Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
math today...

Mercury said:
Going from 2Ghz to 3Ghz is a 50% increase, while going from 3Ghz to 2Ghz is a 33.3% decrease.

I do hope you were joking, and that the decrease was for someone else's benefit. Can't really tell sarcasm over the internet.

I guess they need to re-examine the state of math classes in schools today...

It was most certainly not sarcasm on his part. 50% of 2GHz is 1GHz. So, going from 2GHz to 3GHz is a speed increase of 50%. 50% of 3GHz is 1.5GHz - so a 50% decrease from 3GHz would result in a 1.5GHz machine. To go from 3GHz to 2GHz you are seeing a reduction in speed of 1/3, or 33.3%.

2GHz is a 33.3% reduction in speed from 3GHz.





blakespot
 
iswimbikeandrun said:
the jumps from 700->833->1GHZ are around 5%, where as the jumps from 2.6->2.8->3GHz are around 14%, your theory is wrong.
Your math is confusing to me.

Here is what I get:

700 to 833 --> 19%
833 to 1000 --> 20%

2.6 to 2.8 --> 8%
2.8 to 3.0 --> 7%

BTW, not sure where you got 700 or 833 to begin with.

Sushi
 
I would not be suprised by a 3Ghz Processor in the next Power Mac revisions. Remember last year before the Power Mac G5 was announced? we heard Rumors that the G5 would be released @ clockspeeds of 1.4, 1.6, and 1.8Ghz however in the end everyone was suprised by a 200Mhz bump across the line with a 2Ghz max. Right now we are hearing rumors of Clockspeed ranging at 2.2-2.6Ghz for the G5, Apple could very well Surpise everyone (as they did last year) by topping out at 2.6-3.0Ghz Range or at least produce a lineup that runs in 2.4-2.8Ghz range. Does a 1Ghz jump sound that unbelievable? True that's a huge jump in speed but hey we went from 1.42Ghz to 2.0Ghz last year that was a 580Mhz jump!! In fact we actually went from 1.42 to 2.0Ghz in about 6-8 months as the G4 1.42 was annouced in Jan 2003 and the G5 2.0Ghz was annouced/shipped in June/Sept of 2003. Most folks would have laughed at that kind of jump in clockspeed but hey it happened. 1Ghz may be a stretch but it is feasible and would make quite a splash for Marketing. Why can't Apple announce 3Ghz at WWDC they could annouce then but ship in September with say a lineup ranging from anywhere form 2.0-2.8Ghz available immediately at announcement at WWDC.
 
blakespot said:
I guess they need to re-examine the state of math classes in schools today...

It was most certainly not sarcasm on his part. 50% of 2GHz is 1GHz. So, going from 2GHz to 3GHz is a speed increase of 50%. 50% of 3GHz is 1.5GHz - so a 50% decrease from 3GHz would result in a 1.5GHz machine. To go from 3GHz to 2GHz you are seeing a reduction in speed of 1/3, or 33.3%.

2GHz is a 33.3% reduction in speed from 3GHz.

It would help if you guys actually read and understood Mercury's posts.

You are telling him his math is horrible by quoting him explaining what is wrong with iswimbikeandrun's math. It is iswimbikeandrun whose math is pretty messed up.

Here is Mercury's description of the problem:

Mercury said:
I hate to be teaching math, since the highest I took was college algebra, but if you're using your method, you divide at the end by the original speed. So, (833-700)/700, not 833. We're doing increases. That gives you .19. His math is exact; yours is flawed.

Mercury is correct. To get the increase or decrease, you compute ((new speed - old speed) / old speed) * 100. For 2Ghz to 3Ghz that will give you 50%. For 3Ghz to 2Ghz that will give you -33%.
 
Online membership?

Mike Linnig said:
I joined ADC a year or so ago at the lowest level (free). That was enough to get the discount.

Did you join at the Online level? According to the ADC website the hardware discount isn't available for Online members:

http://developer.apple.com/membership/

Anyone can access the store but only those with valid memberships (Select, Premier, and Student) can actually make a purchase. Have you actually made a purchase at the ADC store with your Online membership?

thanks

-A.
 
snooziums said:
I need to be able to use my SCSI and standard ATA drives, which use PCI cards to access them.
Why? By today's standards those drives are likely both tiny and slow. What is the advantage of using them over the superior SATA drive(s) that comes with the Power Mac?


Having standard PCI cards is very important, as I have yet to see any PCI-X cards on the market.
Just because you haven't encountered them, doesn't mean they don't exist. You likely just aren't in the market for fibre-channel network cards or the like.


