could modify some things fairly easily, it said soemthing to the effect that with XP you could move one scale to change everything (without changing the actual screen resolution) whilst (claims the article) in OSX its not that easy and not everything can be managed...

[aswitcher]

Here's the article

http://www.macobserver.com/columns/devilsadvocate/2004/20030330.shtml

[thatwendigo]

Ok, I dont have a Mac (YET!) so I cannot verify this claim

Go to any store that they let you handle macs (like the Apple Store! :D ) and see if they'll let you try out what I just told you. It will show you how ridiculous a claim that is, and how it's likely FUD.

but whilst I recall it admitted you could modify some things fairly easily, it said soemthing to the effect that with XP you could move one scale to change everything (without changing the actual screen resolution) whilst (claims the article) in OSX its not that easy and not everything can be managed...

Well, I could believe that there would be a few things that you could alter in Windows that you can't on OS X. Windows is rather famous for its vaunted alterability, after all. However, I don't really buy the whole "easier" thing, without more information on just what they're talking about.

[chasingapple]

All of a sudden I am more interested in these new G3's then I am the G5's, please is there a place I can read up on these beauty CPU's?

P.S. I am a huge fan of my 2 G4's, they are fast :D

[rdowns]

Uh, hey... The 750vx is a G3. It's a super-optimized design that beats the G4 clock-for-clock, is higher frequency to begin with, has a lower heat profile, a higher system bus, and generally beats the crap out of the current chip. A 2.0ghz 750vx is supposed to run at 11 watts, which is a significant drop over the G4 while also making the G5 look a bit ridiculous as a laptop chip.

So, yes, the G3 has had an amzing advancement sice IBM completely took it over. I wish they'd use it in the PowerBook, perhaps as a dual-chip design.

Does anyone have any real information on the mythical 750vx? All I've ever seen has been posted in this forum. Links????????

[D*I*S_Frontman]

I have to echo the last few posters regarding suped-up G3s.

From what I remember posted here @ MR, supposedly IBM was working on a "G3+AltiVec" chip, one which might not be as FSB-hampered as the G4s seem to be. Cooler running as well.

Seems to me that IBM should be able to put a G4-killer like that together for mobile computing. Apple's marketing geniuses would have to rename it, of course. The reality distortion field is not strong enough to sell the average Joe on the idea that he should pay top dollar for a G3-based laptop, no matter how much better than the Moto G4 it might be.

Dual G3+Altivec portables with solid performance and which sip the battery's juice slowly--could be promising. Let's call it the "GM1" or "G1M" chip, denoting a design specifically for mobile applications rather than a G3.

So is all this mere MR mythology?

[JFreak]

Here's the article
http://www.macobserver.com/columns/devilsadvocate/2004/20030330.shtml

this article makes one solid point: apple has decided to aim to 100dpi resolution instead of letting the resolution scale. this might be under development in apple r&d and there's a lot more to think than the article has mentioned. (actually, the article only mentioned what microsoft has done ten years ago, nothing more.)

[aswitcher]

this article makes one solid point: apple has decided to aim to 100dpi resolution instead of letting the resolution scale. this might be under development in apple r&d and there's a lot more to think than the article has mentioned. (actually, the article only mentioned what microsoft has done ten years ago, nothing more.)


What I felt the article made clear was that Apple could go for much higher resolution screens and are well situated to do so, but its going to require some sort of signficant upgrade to the osx (say in 10.4 or even a 10.3.x) to allow this to happen. When they do they can bring out new screens with higher resolutions - such as the G5PB - and allow people to view sharper images/video etc. That's one of the reasons I was hoping the G5PB and the larger screens were out by now so I could know if they were headed in that direction.

[Sun Baked]

While I could be reading this wrong, I'd be curious to see what sources you have that say the G5 treats two 64-bit RAM slots as one 128-bit. The implication of the above passage is that it accesses the two slots separately, reading and writing at both edges of the clok cycle.

I'm always open to the idea I could be wrong, and I'm not an electrical engineer. I just know a few of them and tend to read things for myself.This is all in the discussions from when the G5 was first released, Apple uses 128-bit memory bus in all the documents and the stuff on reading and writing on the rising and falling edge is a typical DDR explanation.

Apple also uses quite clearly that the elastic-Bus transfers data in two directions at the same time.

Of course if you want to catch up on all 2500 posts about the G5 and the architecture... :eek:

http://episteme.arstechnica.com/eve/ubb.x?a=tpc&s=50009562&f=8300945231&m=9080959175

You can read the posts or take Apple at their word when they say the DDR400 memory is 128-bit instead of 2 64-bit DDR400 channels or even the crazy math DDR800 in Intel speak.

---

As far as reducing the speed of the System Controller and getting rid of the stuff that's not needed.

PowerMac G5 System Controller Heat Sink (http://www.970eval.com/g5-disassembly/P1010061.jpg)

[JFreak]

well, yes. currently osx ui elements seem to have fixed pixel dimensions, meaning menubar on 12" ibook is smaller than menubar in 14" ibook, and they need to change that.

microsoft's implementation is kind of technical and therefore almost nobody uses it, but i can see apple adding a simple and easy-to-use dimension calibration into displays preference pane (next to color calibration tab for example) - they only need to put in one slider that scales the screen elements without telling user anything about resolutions.

[Benjamin]

Well even tho there is a lack of 90nm chips that are coming out for apple/ibm, I think this is in someways pretty good since that means that demand is over production. And unlike moto IBM will commit to this venture then scaling it down.

[eSnow]

Uh, hey... The 750vx is a G3. It's a super-optimized design that beats the G4 clock-for-clock, is higher frequency to begin with, has a lower heat profile, a higher system bus, and generally beats the crap out of the current chip. A 2.0ghz 750vx is supposed to run at 11 watts, which is a significant drop over the G4 while also making the G5 look a bit ridiculous as a laptop chip.

The only drawback is that it is vaporware.
The IPC of all other G3's (750 - 750gx) is much worse than that of the G4, so I'd be cautious about projections for the 750vx. The 74xx is much more than a 750 with AltiVec bolted on - even if people try to negate this on the rumor boards.

[Thor74]

IF you believe everyhting on rumor boards/sites, and IF you are long time mac fan... this all reaks of deja-vu... We all hated Motorola and it's lack of drive on the G4. Now we all get to scream at IBM for not being able to pump out the G5 fast enough... At some point, we will have to turn our eyes to the main company - Apple - and place just a tiny bit of blame on their processor decisions.

