Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Re: We need the extra power

Originally posted by mk_in_mke
Guys,

This is some good news and I am all excited about it... I have a ACD 22" + 2x 1Ghz PMac + 1.5 G RAM + 2 x 80 Go + FCP 3 + DVD SP 1.5 + PShop... All the bells and whistles... This works just fine and I am very happy... But gee, just to get the extra power, i would buy the dual Pmac + PPC 970!!!

This is getting outrageous: the more we have, the more we need... :D

Now, the wait is going to be a nightmare... This explains why Apple has delayed their WWDC...

Time to save some money (I will cut on the Sapphire and drink Gilbey's, I will cut on the caviar and have some salmon eggs, I will cut on my mistresses, nor more jewels, diamonds...)... Here I come Apple!!!

:D

Take care guys!!!

Michel

If the 970 doesn't come out, can I be your mistress? ;)
 
New processors, new look??

Remember all the talk of the mainboards being longer and narrower than current G4 boards?

Also remember the patent on the mighty color morphing power case?

Why not a whole new look for the PowerMac?? It's been long enough, and frankly, this behemoth of a case under my desk is cramping my legroom.

Now I know we don't want to prohibit expansion, but for what most of us are adding (extra drives, an extra audio or video card), does it really need to be that big?

I think it's high time Apple took the lead and redesigned the professional desktop. Quiet, space-efficient, powerful computers that are works of art.

Send Mr. Ives the challenge.
 
Re: The G3-4 chips still have some life left, but...

Originally posted by tazznb
If Apple only puts the 970 in its highest end PM, no one will buy the other PMs without it, and at a realistic price point.

"The IBM's production can not match the demand for the G5".

This is IBM taking up the chip process, now. Not Moto; If Moto were doing this I can agree the chips not being readily available to use for all users for another 2-5yrs :D

IBM's manufacturing will also cut the prices down, at least when compared to relative clockspeeds on the G4. Apple should save on palettes of chips. Beyond that, the supply line and reliability should improve. I think Moto's inability to deliver has been a major issue with shipping, etc for Apple. IBM should solve that problem - they are a company that tends to deliver ahead of schedule.

The exciting thing to watch for may be a .09 micron PPC970. They may be ready by December - and that may take them up to 2.5 GHZ while increasing power efficiency.
 
Re: New processors, new look??

Originally posted by paulie
Remember all the talk of the mainboards being longer and narrower than current G4 boards?

Also remember the patent on the mighty color morphing power case?...

I think it's high time Apple took the lead and redesigned the professional desktop. Quiet, space-efficient, powerful computers that are works of art.

Send Mr. Ives the challenge.

Quite agree. I made this point in a recent Discussion board thread I started, but the only people who posted disagreed. What the...?:confused:

My point is that Steve's stated aim with the TiBook was to produce "the worlds sexiest computer". Few people would be biased enough to ascribe that title to the current PM.

I own a PM fan and wouldn't buy anything else. I too would like to see the gauntlet thrown down to out friend Jonathan Ives and his team.

When I buy a new 970 PM, I don't just want power for my money, I want sex too!:)
 
Yeah, this is all good news.

Personally I think my G4 533 PM, just gone 2 years old has still got some legs.

As a home user, I'd like to think I could hold out 18 months till the 980 hits, but the temptation to just run out and grab a 970 PM a few months after they hit is going to be HUGE!

But most of all, I just want to see the 970 out in the market place, to guarantee the future of the PM line, so that whenever I do upgrade, there'll still be such a thing as a PowerMac!
 
new PPC 970

Originally posted by MacManiac1224
If they are in production now, when can we expect to see them appear in new machine? is it possible they can appear in June at WWDC?