In addition to having PCI cards for standard ATA and SCSI (both internal and for external drives), I also have a PCI card that allows for five USB ports including one internal port (inside of the computer), as well as both firewire 400 and 800.
Well, the PM has 3 USB 2.0 ports and firewire 400 and 800 built in, so I'm not sure what the problem is here unless you desperately need those two extra USB ports and one interal (?) in which case you could buy a $15 USB hub. Also, your firewire ports will no longer be bottlenecked by the 33MHz PCI bus they currently operate on (and which basically counteracts any speed gain from FW800 in the first place). On the PM they are integrated into the 800MHz hypertransport I/O subsystem.


Since the new PowerMacs do not seem to have any internal drive bays, the Zip, Jaz, PCMCIA card readers, and Tape drives mush be placed externally, however that requires having an external SCSI port, which no new Macintosh comes with, thus I have to keep the SCSI PCI card.
That would be a tragedy if anyone used Zip or Jaz drive anymore. As for backup, you do realize it comes with a DVD burner, don't you?


The 1.6 GHz PowerMac G5 appears to be the only new Macintosh that will run standard PCI cards. The 1.8 and 2.0 have PCI-X expansion slots, which are incompatible with any normal PCI cards, and there does not appear to be any PCI-X cards on the market right now.
Wrong and wrong. They are compatible and there are PCI-X cards for things like gigabit ethernet and fibre channel.


If Apple cuts the PowerMac G5 1.6 tower, they will be cutting out the use of PCI cards when there are no PCI-X cards yet on the market. So, basically, the 1.6 G5 tower needs to stay, or Apple needs to find a way to use older PCI cards, or they need to come out with some PCI-X ATA and SCSI, controllers.
They are compatible, so it's not an issue.
 
The Cheat said:
That would be a tragedy if anyone used Zip or Jaz drive anymore. As for backup, you do realize it comes with a DVD burner, don't you?
Also I haven't seen a SCSI jazz drive in a lonng time and since PCI X is backwards compatible it really doesn't matter. If for some reason you feel the need to add SCSI or ATA to the new power mac you can. exactly as you stated later in your post
 
ktlx said:
It would help if you guys actually read and understood Mercury's posts.

You are telling him his math is horrible by quoting him explaining what is wrong with iswimbikeandrun's math. It is iswimbikeandrun whose math is pretty messed up.

Mercury is correct. To get the increase or decrease, you compute ((new speed - old speed) / old speed) * 100. For 2Ghz to 3Ghz that will give you 50%. For 3Ghz to 2Ghz that will give you -33%.

Finally - someone who understands!

For the record:

Mercury is perfectly correct.
iswimbikeandrun is wrong. (and very frustrating to read when he can't even use his wrong formula consistently)
biaachmonkie and blakespot need to pay closer attention to what Mercury has said.
 
Too Hot, Literally?

JasonElise1983 said:
here is my take on it. They are clearing out supplies for WWDC. There will be kind of update, but noone is going to be able to know what it is. Everyone knew the G5 was coming last year, but noone knew what it was until WWDC. Displays are definately coming, and i believe G5 iMacs are as well. Here is what i would bet on being there.

PowerMac G5
Dual 2.0Ghz $1799
Dual 2.6 Ghz $2499
Dual 3.0Ghz $2999

iMac G5
15" 1.6Ghz $1299
17" 1.8Ghz $1799
20" 1.8Ghz $2099

Displays
17"
20"
23"
30"

I might agree with you if the G5 did not have cooling issues, but I think the current fanlessness of the iMac precludes this. Unless they are remodeling the case. This might be reasonable since they have known about the temp of the 130 nm G5 for a while now. Here's hoping...
 
970fx designed for these purposes

jwhitnah said:
I might agree with you if the G5 did not have cooling issues, but I think the current fanlessness of the iMac precludes this. Unless they are remodeling the case. This might be reasonable since they have known about the temp of the 130 nm G5 for a while now. Here's hoping...

What cooling issues? Just how hot do you think the G5 is?? :confused:

Page 9 on pdf

The 970fx dissipates 12.3 watts @ 1.4ghz 24.5 @ 2Ghz these are typical numbers

This motorola sheet is old but it shows that a 1.25Ghz G4 @ 130nm is probably equivalent to a G5 at 1.4+Ghz. Likely to dissipate around 17 watts.