[Phinius]

This PDF http://www.motorola.com/mot/doc/0/786_MotDoc.pdf is from a presentation Moto did back in June 2003. They reckon the G4 is going all the way to 2Ghz and Dual Core. I'm not sure exactly what they mean with DDRI and DDRII built in - I presume they are adding a memory controller which would cure the current problem with the G4 MPX bus maxing out at 167Mhz. That would be nice. RapidIO isn't going to be too popular with Apple I'd guess after they've plumped for HyperTransport in the G5.

Anyway, that would still make for a pretty damn fine laptop even if it came out in Jan 05. I'd suspect a 2Ghz G4 with built in memory controller would wipe the floor with even a G5 2Ghz bearing in mind the G4's shorter pipelines.

Could you imagine the uproar if Apple did come out with another G4 Powerbook after this one though from the "G5 or I'm buying a Dell" crowd?

Motorola also stated last year that the plan is to double the frequency of the G4 about every 18 months. That essentially means the G4 would move to 2GHz about July 2004. So, expect either a dual-core 2GHz G4 chip or a single 2 GHz G4 processor about that time.

[pgwalsh]

The problem is not the etch-size reduction to 90nm. It's the transition to Strained Silicon on Insulator (SSOI) instead of Silicon on Insulator (SOI).Maybe I missed your piont, but isn't Intel also haveing 90nm woes?
---- On Another Note ---
I’ve reevaluated my expectations for Apple and their product release cycle. Since it’s getting more difficult for processor manufactures to ramp up speeds and there’s been notable cooling problems with the transition to the 90nm process, I wont expect clock speed upgrades as frequently as in the past. However, I’d like to see more than one speed bump a year. It will be a good thing when IBM gets things ironed out, but I don’t expect the speed increase that we saw with Intel going from 2.0Ghz to 3.0Ghz or farther.

I don’t think it would be a bad thing for Apple to have revision releases with newer better components beyond the stock components (Video cards, DVD drives etc.). If they would also mention other improvements, such as chipset revisions, in the release, that would be a good thing.

[blackfox]

Muahaahaahahahahahhahahahahahaaaaaaaaahaahaaaaaaaaa. You. have. no. clue. Actually, we are indeed not "in the risk", because all Apple can offer in the notebook market is already eating dust by the shovel.

Test drive any Centrino out there and report back.
Sorry, don't agree. I do believe that the pentiumM and the Centrino chipset are superior to G4s, but not by a ridiculous margin. Perhaps 10-15% at most...again, in a notebook, performance is not the only factor...both offer good performance and battery life...choose as you will.

[Dont Hurt Me]

Motorola also stated last year that the plan is to double the frequency of the G4 about every 18 months. That essentially means the G4 would move to 2GHz about July 2004. So, expect either a dual-core 2GHz G4 chip or a single 2 GHz G4 processor about that time.LOL :D sorry but thanks for the laugh, I dont see why they didnt just try to pump up the 130 nm process G5 to 2.2 or 2.4? also why not release 970 imacs at say 1.6 & 2.0 . . didnt they get good G5 yields at 130nm?

[thatwendigo]

Does anyone have any real information on the mythical 750vx? All I've ever seen has been posted in this forum. Links????????

A Google search shows only rumor sites at this point. If it's official, then it's buried some way that the search engines don't touch on it.

What I felt the article made clear was that Apple could go for much higher resolution.....

For something that was "made clear," you seem to have a hard time explaining just what the difference is. Maybe you should find the source so that we can register out opinions on it.

As far as reducing the speed of the System Controller and getting rid of the stuff that's not needed.

Interesting. That's about the size of the heat pipe in the eMac, which means that the ASIC's either taking advantage of convection in a much more efficient way, running between 15 and 30 watts, or there's something funky going on.

At some point, we will have to turn our eyes to the main company - Apple - and place just a tiny bit of blame on their processor decisions.

Except IBM hasn't pulled a Motorola on us yet. The G5s are a stronger design than the G4s, they're not jumped-up embedded processors, and they have legs. On top of that, they're intended for use in machines that IBM will be producing, too, so we're going to see push for them.

You also, rather conveniently, leave out that IBM is having trouble with doing something that Motorola never really did for us - technical innovation. The jump to 90nm and adding SSOI are big deals.

LOL :D sorry but thanks for the laugh, I dont see why they didnt just try to pump up the 130 nm process G5 to 2.2 or 2.4?

You wouldn't, given that you don't even understand the basic heat profile of the G5. With the original generation, running at 2.0ghz was over 50watts. That's the hottest processor Apple's used in a long time, if not ever. Pushing the clock up on the same design means decreasing overal speed gain for increasing heat, which Intel has found out with their 90-100 watt chips.

Yes, it probably could be done if they were going the route of the kludge and pray, like Intel does.

also why not release 970 imacs at say 1.6 & 2.0 . . didnt they get good G5 yields at 130nm?

Because they're too hot. Jesus, how many time do I have to say this, DHM?

G4 iMac: G4 @ 1.25 (~15-22watts), 167mhz bus (unknown wattage), PC 2700 RAM (cooler), and PATA hard drives
G5 iMac: G4 @ 1.6 (~30-35watts), 800mhz bus (needs it own heat pipe), PC 3200 RAM (hotter), and SATA hard drives (faster typically = hotter)

Never mind the cost... Everything you just put in the case made the heat jump.

[pgwalsh]

You also, rather conveniently, leave out that IBM is having trouble with doing something that Motorola never really did for us - technical innovation. The jump to 90nm and adding SSOI are big deals.

You wouldn't, given that you don't even understand the basic heat profile of the G5. With the original generation, running at 2.0ghz was over 50watts. That's the hottest processor Apple's used in a long time, if not ever. Pushing the clock up on the same design means decreasing overal speed gain for increasing heat, which Intel has found out with their 90-100 watt chips.


So perhaps they should have implemented SSOI on 130nm and bumped the processor speed while getting the 90nm SSOI process right. Of course hindsight is 20/20, but we're talking big dollars here. I'd imagine SSOI on 130nm process would bring the heat down some and possibly allow higher clock rates for the super-sized heatsink PowerMacs and the ability to utilize the lower clocked 1.6/1.8 in the iMac. :cool:

[Bengt77]

Sorry, don't agree. I do believe that the pentiumM and the Centrino chipset are superior to G4s, but not by a ridiculous margin. Perhaps 10-15% at most...again, in a notebook, performance is not the only factor...both offer good performance and battery life...choose as you will.