I seems to me very clear that the new PowerMac will be announced in Jine at WWDC. Why?
- IBM PPC970's father is giving a speach on that processor and architecture
- it will probably be linked with the new serveur from IBM using the PPC970 (check IBM road map)
- MacOSX 64bit might be demoed
- Apple is also present at the WWDC
- and last but not least , there will be no Keynote in MacWorld NY in July.... Why??
because the last Apple products will have been presented and demoed 3 weeks before at the WWDC!!!!!!!!!!!
So be prepare for June, it is going to be hot.
In addition Macbidouille.com is usually a very reliable mac info site.
All their info concerning the PPC970 and the corresponding motherboard make a lot of sens. The mobo will have AGP8x and hypertransport technologies (imagine the effect on a dual PPC970, knowing that the hypertransport mimics 2 virtual processors from 1 physical unit....). They also reprt that PCI express will be present on the PPC970 inly next year since it is not even implemented yet in the PC world.
Why hypertransport?? because IBM also needs it for its server based on PPC970...so Apple had no reason for not implementing it on its mobo.
 
Re: new PPC 970

Originally posted by eric67
- IBM PPC970's father is giving a speach on that processor and architecture
- it will probably be linked with the new serveur from IBM using the PPC970 (check IBM road map)

Of course the PPC970 speach by Peter Sandon will be linked to a new server from IBM - IBM publically stated that they were going to use the chip in a blade server!

- MacOSX 64bit might be demoed

Don't hold your breath.

- Apple is also present at the WWDC

Of course they are - it's their bloody show in the first place!

- and last but not least , there will be no Keynote in MacWorld NY in July.... Why??
because the last Apple products will have been presented and demoed 3 weeks before at the WWDC!!!!!!!!!!!

And what impact would this have on consumers at large? WWDC is a developer's conference! I know, I'm going, and I can tell you right now that whatever Apple announces will be buried in NDAs so that no one that goes will be able to say a damned thing, so don't expect to hear what gets announced from those of us that go.

The mobo will have AGP8x and hypertransport technologies (imagine the effect on a dual PPC970, knowing that the hypertransport mimics 2 virtual processors from 1 physical unit....).

Your confusing HyperTransport with HyperThreading. The former is a bus technology from a consortium of computer developers that, iirc, includes Apple & nVidia. The latter is a technology found in the latest Pentium 4 that allows a single processor to appear as more than one to the OS.

They also reprt that PCI express will be present on the PPC970 inly next year since it is not even implemented yet in the PC world.

I would think that HP (one of the creators of the technology) would have shipped some in their servers, but I could be wrong. None the less if this came to pass it would be the first consumer release of PCI Extreme that I know of.
 
Originally posted by Frobozz
This is going to be the best year since 1984 for the Mac. I have no doubt.

1994 was pretty big. I think the introduction of the PowerPC in the first place was huge.

Now, another issue is the IBM vs. Moto issue. Moto was Apple's processor supplier ever since the Apple I, when Woz decided to get the Moto processor because it was cheaper than the Intel processor. (It wasn't Moto then, Moto later bought out that company.)

From 1984 to 1994, Apple used the 68k processors from Motorola. Consistently, the 68000, 68020, 68030, and 68040 were outperformed by the Intel 8086, 80286, 80386, and 80486, respectively.

In 1994, Apple decided to ally with Motorola and long-time rival IBM in the AIM alliance. This alliance would retain Motorola as a direct chip vendor to temper the alliance with the hated IBM. (IBM was the Microsoft of the time, remember.) The idea was to change IBM's POWER series of high-end workstation processors into a viable PC processor.

The first PowerPCs were the 601, 603, and 604. There may have been another, but Apple didn't use it. The 601 was the main chip, the 603 was the low-power chip, and the 604 was the big, powerful chip. The 601 was produced only by IBM. Motorola bought the 601s from IBM, stamped "Motorola" on them, and sold them again to Apple. No Mac user wanted to buy IBM, so this was necessary.

(Remember, we were more cultish back then.)

Anyway, the 603 and 604 were designed at the joint Motorola-IBM facility, Somerset. Somerset lasted well into the G3 period.