Thus I do not see why people hold on to this myth that the G5 cannot work in an iMac. The 970fx was developed primarily for this application where heat dissipation must be kept minimal. Sure the iMac will have to undergo some changes but they won't be that difficult.
 
jwhitnah said:
I might agree with you if the G5 did not have cooling issues, but I think the current fanlessness of the iMac precludes this. Unless they are remodeling the case. This might be reasonable since they have known about the temp of the 130 nm G5 for a while now. Here's hoping...

You're assuming that they'd just grab the old G5s at those clock speeds and toss them in. I would think that Apple would want to use the 90 nm G5 (aka the PPC 970FX) and just use some of the chips from the lower end of the acceptable yields. Assuming IBM has been able to make at least 2.6 GHz PowerMacs a reality, I can't see why 1.6 and 1.8 GHz (or maybe even 2 GHz) isn't feasible for the iMac G5... assuming it does in fact debut at WWDC.
 
The Cheat said:
Finally - someone who understands!

Mercury is perfectly correct.
iswimbikeandrun is wrong. (and very frustrating to read when he can't even use his wrong formula consistently)
biaachmonkie and blakespot need to pay closer attention to what Mercury has said.

Actually, no.

I paid perfect attention to the message I responded to, which was Mercury saying:

I do hope you were joking, and that the decrease was for someone else's benefit. Can't really tell sarcasm over the internet.

In response to "biaachmonkey" who said:

AHHH!!! Your math is so horrible, please do not try teach it.

Going from 2Ghz to 3Ghz is a 50% increase, while going from 3Ghz to 2Ghz is a 33.3% decrease.
.

You'll note my response addresses this sequence, which I quoted in my response. I don't believe my response could be called incorrect or straying from what I quoted.



blakespot
 
The Cheat said:
...Also, your firewire ports will no longer be bottlenecked by the 33MHz PCI bus they currently operate on (and which basically counteracts any speed gain from FW800 in the first place). On the PM they are integrated into the 800MHz hypertransport I/O subsystem...

Unfortunately, in actual practice this does not seem to help. The Firewire ports on the new G5s actually perform WORSE than the previous generation G4 Power Macs and the current generation G4 PowerBooks. Check out this link from Barefeats:

http://www.barefeats.com/fire43.html

That said, native FW800 would be faster than FW400. But (apparently) the Firewire implementation on the G5s seems to be somewhat lacking (individual mileage may vary).

I also read one post (unverified, seems a little strange) that the Power Mac G5s have only a single Firewire channel, which would mean that all Firewire traffic goes through a single point. Apparently, the earlier G4s used dual Firewire channels.
 
The Cheat said:
Why? By today's standards those drives are likely both tiny and slow. What is the advantage of using them over the superior SATA drive(s) that comes with the Power Mac?

Not arguing with you, because I agree I would much rather use SATA than anything else...but SCSI is not the slower of the 2. :(

SATA vs SCSI vs FW RAID

Ok, 2 more things...for all you people believing in a release of like 2.0, 2.2, and 2.6, or 2.4, 2.6, 3.0. That will never be. The speeds should have an equal speed diference between each processor allowing for the multiplier. Apple/IBM will NOT decrease their bus speeds as some have suggested (900mhz 3x multiplier?????) so hypothetically, speeds should be 2.2, 2.6, 3.0. with respective bus speeds of 1.1, 1.3, and 1.5.

There will not be any single processor G5 towers released. That is why they are specifically selling those at cost along with the fact they are the lowest sellers. Even if apple decides to keep the current 1.8/2.0ghz towers as a lower level G5 and split it into a consumer/pro line, I would welcome that Idea. The single processors will be released in the G5 iMac. Powerbooks got the 1.5ghz G4 upgrade, because they are a ways away from receiving a G5. iMacs however, went unchanged, leading to the conclusion of very immediate releases that are big enough to be done at WWDC, leading to the speculation of a G5/updated form factor. Can anyone say HEADLESS. :) The current iMac uses a 7455, which can average 35w with a max of 50w heat dissapation . The proposed 975/980 (we will just call it the POWER5 derivitave for clarification purposes...) G5's are 65w average, however, that is at their higher speeds. A scaled down version of any SINGLE G5 will easilly run in an iMac without any heating issues.

and lastly, I do not believe you will see any 90nm chip in the next G5. Considering the xServe's 4-6 week waiting period due to low processor yields. Business clients are not going to wait any longer for a server, since when you order a business class server you are generally dealing in mission critical data and your time period is not as leniant as consumers. Introducing a 970fx/any 90nm proc'd G5 tower would significantly increase those delays on both ends, and Apple would loose alot of corporate business, which they desperately need.