Ummmm ... I think that's exactly what he meant too. Maybe the sarcasm was too subtle for some, but I think it was still quite obvious. Though, if I'm wrong and he actually meant the G4 was superior, then please correct me. :(

[ingenious]

I assume you know nothing about PC laptops. First off, this "luggable" nonsense has been false since the Centrino was released. My Centrino machine is the same weight and nearly as thin (like 1.25") as an Apple Powerbook - in short, it is every bit as portable. Second, the screen on my laptop absolutely wipes the floor with the junk Apple puts on its laptops - mine is a 14" screen that does 1400x1050 resolution. You'd have to get a lunch tray 17" powerbook to get that kind of screen space from Apple (and for me, screen space was really the deciding factor). As for speed, the Pentium M (the processor part of Centrino) is clearly faster than a G4 -- I do take your point however that for standard laptop tasks, this may not be noticeable. Price wise, PCs win. As for physical build quality, Apple gets the nod - contrary to popular belief, my DELL is quite solid (and its a lattitude "business model", so its silver and black, not that fisher price blue nonsense they put on inspirons), but it isnt brushed aluminum either.

So, in conclusion, both the Centrinos and Powerbooks have advantages and disadvantages. Either is a solid choice in my opinion. I just had to respond though, because I get the feeling people haven't touched a PC laptop since the centrinos were released - so they love to talk about what junk they are.

PS: I am not a "PC Troll" -- I am writing this on my first mac, a PM G5 1.8 single.

well, you're starting to sound like one! if you don't like macs, then dont come to "Mac"rumors! Dells seem too flimsy for me! when i touch my cousins, it feels like im going to break right thru the PLASTIC! Centrinos seem to be underclocked from what ive read... is that true? i dont use those ugly horrid frustrating things if it dont have to, sorry! :P

[teksmex69]

well, yes. currently osx ui elements seem to have fixed pixel dimensions, meaning menubar on 12" ibook is smaller than menubar in 14" ibook,


XP has that problem as welll. (I'm typing from it right now, with a desktop of 1600x1200 on a 16" laptop display)

Various desktop environments, most popularly run under x windows, DO have scalable elements. 1600x1200 or higher on a notebook looks awesome when your entire desktop scales correctly.

People who state that there's no need for higher resolution notebook displays, because the desktop environment elements will become too small to be comfortably seen, are simply making excuses for their OS. (cough, windows, cough, mac)

[MrSugar]

........................


Dell Inspiron 8600
Pentium-M 1.5ghz
XP Pro with Plus! and Digital Media
512MB RAM 1-DIMM
ATI Radeon 9600 Mobile 128MB
4x CD/DVD-R with RecordNow! and MyDVD Deluxe
80GB HD
Dell Wireless 1450 Internal 802.11
Dell TrueMobile Blutooth
McAffee Security Center with 1yr Enrollment
Dell Jukebox Plus
Dell Picture Stuiod with Photo Album Premium
Cost: $2,686

Apple PowerBook 15"
Motorola G4 1.5ghz
MacOS X 10.3.3
512MB RAM 1-DIMM
ATI Radeon 9700 128MB
SuperDrive
80GB HD
Airport Extreme
Blutooth
iLife
Cost: $2,599



I have. So what?

Dude, why is it that I always love your posts. This was beautiful, thank you for putting everything into perspective. Wow... nice work.

[thatwendigo]

Dude, why is it that I always love your posts. This was beautiful, thank you for putting everything into perspective. Wow... nice work.

There's two things I didn't directly point out. One is that, for about $120, you can get a 7200 RPM drive in the Dell (something I think should be an option on the PowerBooks), and that the PowerBook has a better graphics card than Dell. Take a look... Radeon 9700 vs. a Radeon 9600, with the same VRAM.

Also, thanks for the kind words. :D

[pigwin32]

Motorola also stated last year that the plan is to double the frequency of the G4 about every 18 months. That essentially means the G4 would move to 2GHz about July 2004. So, expect either a dual-core 2GHz G4 chip or a single 2 GHz G4 processor about that time.
...did they provide any information about making the fsb faster than 167MHz...

[pgwalsh]

Pentium® M Processor,1.7 GHz, 15.4 UltraSharp™ Wide Screen UXGAPentium® Microsoft® Windows® XP Professional
1GB DDR SDRAM 2 Dimms
128MB DDR ATI's MOBILITY® RADEON™ 9600 PROTURBO
80GB Hard Drive
4x CD/DVD burner (DVD+RW/+R)
Dell® Wireless 1450 Internal Wireless (802.11a/b/g, 54Mbps)
Productivity Pack including WordPerfect® and Money®
Dell Jukebox powered by MUSICMATC

$2715.00

1.5Ghz PowerPC G4
OS X 10.3
Radeon 9700 128
80GB Ultra ATA 5400RPM
SuperDrive
Airport Extreme
Bluetooth
ILife

$2999.00

Once you upgrade the RAM, Hard drive, and video card, your paying a bit more... Actually if you just don't upgrade the Ram, you're doing quite well with Apple. I'd still buy the PowerBook.. but not with Apples memory offers.. Jackasses.

To note.. The Dell is faster has a bigger and better screen and the wireless card works on all three a/b/g.. Smooth. But's it's a friggin dell.

[thatwendigo]

Once you upgrade the RAM, Hard drive, and video card, your paying a bit more... Actually if you just don't upgrade the Ram, you're doing quite well with Apple. I'd still buy the PowerBook.. but not with Apples memory offers.. Jackasses.

That's your mistake. Don't buy RAM from Apple. You also left out about $300-400 in software to equal the security, innate features, and other advantages of using OS X.

To note.. The Dell is faster has a bigger and better screen and the wireless card works on all three a/b/g.. Smooth. But's it's a friggin dell.

It also doesn't use OS X, which costs it in smoothness. I'm not going to argue the Cenntrino being faster at higher clock, and that the screen is probably a bit better, but the addition of 802.11a is kind of a wash.

[gensor]

Don't laugh, I know they are the worst.

April 22: Reader Mail

David K. Every writes: Fabrication is a multi-stage process, with each stage taking time.

You can be making a ton of chip at once, but they've got to go through all the steps before getting to the other side.

So there are huge latencies (as long as 10 weeks) between starting a run, and getting finished chips out the other side.

And manufacturers usually want a significant number of chips in reserve before even considering making a model on them (more latency while that reserve is manufactured).

If something has a problem, it can stop this manufacturing pipe, and take long time to get ramped back up, and you may need to flush your reserves (throw away the bad chips).