The G3 was based on the 603e (an enhanced 603, more powerful than the 601). In mid-1997, it was put into PowerBook and Power Mac form. By 1998, the entire line was G3. It wasn't too much later that Somerset broke up, as Motorola pursued an AltiVec-enabled G4 and IBM pursued copper processors (first copper processor was the copper G3, introduced in the blue-and-white G3.)

The point?

First, there is precedent supporting the notion that 970 will spread to the entire line quickly. The G3 was enough of a boost over the 604e to do so. Second, the IBM thing isn't new. It dates from 1994.
 
Simple recompile for 64-bit - not really....

Originally posted by Macpoops
Number 3. OS X was written as 64 bit clean. Which means all that would be required to make use of the 64 bit processor would be a simple recompile. NOT a complete rewrite of the OS.

This is not really true - especially when you look at the whole OS.

The Mach kernel may be 64-bit clean, but even so it is likely that some adjustments need to be made (OS X only runs on systems with 2GB or less of memory, does a change need to be made for a 64-bit system with 4GB or more?).

The real work, though, is all the libraries and applications. While many apps can remain 32-bit (the clock, finder, etc), one would need to have 64-bit versions of Cocoa and Carbon. Even if the code were 64-bit clean (possible for Cocoa, highly unlikely for Carbon), all of the APIs are 32-bit. You therefore need to produce separate 32-bit and 64-bit APIs.

Think about the simple "malloc" call which allocates memory. A 32-bit program needs "malloc" to return a 32-bit pointer to the new memory. A 64-bit program needs to get a 64-bit pointer.

When you pass that memory to a another API, you'll need a version that expects the same size pointer that you got.

One also has to worry about code that was once 64-bit clean becoming unclean. Apple has hundreds of people working on the code - any one of them could have made a mistake and assumed that a pointer was 32-bits. Those mistakes have to found and fixed.
 
I thought the PPC970 could run 32-bits code as is, so no rewrite is necessary to run the current codebase?

Ok, to take advantage of the 64-bit addressing space the OS would have to be rewritten, but at first I think you could run 10.2 without modification.
 
Re: Re: new PPC 970

Originally posted by Rincewind42
Your confusing HyperTransport with HyperThreading. The former is a bus technology from a consortium of computer developers that, iirc, includes Apple & nVidia. The latter is a technology found in the latest Pentium 4 that allows a single processor to appear as more than one to the OS.


Oups, big mistake, I am very sorry.
I am so excited to see the future development of the PPC970 in Apple machines that I was reading a bit too fast....
Hypertransport is something good of course, but I think the best , if true, will be the usage of PCI express by Apple in 2004 (ok it is a rumors), but what I can see positively is the integration by Apple in its machine of the latest technologies, and if regarding connectics (FireWire, USB,...) Apple has also been ahead, regarding mobo-related innovation, Apple has always been behind, far behind..... just look how long Apple desktop and notebook had a less than 100MHz mobo..... where is the AGP8x?? (OK I know the gain compare to AGP4x is small, but marketing wise....) graphics cards....I mean iMac are still delivered with Geforce2MX or 4MX, I mean get real those cards are 2 or 3 years old.....

ok my 2 pennies remarks...

by the way I love Mac, I just would like to see Apple machines more present around us, in our normal life, and not retricted to small business or special area...
 
Re: Simple recompile for 64-bit - not really....

Originally posted by AidenShaw
Think about the simple "malloc" call which allocates memory. A 32-bit program needs "malloc" to return a 32-bit pointer to the new memory. A 64-bit program needs to get a 64-bit pointer.
You are funny ;)

man malloc

void *
malloc(size_t size);

size_t is already 64 bits wide

Pointers are based on register size, developers should not rely on pointers being 32 bits (this is a very bad habit), it's not difficult at all to code without knowing the pointer size and to keep portability in mind.
When you declare a pointer you do not precise it's size anyway, thus you should not make an assumption about it.
 