So, considering that, along with the fact any 970 based processor will not surpass 2.6-2.8ghz on a commercial level. (it has reached 3.2-3.4 in developemental tests, but the chances for that on a large level are close to 0%) SJ did not make a promise he did not intend to keep. So even if they release 970 based revised G5's, he still needs to release another processor after summers end just to keep his word. To me, that seems like alot of work and $ to spend on an outdated processor, when the POWER5 has no forseen limits currently, and is IBM's baby that they have spent 1.5 years developiing, alongside with Apple.

You will see 975/980 equipped powermacs, you will see G5 equipped (hopeless) headless iMacs, (which will also negate the need for eMacs, and fulfill the desires of all you cube lovers out there wanting a small form factor G5). And I think you will see something totally unexpected from Apple as well. :) Now I could be totally wrong, but Id be willing to bet quite alot that Ill be pretty damn close come WWDC day...

:) Keep the optimism, Apples not in business to dissapoint.
 

Attachments

  • headless-iMac.jpg
    headless-iMac.jpg
    87.1 KB · Views: 429
SyndicateX said:
... I do not believe you will see any 90nm chip in the next G5.

I was kind of liking some of what you said until I came upon the above statement. IMO, there is absolutely no question that the next G5 Power Mac will use the 90nm 970fx processor. I don't even think this is worth debating.

As to the FSB speeds and bus multipliers. That is an issue which seems to be seldom discussed. I'm not sure that Apple can deliver a FSB that runs at 1.5GHz. Would anyone like to take a guess as to how fast they can actually run the FSB? I know that some AMD systems are shipping with FSB speeds over 1GHz, but not by much (1.1 or 1.2 GHz?). My guess is that there is a limitation here and 1.5GHz seems a little high.

Also, does anyone know if IBM has ever admitted to supporting a FSB multiplier other than 2X? I think Arstech even asked this question of the IBM engineers and they wouldn't give a direct answer (but that was for the original 970 chip, not the fx).
 
Just a thought

It's an accnowleged fact that 970FX supplies caused G5 XServe delays. IBM has said already some time ago that the production numbers are now much closer to what they would like them to be. It is also known that G5 XServes are shipping in volume. But there are still waiting times, 5 - 7 weeks shipping time.

Could it be that there are now enough 90nm 970FX sitting at Apple's assembly lines but there is such a big backlog of orders that they can't put them fast enough together?
 
The Cheat said:
Finally - someone who understands!

For the record:

Mercury is perfectly correct.
iswimbikeandrun is wrong. (and very frustrating to read when he can't even use his wrong formula consistently)
biaachmonkie and blakespot need to pay closer attention to what Mercury has said.
My formula was and still is (new-old)/old, I am sorry I didnt use it consitantly, I was pretty tired when I posted, but yeah, biaachmonkie is right. I have been pretty fried latly due to my quantum mechanics final.
 
PowerMac just went to 7-10 day delivery

Power Mac G5s: 7-10 day wait at The Apple Store
Saturday, May 29 2004 @ 9:11 AM CT | Article Link
The wait for new Power Mac G5s at The Apple Store moved from 'Same Day' delivery to 7-10 days this morning.
:D
 
gensor said:
Power Mac G5s: 7-10 day wait at The Apple Store
Saturday, May 29 2004 @ 9:11 AM CT | Article Link
The wait for new Power Mac G5s at The Apple Store moved from 'Same Day' delivery to 7-10 days this morning.
:D


Ok, I doubt Apple could afford to have a month of constraint on G5 2.0Ghz.
I also doubt they would push out the 3Ghz without the fanfair of WWDC in 1 month.
This makes me think we are going to see a speed bump now before WWDC and hopefully the real deal 3GHz later this year...
 
aswitcher said:
Ok, I doubt Apple could afford to have a month of constraint on G5 2.0Ghz.
I also doubt they would push out the 3Ghz without the fanfair of WWDC in 1 month.
This makes me think we are going to see a speed bump now before WWDC and hopefully the real deal 3GHz later this year...

A solid theory - if we do indeed see updates before WWDC, then it's going to be marginal updates to 2.2/2.4/2.6 (or something similar), i.e. nothing worth showcasing at WWDC. And, with WWDC then a few weeks away, there is no way Jobs would go ahead and release the 3 GHz model then, so we would have to expect it later in the year.

So if there are no PowerMac updates at WWDC, due to updates being released earlier, then what does that leave for WWDC? G5 iMacs? New displays? 4G iPods? Or just a 2-hour long presentation on Tiger? ;)
 
blakespot said:
I guess they need to re-examine the state of math classes in schools today...