So little problems in a single stage (like final packaging) can take weeks (months) to discover, then stall the pipe, flush the cache (reserve), and make you start over.

Two minor problems could mean many months.

Interestingly enough, IBM sources have recently echoed much of what David writes above; apparently, although the PowerPC 970FX packaging delamination issues we've reported recently were relatively simple to fix, they caused tremendous delays due to the way semiconductor ramp-ups are carried out.

Thankfully, it looks like we're getting close to the end of the tunnel; IBM sources now report that 970FX yields are strong, and the 975 is soon to begin ramping up for full-blown production in about three weeks' time.


My comment: What Apple product uses the 970FX vs the 975?

[Sun Baked]

...did they provide any information about making the fsb faster than 167MHz...The were supposed to switch over to RapidIO along with the integrated memory controller.

But there were rumors that Motorola canceled their G5 project -- of which the 7457-RM (RapidIO + Memory Controller) most likely would have fit.

What is worse, is that after the cancelation rumor -- Motorola began talking widely about adding RapidIO and integrated memory controllers to all their PPC product, with zero news about doing this for the chips Apple uses.

So yes there is news about forward movement for Motorola's PPC and e-Book line, just not chips Apple would use.

Considering we used to get real good info about future product about a year out, lack of info insn't good.

[Phinius]

LOL :D sorry but thanks for the laugh, I dont see why they didnt just try to pump up the 130 nm process G5 to 2.2 or 2.4?

Who's they? If you mean IBM, well IBM lists the topend speed of the 970 as 1.8 GHz. Seems the 970 was pushed to 2 GHz for Apple's exclusive use. If you believe that the 130-nm process could achieve 2.4 GHz, well the Power4 has only gone to 1.7 GHz on that process size and the 970 is based on that core architecture.

also why not release 970 imacs at say 1.6 & 2.0 . . didnt they get good G5 yields at 130nm?

IBM's own estimates are that the 1.8 GHz 970 uses an average of 42 watts. The small, compact iMac processor box is not designed for dissipating such high amounts of heat. If Apple had plans to move the iMac to the G5, it was wise for them to wait until the processor moved to a smaller process size a few months after the first PowerMac G5 was available. Unfortunately, IBM has had a few delays in providing a adequate supply of 970FXs, so that maybe a reason why the iMac has not had a update recently like the eMac, iBook and PowerBook.

[jade]

I find that extremely unlikely. In System Preferences, you can adjust the font smoothing by degrees (with recommendations for monitor type). On top of that, all you need for an active demonstration of how silly a claim this happens to be is to go to the Dock preferences. See the Dock Size slider? Play with it. For extra fun, turn magnification on and watch icons smoothly shifting in size.

Sure looks like a "simple scaling" problem to me. :rolleyes:

We will have to go back to disagreeing. On a PC you have the broad option of making your settings larger or smaller without effecting resolution. ie like the Mac os x zoom feature but it effects everything from icon size to menu type to dialog boxes. On a mac you have to change each individual componenet, and not all components are changeable. Windows has desktop themes like small medium and large. Potential, with Apple's superior rendering engine, this should work better, but it is currently limited by software. That is what the other posters are getting at.

It takes a lot more work to adjust your global settings to match on a mac.


Next up PC laptop screens

Powerbook screens have fallen behind in brightness and veiwing angle. PC makers like Sony and Fujitsu, as well as LCD monitor makers like Sharp and NEC-Mitsubishi are using LCDs that offer superior crispness and viewing angles more on par with high definition TV. Next time you go to best buy check out the sony xbrite screen on the high end notebooks or on Sony's wacky all-in-one. It is a mazing how much brighter and sharper these screens look. The resution on powerbooks tends to be a little low, but that isn't always a concern, bur crispness and brightness are important to most users across the board.


And those centrino notebooks:

Well the ones with the larger screens typically have a longer real-world battery life than the powerbooks. So they win on the battery count (about 20-30% more. The only apple notebook that truly tops the Centrinos in battery life is the ibook. Clock for clock the centrino varient of the p4 is faster than the g4 powerbooks, but these new upgrades should put the speeds at the top end very close to the rated speeds of those centrino notebooks.

thatwendigo: I know you like to use Dell for comparison, but I'd like to counter, that the powerbooks are very price competitive with the high-end PC notebooks, where Apple really misses is at the $1000-$1500 midrange notebook lineup. In this price range you can find a wide variety of fairly small widescreen notebooks, as well as larger screens for the consumers than the ibook offers. But with every ibook revision Apple is getting closer to providing a compelling option for PC users shopping in this price range.

[cb911]

people have actually been blaming Apple for the late arrival of the G5 chips? :confused: i must have missed that one.

well as far as i can remember i've only ever blamed the chip manufacturers. like when Moto was coming out with the new 1.2GHz G4 (i think) chip for the PowerBooks they were having terrible problems with low yields. i actually remember that at that time i seriously thought that Apple would be in serious trouble because of it. people were specualting that Apple would use a low-power version of a G5 instead. also at that time Steve Jobs himself was very angry at the suituation Moto had left them in. i'll always remember him calling Motorola 'Scum-o-rola'. :D

well Moto has now sorted out their issues, as we can see with the release of the new 1.5GHz G4 PowerBooks. no doubt IBM will work through this. lets just hope the situation doesn't get really bad, and people don't start making jokes and funny names for IBM. :D

[JFreak]

IBM's own estimates are that the 1.8 GHz 970 uses an average of 42 watts. The small, compact iMac processor box is not designed for dissipating such high amounts of heat.

who says imac cpu should run at the same clock speed as the powermacs do? smaller clock speed equals less heat, so imagine a 1.2GHz imac G5 having 600MHz bus. i'd be amazed if G5@1.2GHz would use more than 20W.

[Phinius]

You also, rather conveniently, leave out that IBM is having trouble with doing something that Motorola never really did for us - technical innovation. The jump to 90nm and adding SSOI are big deals.

Motorola came up with Altivec, that alone is substantial innovation. There is also adding an L3 cache to the G4, bumping up it's pipeline stages (which IBM has never done to the G3), SOI, low-k dielectric and of course the most overlooked of all-more than tripling the frequency of the 7400 G4-moving it from 400MHz to 1.5GHz since 1999.