Re: Re: Re: new PPC 970

Originally posted by eric67
....I mean iMac are still delivered with Geforce2MX or 4MX, I mean get real those cards are 2 or 3 years old.....

Umm How is the Geforce 4mx 2 to 3 years old. the 2mx is but the 4mx just came out last year. It was offered in the PM just before the Ti came
 
Re: Re: Re: new PPC 970

Originally posted by eric67
Oups, big mistake, I am very sorry.
I am so excited to see the future development of the PPC970 in Apple machines that I was reading a bit too fast...
Replace PPC970 by PPC980 and you'll be probably right. The POWER5 has this multithread on a single core feature, the PowerPC 980 (Q3 2004?) is expected to be based on the POWER5 core plus AltiVec.
 
New article

I know MacWhisper's reputation has been tarnished, but it's worth a read:
April 10, 2003
Update: 970 PowerBook Production: Accelerated 970 Production

Further discussions with Apple OEM assembly partners have given us up to date information on motherboard production for both PowerBooks and PowerMacs, both using the IBM PPC 970 processor.

First, we have been told, and we have second-sourced the claim that 15.4-inch PowerBook motherboards are now in production and are shipping regularly to Apple's contract PowerBook assembler. Next, we have been told, and we have also second-sourced a claim that bid requests for a fully-designed 970-based board for the 17-inch PowerBook were received by two assembly plants this past Friday, with a submission deadline for replies of April 30th.

Finally, a reliable source in engineering management at the ODM supplier providing the upcoming PowerMac motherboards has informed us that those boards went into volume production this past Friday, and that first shipments to Apple's final assembly partner for the new PowerMacs is to take place "about April 15th."

We will add that our sources seem consistently taken aback by what they all characterize as the unexpected and very unusual hurry involved in all work on these new desktop and portable Apple products. Every step in each process is being scheduled far tighter than is normal for a new production run.
 
printf ("%d\n",sizeof(void *));

Originally posted by mathiasr
size_t is already 64 bits wide


printf ("%d\n",sizeof(void *));


:confused:

Note that I said "returns".

One *shouldn't* count on the size of a pointer, I agree. Disciplined, skilled programmers usually don't.

"Should not" and "does not" are different.
 
Re: New article

Originally posted by mproud
I know MacWhisper's reputation has been tarnished, but it's worth a read:
Yeah, well Jack Campbell and his site, MacWhispers, has put its neck on the line with some pretty bold stuff. It has actually put a shelf life on that site. If that stuff doesn't come to fruition by WWDC...he might as well fold up his tent and go home.
 
Re: Simple recompile for 64-bit - not really....

Originally posted by AidenShaw
Even if the code were 64-bit clean (possible for Cocoa, highly unlikely for Carbon), all of the APIs are 32-bit.

Carbon is a complete reimplementation of the traditional MacOS APIs, so the implementation in MacOS X is very likely to be 64-bit clean (or nearly so)

Originally posted by mathiasr
man malloc

void *
malloc(size_t size);

size_t is already 64 bits wide

size_t should not be 64-bits on a 32-bit platform. It should be the maximum size that a pointer can refer to, which on a 32-bit PowerPC platform is 32-bits worth.
 
I just thought of an interesting thought.
Normally Apple makes their "Northbridges" I understand. But Apple could use nVidia !
nVidia as a Northbridge for the Athlon processors called nForce that uses dual DDR channels. The future Athlons will have a HyperTransport bus. nVidia would follow so there Northbridge chip would support 1) HyperTransport 2) AGP 8X 3) dual DDR SDRAM 4) PCI bus.
This is very,VERY close to what Apple needs for their new 970 systems!
Is there any other rumors about nVidia and Apple?
Maybe ATI?