It was most certainly not sarcasm on his part. 50% of 2GHz is 1GHz. So, going from 2GHz to 3GHz is a speed increase of 50%. 50% of 3GHz is 1.5GHz - so a 50% decrease from 3GHz would result in a 1.5GHz machine. To go from 3GHz to 2GHz you are seeing a reduction in speed of 1/3, or 33.3%.

2GHz is a 33.3% reduction in speed from 3GHz.

blakespot

You obviously misunderstood me. That is the exact math I did. Read my post; this is exactly what I said. However, I wasn't sure whether or not he was being sarcastic in analysis of mine. I said I hoped he was reiterating what I said (and the decrease of 33%) for someone else's benefit, since he would be wrong to address it to me, because we both did the right math. If he is not being sarcastic, then he is wrong, as we are using the same math. You would also be wrong in agreeing with him.

He said my math is horrible. It is the exact same formula and the same results he used. If he is not being sarcastic, then he and you are wrong; since we did the same thing, the correct thing.

So that everyone can understand exactly what was being said:

iswimbikeandrun does incorrect math.
Mercury(me) does correct math, shows him where it is wrong.
Biiachmonkie(assuming he was not being sarcastic about my math being horrible) incorrectly thought I was doing things the wrong way, and addressed his post to me, and did proper math, asserting that there was a difference between his and mine. There is not.
I(Mercury) called him on it, and said I hoped he was being sarcastic(about me not teaching the math, not his actual numbers, which are fine) since we did the exact same thing, the correct thing.
Blakespot then incorrectly assumed I was wrong because I told biaach he was wrong about calling me wrong. His math was correct, his analysis of mine was not.

To further clarify: This mess began when biiachmonkie(assuming he was not being sarcastic about my math being horrible) misinterpreted what I had to say as being incorrect. We both did the proper math; however, he thought mine was wrong, which it was not. That's where this confusion started. As for the state of math classes, it doesn't really seem fair to say that the quality of education has decreased markedly, as you seem to imply, blakespot, when there is at most a decade between our last math classes. Unless, on the other hand, I can decry the '80s and early '90s trouble with reading comprehension... :p

Thanks to ktlx and The Cheat for taking the time to read and try to explain.
If this needs further clarification, I'll do a post-by-post analysis...

Edit: Edited for maximum explainability.
 
~Shard~ said:
So if there are no PowerMac updates at WWDC, due to updates being released earlier, then what does that leave for WWDC? G5 iMacs? New displays? 4G iPods? Or just a 2-hour long presentation on Tiger? ;)

G5 iMacs...well overdue. Would be a very good thing to appear to celebrate the 20th anniversary year of the mac. If they can only get the price down as well...

New displays...yep...due a while back. I suggest both the 20 and 23 will not support HD Video with room for editing tools...

4G iPods...definetly the right time to get them out as ipod minis are few and far in between and the opposition is pumping lots of players into the market...wonder how much more of a pda it will be...

Tiger...30 mins...special whole screen icon/bar/text resizing to cope with new screens and new PB G5 screens next year...faster...voice recognition/speaking...selective file/directory encryption...

also

iTune Store and Pod success...15 mins... :rolleyes:

.Mac numbers and new stuff?? 10 mins if it happens... :)

Mac Office 2004...sales, performance etc... :rolleyes:

Maybe mention or hint of the next iLife addition... :D ...iHome, your personal organiser for all your home life issues like bills, reminders, recipes, shoppinglists, service schedules, warranties, purchases, personal finances (able to pull down your bank and cc statements each month and track your spending) etc linking in with iCal to help you organise and keep track of your home life...maybe a cool database of all your music and computer discs, books, videos, etc like some third party products have - with barcode scanning through isight ;)

Better home TV recording support with iMovie and iDVD...a Mac digitial SD/HD TV tuner! :eek: ;)

New Pro Apple Mouse :)...with multibuttons and a wacky wheel... :p
 
aswitcher said:
Ok, I doubt Apple could afford to have a month of constraint on G5 2.0Ghz.
I also doubt they would push out the 3Ghz without the fanfair of WWDC in 1 month.
This makes me think we are going to see a speed bump now before WWDC and hopefully the real deal 3GHz later this year...

I think we have learned that specials, rebates and extended shipping times often have no correlation to the release of new products.
 
The 1.6GHz Power Mac G5s are also showing a 7 to 10 day wait. Given the ADC promo on these machines and all of the speculation on why Apple is trying to reduce inventory on the 1.6GHz model the waiting periods at the Apple Store make no sense. Maybe these delays have nothing to do with the availability of the physical hardware. Or, it could just be a temporary mistake on the part of the online store.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.