The G4 is not at a deadend, Motorola and Apple both have publicly stated that the G4 will continue to be improved. Next up, the G4 will move to a smaller process size, with a much faster bus speed, on-board controller, RapidIO chip interconnect, dual core, and a topend frequency of 2GHz. Those improvements should be coming about July. After that, Motorola has architectural improvements on their roadmap which will bring the G4 up to 3GHz (likely at the 65-micron process shrink in 2005). That would probably mean bringing the pipeline stages from the current 7 to at least 10.

[Phinius]

So perhaps they should have implemented SSOI on 130nm and bumped the processor speed while getting the 90nm SSOI process right. Of course hindsight is 20/20, but we're talking big dollars here. I'd imagine SSOI on 130nm process would bring the heat down some and possibly allow higher clock rates for the super-sized heatsink PowerMacs and the ability to utilize the lower clocked 1.6/1.8 in the iMac. :cool:

IBM had planned on implementing SSOI on the 65nm process size. They pushed it up to the 90nm process size.

Spliting the Power4 to make a 970 would have been impractical at the 180-nm process size. It would have been too large and expensive, not be mention the Power4 uses at least 125 watts.

At each process shrink the PowerPC chips are beginning to look much more competitive. The Intel X86 designs from Intel are having more and more problems with excessive heat loss. Even on battery power the default frequency for all Pentium-M's is 600MHz and the Pentium4-M is 1.2GHz. Not to mention the desktop Pentium 4 which is barreling towards a topend 125 watt use next year.

[potterfast]

http://www.apple.com/motion/

Can anyone identify the monitor in the little G5 picture in the center?

[Spazmodius]

This PDF http://www.motorola.com/mot/doc/0/786_MotDoc.pdf is from a presentation Moto did back in June 2003. They reckon the G4 is going all the way to 2Ghz and Dual Core. I'm not sure exactly what they mean with DDRI and DDRII built in - I presume they are adding a memory controller which would cure the current problem with the G4 MPX bus maxing out at 167Mhz. That would be nice. RapidIO isn't going to be too popular with Apple I'd guess after they've plumped for HyperTransport in the G5.

Anyway, that would still make for a pretty damn fine laptop even if it came out in Jan 05. I'd suspect a 2Ghz G4 with built in memory controller would wipe the floor with even a G5 2Ghz bearing in mind the G4's shorter pipelines.

Could you imagine the uproar if Apple did come out with another G4 Powerbook after this one though from the "G5 or I'm buying a Dell" crowd?

Wow, dual-core G4 would rock, so long as it was energy-efficient enough to put in a Powerbook. If not, it might make a nice upgrade to iMacs and eMacs (or whatever the descendant of these boxes would be).

Could you elaborate on the DDRI vs. DDRII speculation? I'm not up on the jargon, and I wondered if you think a faster FSB frequency is inevitable, given the content of the Moto roadmap.

[avkills]

Motorola came up with Altivec, that alone is substantial innovation.

You need to check your research again, IBM, Apple and Motorola all developed the G4s Vector Processing Functions as a team. Motorola calls it Altivec. IBM calls it VMX. Apple calls it Velocity Engine. IBM and Apple developed the optimized Vector Processing for the G5.

-mark

[thatwendigo]

We will have to go back to disagreeing...It takes a lot more work to adjust your global settings to match on a mac.

I stand corrected, then. I haven't done much with altering sizing of elements in XP, but I've also found that OS X at high resolution doesn't bother me at all. My desktop is at 1280x960, and I like it there, but most of my family prefers their machines down lower than that.

Powerbook screens have fallen behind in brightness and veiwing angle. PC makers like Sony and Fujitsu, as well as LCD monitor makers like Sharp and NEC-Mitsubishi are using LCDs that offer superior crispness and viewing angles more on par with high definition TV. Next time you go to best buy check out the sony xbrite screen on the high end notebooks or on Sony's wacky all-in-one. It is a mazing how much brighter and sharper these screens look. The resution on powerbooks tends to be a little low, but that isn't always a concern, bur crispness and brightness are important to most users across the board.

Fair enough, then.

Well the ones with the larger screens typically have a longer real-world battery life than the powerbooks. So they win on the battery count (about 20-30% more. The only apple notebook that truly tops the Centrinos in battery life is the ibook. Clock for clock the centrino varient of the p4 is faster than the g4 powerbooks, but these new upgrades should put the speeds at the top end very close to the rated speeds of those centrino notebooks.

I'd buy that the Centrino is a better battery manager, largely because it's a vastly newer design. Unlike the G4, it's intended as a computer processor for mobile applications, and so Intel built some things into it that make it act that way. Apple's done a damn fine job of making the G4 work in the real world, but there's only so much you can do without full support.

thatwendigo: I know you like to use Dell for comparison, but I'd like to counter, that the powerbooks are very price competitive with the high-end PC notebooks, where Apple really misses is at the $1000-$1500 midrange notebook lineup. In this price range you can find a wide variety of fairly small widescreen notebooks, as well as larger screens for the consumers than the ibook offers. But with every ibook revision Apple is getting closer to providing a compelling option for PC users shopping in this price range.

I agree. Apple's weak point is the midrange, because they don't compete at the lowend (except with the eMac), and that's traditionally been a serious issue. However, given that the 970 is only likely to get cheaper in volume as IBM refines the process, and perhaps even cheaper when the technology is applied to the 975/980 successor chips, I think that we might see some even better options in the consumer line soon.

However, it's going to take time. As I keep telling peopel, the PowerPC is a more expensive platform just at the basic levels of boards and chips. If enough others adopt the standard, this will almost certainly come down, and the AMD and Sony deals are likely just the tip of the iceberg.

[QUOTE=Phinius]Motorola came up with Altivec, that alone is substantial innovation. There is also adding an L3 cache to the G4, bumping up it's pipeline stages (which IBM has never done to the G3), SOI, low-k dielectric and of course the most overlooked of all-more than tripling the frequency of the 7400 G4-moving it from 400MHz to 1.5GHz since 1999.

No, Apple, IBM, and Motorola worked together on the SIMD units in the PowerPC architecture, and each used a different name. The only 'AltiVec' invented by Motorola is their brand name for a joint technology design. Higher pipeline stages only make for faster staging of the clock in revisions. In terms of the design of the PowerPC, the Deep and Wide approach is more efficient and less error-prone. SOI is not a Motorola invention, and they really ought to have done more with the frequency.

Incidentally, Tom's Hardware (http://www6.tomshardware.com/cpu/20040201/prescott-01.html) shows a timeline of Intel and AMD jumps in the same timeframe.

Motorola: 400mhz-1.5ghz
Intel: 1.5ghz-3.4ghz
AMD: 1.0ghz-2.4ghz

Unlike Motorola and Intel, AMD actively improved their designs while also scaling the clock. We had to go to IBM to get the same effect.

The G4 is not at a deadend, Motorola and Apple both have publicly stated that the G4 will continue to be improved. Next up, the G4 will move to a smaller process size, with a much faster bus speed, on-board controller, RapidIO chip interconnect, dual core, and a topend frequency of 2GHz. Those improvements should be coming about July.

You believe Motorola's roadmap? After all this time, you trust a press release from them?

[Phinius]

However, given that the 970 is only likely to get cheaper in volume as IBM refines the process, and perhaps even cheaper when the technology is applied to the 975/980 successor chips, I think that we might see some even better options in the consumer line soon.

If the successor to the 970 is based on the Power5, then it's likely to be a larger and more power hungry chip (since the Power5 is about 25% larger than the Power4 and uses about 25% more watts) and therefore more costly to make than the 970.

As I keep telling peopel, the PowerPC is a more expensive platform just at the basic levels of boards and chips.

The PowerPC is more expensive for Apple due to the higher costs involved with having to develop so much of their own hardware and selling it at a much lower volume than Intel does.

If enough others adopt the standard, this will almost certainly come down, and the AMD and Sony deals are likely just the tip of the iceberg.

I'm not aware of any statement from IBM, AMD or Sony that says the 970 will be used by AMD or Sony. Only rumor sites have stated such.

No, Apple, IBM, and Motorola worked together on the SIMD units in the PowerPC architecture, and each used a different name. The only 'AltiVec' invented by Motorola is their brand name for a joint technology design.

Well, where was IBM's Altivec or VMX implementation for Apple when Motorola came out with Altivec in 1999? Why didn't the great and powerful IBM wait until 2003 to make a Altivec processor for Apple's use if the company was so much better than Motorola? Is that somehow making better PowerPC advancements than Motorola in that time period?

How about IBM's failure to match or beat Motorola with SMP or L3 cache capabilities on a PowerPC processor for Apple's use.

Or why is it that IBM never showed their superior capabilities by matching or beating the frequency advances that Motorola was making with the G4? The G4 is at 1.5 GHz and the G3 is at 1.1 GHz.

The IBM's G3 has never advanced beyond 4 pipeline stages and yet Motorola invested the money to move the G4 to 7 pipeline stages.

Higher pipeline stages only make for faster staging of the clock in revisions.

More pipeline stages makes for greater frequencies and as a result higher performance. There is however a point of diminishing returns where it's not worth it. An Intel research paper stated that the point of diminishing returns was at about 50 pipeline stages, an IBM research paper stated it was at about 20 pipeline stages.

In terms of the design of the PowerPC, the Deep and Wide approach is more efficient and less error-prone.

Your just pulling this stuff out of a hat aren't you. The Power4 and 970 are very error prone.

SOI is not a Motorola invention, and they really ought to have done more with the frequency.

Motorola more than tripled the G4 frequency in 4 1/2 years. Compare that to Intel increasing the Pentium 4 frequency by about 2.25 times in 3 1/2 years. Raising the frequency is only one of many ways to get performance. Adding more cache is another way and putting two processors on a chip yet another method.

Also, the fastest processor that Intel makes runs at 1.5 GHz and it's called Itanium, not Pentium. So clock frequency is not necessarily a good way of judging a processors ability when comparing two different designs.

a timeline of Intel and AMD jumps in the same timeframe.

Motorola: 400mhz-1.5ghz
Intel: 1.5ghz-3.4ghz
AMD: 1.0ghz-2.4ghz

Unlike Motorola and Intel, AMD actively improved their designs while also scaling the clock. We had to go to IBM to get the same effect.

Hogwash. Motorola, Intel and AMD all actively improved their designs while scaling the clock. Motorola moved the bus speed up, added SMP, L3 cache, onboard L2 cache, temperature diode, power management features and of course Altivec. Motorola also added SOI and a low-k dielectric to their chip making process.

Meanwhile IBM basically split the existing dual-core Power4 in two to come up with the 970. IBM also added a less capable VMX or Altivec to the 970 than what Motorola uses on the G4. Seems IBM was in a hurry to get the project done.

IBM did not create a whole new chip architecture for Apple's exclusive use. Nor would any chip manufacture make a design solely for Apple's use in a shrinking marketshare. Yet people expected Motorola to have come up with a high watt usage processor design that they couldn't hope to sell anywhere but to Apple.

Realisticly Motorola's plans have always been to create processors for Apple's use that could also be sold in the embedded market (in fact Motorola's website explains that host processors are introduced in computers first and work on down through the embedded market). Which means keeping the power use down. There were obvious setbacks for Motorola's 2001 roadmap timeframe (still on Motorola's website). Details of the upcoming dual-core G4 design were mentioned for introduction at the .13-micron level and not at the .090-micron process size (perhaps even more pipeline stages). This is basically what the so-called Motorola G5 was going to be. Yet, it would never intended to be a Pentium 4 trump card. Apple realized that and started working with IBM close to 4 years ago to bring out a PowerPC version of the Power4 architecture. The 970 simply brings Apple up to a whole new level of performance. It is not meant to be a complete replacement for a low power use chip like the G4. Lower power use is not it's foretay, much like the Pentium 4 does not make an ideal processor for portable use.

You believe Motorola's roadmap? After all this time, you trust a press release from them?

That pdf file is from a Motorola presentation at a developers conference last year. It is not in any of the companies press releases.

To jump to the conclusion that Motorola is planning on or has stopped ongoing development of the G4 flies in the face of statements to the contrary from both Motorola and Apple. Yes, I stated Apple. An Apple spokesperson recently stated that Apple has no intentions of dropping the G4 in the near future and they were glad that Motorola is continuing to develop it. The 7457A that is found in the PowerBook and it's use in the iBook are recent examples of that. Using your sense of jumping to unfounded conclusions...has IBM stopped development on the 970 since there has been no more speed advancements since August of last year.

[jade]

I stand corrected, then. ....
Fair enough, then. ....

I'd buy that the Centrino is a better battery manager, largely because it's a vastly newer design. .....

I agree. ......


We agree again.


Since IBM agreed to open up the powerpc, Sony and AMD have decided to work on it in conjunction with IBM. This announced at their recent conference, and can be found in quite a few news sites. Where this will lead is of course unclear, but I agree with thatwendigo, that this may bode well for Apple.

Unless Moto makes serious revisions to the g4, it will continue to fall behind the PC counterparts.. At this point in tim it is still substandard in DDR support and FSB....I'l believe it when I see it in the case of Moto and IBM. In the past IBM has been more reliable about keeping their prmises.

[thatwendigo]

You know, I had a whole long reply to this, but I just deleted every word of it, because I'm not going to pander to your insulting and demeaning tone. Most of what you said I already know, and most of what you disagree with in my posts just shows that you either didn't read all that carefully, implied things I didn't say, or otherwise couldn't be bothered to at all fairly represent what I said.

The G4 is not being developed for Apple. It's being developed for Motorola, just as the 970 is being developed for IBM. In both cases, they are CPUs that are created for specific purposes, sometimes with Apple's input on the design, and which have largely been bought in volume by our favorite computer company. That being said, the G5 is not a low-power chip, and if you at all paid any attention to my general attitude about it, you'd know that I don't want there to be a portable based on it for the simple reasons of battery and heat issues. However, it's far easier for you to assume, belittle, and misstate, so I'll leave you to that.

Whether or not the G4 is still being developed (I fully believe that it is), I find it extremely unlikely that Motorola has magically gotten their act together. They're promising a lot of things when they couldn't even deliver simple bumps or proper quantities in the past.

[avkills]

In 1999 I don't think IBM cared about SIMD because the benefit to servers is about zero. IBM was making the low-end for Apple, so I don't get the ribbing at IBM for not automatically implementing SIMD (VMX). :confused:

Apple wanted IBM for the next Pro chip and as far as I am concerned they delivered. They will solve the 90nm problem soon enough. I seriously doubt IBM threw billions of dollars into a plant to fail.

-mark

[SiliconAddict]

Motorola also stated last year that the plan is to double the frequency of the G4 about every 18 months. That essentially means the G4 would move to 2GHz about July 2004. So, expect either a dual-core 2GHz G4 chip or a single 2 GHz G4 processor about that time.

With what?? A 200Mhz system bus?!?! I don't know about everyone else but the damn FSB needs to be addressed by Moto else what the hell is the point of a 2Ghz (fire hose) and a 333Mhz+ memory bus (garden hose) going through a 167Mhz+ FSB (straw). OS be damned at this point. This is a hardware issue.
If Moto addresses this and breaths some new life into the PowerBook line, then the G5 PowerBook can come out in 2038 for all I care. For now there are massive bottlenecks in the PowerBooks overall architecture that has yet to be addressed by Moto.
And I wish to GOD people would stop making the excuse that the OS makes up for any hardware deficiencies. Stop the BS please. I don’t care about Windows, OSX, Linux, Amiga, DOS, BSD, or any other OS. They are all tools for all I give a crap. But tools are dependant on the hardware they run on and the PowerBook and the current crop of G4’s are aging slag. Admittedly very pretty slag. But slag nonetheless. If I’m offending someone with this description of your computer sorry. But looking at this from the outside, someone who WANTS to be looking at this from the inside one day, and having only a minimal vested interested in Apple I’m not seeing a pretty picture being painted by Apple. In fact the mural looks suspiciously like the PowerMac’s pre-G5. The overall system is being strangled and unless some architectural changes are applied to the PowerBook line adding a 2Ghz, 3Ghz, or even a dang 5Ghz CPU isn’t going to make a drastic difference. Flame away all you want guys but I haven’t stated anything that isn’t true. Now you will excuse it Its 4AM and I need some less controversial sleep. With my luck I’ll dream of G5 PowerBooks. :p

[SiliconAddict]

Whether or not the G4 is still being developed (I fully believe that it is), I find it extremely unlikely that Motorola has magically gotten their act together. They're promising a lot of things when they couldn't even deliver simple bumps or proper quantities in the past.

Oh one last question I've been meaning to ask. Please point out what other products use the G4? Seriously. I really want to know.

[thatwendigo]

Oh one last question I've been meaning to ask. Please point out what other products use the G4? Seriously. I really want to know.

http://www.cetia.com/powernode3.asp
http://www.vistacontrols.com/PR_011112.php
http://www.elecdesign.com/Articles/ArticleID/5119/5119.html
http://www.vita.com/vmeprod/company/aitechinfo.html
http://www.vita.com/vmeprod/company/thalesinfo.html
http://www.synergymicro.com/products.html
http://www.spectrumsignal.com/Product_selector/product_search/search_results.asp?category_id=477&category_type_id=45&by=feature&category_name=PowerPC+%28Floating%2DPoint%29&category_type_name=Processor
http://www.dy4.com/products/index.htm
http://www.mangodsp.com/condor_cpci.asp
http://www.eqware.net/Products.html
http://www.vistacontrols.com/product_PPCG4C.php

As I said before... It's still pretty big in the embedded, blade, and other markets.

[thatwendigo]

With what?? A 200Mhz system bus?!?! I don't know about everyone else but the damn FSB needs to be addressed by Moto else what the hell is the point of a 2Ghz (fire hose) and a 333Mhz+ memory bus (garden hose) going through a 167Mhz+ FSB (straw). OS be damned at this point. This is a hardware issue.

Actually, I agree with you for once. The G4 has needed a better bus for quite a while. Frequency pushing only goes so far when you're not keeping the processors fed. The G4 would have done a lot better, and shown its age more gracefully, had Motorola just given us a better FSB.

If Moto addresses this and breaths some new life into the PowerBook line, then the G5 PowerBook can come out in 2038 for all I care. For now there are massive bottlenecks in the PowerBooks overall architecture that has yet to be addressed by Moto.

Screw the G4, unless they're going to make it somewhat equivalent to the still-unannounced 750vx. We need the FSB and memory clock to go along with the processor bumps. Something low power and quick would fit ever so nicely, and hopefully not be as excessively restrictive as the G5 is likely to be.

In fact the mural looks suspiciously like the PowerMac’s pre-G5. The overall system is being strangled and unless some architectural changes are applied to the PowerBook line adding a 2Ghz, 3Ghz, or even a dang 5Ghz CPU isn’t going to make a drastic difference.

You're right on this and I'm just not touching the section right before it, but there really isn't a whole lot Apple can do without either a new chip or some amazing redesign. I trust Apple to come up with something, but I hope that it's a wow product and not another compromise.

Its 4AM and I need some less controversial sleep. With my luck I’ll dream of G5 PowerBooks. :p

Don't worry. That'll stop this Tuesday. ;)

[Phinius]

With what?? A 200Mhz system bus?!?! I don't know about everyone else but the damn FSB needs to be addressed by Moto else what the hell is the point of a 2Ghz (fire hose) and a 333Mhz+ memory bus (garden hose) going through a 167Mhz+ FSB (straw). OS be damned at this point. This is a hardware issue.

If you would look at the above pdf of Motorola's developer presentation in June of 2003, you will see that Motorola has plans for DDRII. That would essentially mean a bus speed of at least 400MHz.

For now there are massive bottlenecks in the PowerBooks overall architecture that has yet to be addressed by Moto.

Motorola added a L3 cache to the G4 in January of 2001. That reduces the number of times the process will go to the much slower main memory and therefore increases the speed that the processor is fed. Unfortunately, most people don't seem to realize that a much bigger cache is yet another another way to speed up the amount of work that the processor accomplishes in any given amount of time. That is why Intel has 1 MB of L2 cache on the Pentium-M and also why Itanium has 6MB of onboard L3 cache and why the Power4 uses up to 64 MB of L3 cache.

The publics focus of frequency and bus speeds as the major determining factors in processor speed is probably a significant reason why IBM and Apple kept the L2 cache the same size when the 970 moved to the smaller .090 micron process size. Keeping the L2 cache small also makes for a smaller die size and therefore less expensive chip manufacturing costs. It also reduces the power use somewhat. That also points towards Apple's interest in incorporating the G5 chip into the PowerBook.

Apple is at least partly to blame for the implementation of the G4 in PowerMacs. The effective bus speed could have been doubled if Apple would have used a dual channel memory setup.

Also, Apple dropped the use of L3 cache in the PowerBooks. That would have sped up the performance, but again, most buyers don't seem to realize the benefit and therefore Apple discontinued its use due also to its added cost.

The overall system is being strangled and unless some architectural changes are applied to the PowerBook line adding a 2Ghz, 3Ghz, or even a dang 5Ghz CPU isn’t going to make a drastic difference.

Apple seems headed towards a G5 in PowerBooks. With Motorola headed towards a higher power use dual-core G4 in the next process shrink, I don't see the next G4 as something Apple would use in the interim until a G5 PowerBook arrives. It seems more likely that Apple will introduce a G5 PowerBook in the next few months, after all, Mac users overwhelmingly want one and therefore Apple is compelled to make it.

Motorola is moving the G4 to DDRII in the next process shrink scheduled for 2004. That means a bus speed of at least 400MHz. Maybe more if Apple sees fit to use a faster bus than that.

There will be a 90-nm PowerPC G4 processor in 2004, made on 300-mm wafers, according to the chief technology officer of Motorola:

http://www.siliconstrategies.com/article/printableArticle.jhtml?articleID=10802061

The G4 will double in frequency about every 18 months and hit 2 GHz according to Motorola's Dirk Wristers:

http://www.siliconstrategies.com/article/printableArticle.jhtml?articleID=10802309

It's obvious that the next process shrink of the G4 will obtain 2GHz. That's due to about a 33% speed improvement that Motorola got moving from 18-nm to 13-nm. The G4 tops out at 1.5GHz on a 13-nm process shrink and if you add a 33% increase moving to .90 microns, then that is about 2GHz.

[SiliconAddict]

It's obvious that the next process shrink of the G4 will obtain 2GHz. That's due to about a 33% speed improvement that Motorola got moving from 18-nm to 13-nm. The G4 tops out at 1.5GHz on a 13-nm process shrink and if you add a 33% increase moving to .90 microns, then that is about 2GHz.


Phinius,
I appreciate the insight but lets be pragmatic about Moto and release schedules they are about as reliable as the weather. Even if Moto dropped a working prototype of a newly designed G4 (G4.5? ;) ) at some developer conference I've still be highly skeptical. I don't trust anything said by Moto until actual units are being shipped out to Apple. They've lost too much creditability IMHO.

[Phinius]

Phinius,
I appreciate the insight but lets be pragmatic about Moto and release schedules they are about as reliable as the weather. Even if Moto dropped a working prototype of a newly designed G4 (G4.5? ;) ) at some developer conference I've still be highly skeptical. I don't trust anything said by Moto until actual units are being shipped out to Apple. They've lost too much creditability IMHO.

Motorola's last four 7400 series of PowerPC processors have arrived annually like clockwork:

7410-January, 2000-533 MHz
7451-January, 2001-733 MHz
7455-January, 2002-1 GHz
7457/47-February, 2003-1.33 GHz

The next 7400 series has not arrived yet, but in the interum Motorola announced the 7447A which tops out at 1.5 GHz. Judging from the June 2003 statements that I have links to in the previous post, it seems that the next major update was scheduled to appear sometime around midyear-to-late 2004. I would not be surprised to see it appear in July of this year.

There seems to be a lot of distorted beliefs about Motorola's recent ability to deliver upgrades in a timely manner. Motorola has not kept up with Intel on the desktop, but then again, the PowerPC processors that Motorola makes for Apple are designed for use in the embedded market also, and not aimed exclusively at any of Apple's market segments. Likewise, IBM's 970 is a derivitive of a processor that is made for the server market and not exclusively aimed at the desktop or notebook market. Although IBM is likely keeping Apple's needs in consideration when designing future versions of the Power series of processors.

Motorola has to keep improving the host PowerPC processors in order to stay abreast of competitors in the embedded market. Luckily the embedded markets that use the 7400 series of PowerPC processors are also clammering for speed improvements that are catching up to what is demanded in the notebook market. That is low power use, yet much higher performance.

As for the 7400 chip being outdated, well, Intel's Israel design team used the Pentium III architecture as the basis for the Pentium-M simply due to it's efficient watt use-to-performance ratio. The next two versions of the 7400 G4 will be getting major updates in much the same way as the Pentium III did when it morphed into the Pentium-M (the current G4 is also comparable in performance to the Pentium III) After the boost to 2 GHz, with a on-board DDR/DDR-II memory controller and RapidIO, comes a .065-micron version in 2005/6 that will put it at 3 GHz. This should bode well for Apple using the G4 in small computer boxes where power use is very important.

Motorola (chip business being spun off as FreeScale very soon) will come out with a dual-core 7400 series chip simply due to what is demanded of them from the embedded market. Whether Apple chooses to use it is another matter, although I can't imagine that Apple's needs were not considered when the decision was made to go ahead with a dual-core chip. Motorola may be planning to have dual-core and single processor versions of the next series of 7400 host PowerPC processors.