PS Are the memory controllers on Macs refered to as Northbridges?
 
new info pn PPC970

Macbidouille.com reports today 15th April 2003, that the new PowerMAc are entering production!!! well let say that the company which will assemble the future powermac has been chosen, and it is "Hon Hai Precision" also better known as"FOXCONN". They will receive the different pieces (RAM, HD and chips,...) from IBM, pointing out ,as already expected that the new PoerMac will have its processor on a ZIF-type socket!! So Apple should not have problem to demo the new PowerMac at the WWDC in June. It is alctually interesting to notice that this info come on timeas it was reported few weeks ago that Apple was looking for companies for assembling the future PowerMac.... at that time if I remember well the dead line was first week of April ....
let's wait and see
 
new info pn PPC970

Macbidouille.com reports today 15th April 2003, that the new PowerMAc are entering production!!! well let say that the company which will assemble the future powermac has been chosen, and it is "Hon Hai Precision" also better known as"FOXCONN". They will receive the different pieces (RAM, HD and chips,...) from IBM, pointing out ,as already expected that the new PoerMac will have its processor on a ZIF-type socket!! So Apple should not have problem to demo the new PowerMac at the WWDC in June. It is alctually interesting to notice that this info come on timeas it was reported few weeks ago that Apple was looking for companies for assembling the future PowerMac.... at that time if I remember well the dead line was first week of April ....
let's wait and see
 
Originally posted by Flynnstone
I just thought of an interesting thought.
Normally Apple makes their "Northbridges" I understand. But Apple could use nVidia !
nVidia as a Northbridge for the Athlon processors called nForce that uses dual DDR channels. The future Athlons will have a HyperTransport bus. nVidia would follow so there Northbridge chip would support 1) HyperTransport 2) AGP 8X 3) dual DDR SDRAM 4) PCI bus.
This is very,VERY close to what Apple needs for their new 970 systems!
Is there any other rumors about nVidia and Apple?
Maybe ATI?

PS Are the memory controllers on Macs refered to as Northbridges?

The PowerPC is Big-Endian, the Athlon (and any x86 processor) is little-endian. The N-Force2 chipset is not compatible with the the PowerPC as it stands... they don't even read data in the same direction!
Also, As far as I remember, Nvidia doesn't use a classic northbridge/southbridge architecture anymore (Intel doesn't either).
And... Just because AMD ties the media controller to other chips via Hypertransport, that doesn't mean that their chipset has a processor bus that is in any way alike with what is required by the 970. The Athlon uses a DDR processor-to-chipset bus. It currently maxes out at 166MHz x 2. The 970, by all accounts, will run something like a 225MHz x 4 (quad-pumped) processor bus. The 970 bus is also comprised of 2 one-way 32bit buses. The buses used by the two PROCESSORS is totally different.
 
Re: new info pn PPC970

Originally posted by eric67
Macbidouille.com reports today 15th April 2003, that the new PowerMAc are entering production!!! well let say that the company which will assemble the future powermac has been chosen, and it is "Hon Hai Precision" also better known as"FOXCONN". They will receive the different pieces (RAM, HD and chips,...) from IBM, pointing out ,as already expected that the new PoerMac will have its processor on a ZIF-type socket!! So Apple should not have problem to demo the new PowerMac at the WWDC in June. It is alctually interesting to notice that this info come on timeas it was reported few weeks ago that Apple was looking for companies for assembling the future PowerMac.... at that time if I remember well the dead line was first week of April ....
let's wait and see
Wasn't the last mac to use a zif socket the B&W G3 or G4 yikes? Isn't the current daughter card socket called a different name? And why would apple go back to the ZIF if the daughter card format worked so well with the powermacs?
 
Originally posted by ffakr
The PowerPC is Big-Endian, the Athlon (and any x86 processor) is little-endian. The N-Force2 chipset is not compatible with the the PowerPC as it stands... they don't even read data in the same direction!
Also, As far as I remember, Nvidia doesn't use a classic northbridge/southbridge architecture anymore (Intel doesn't either).
And... Just because AMD ties the media controller to other chips via Hypertransport, that doesn't mean that their chipset has a processor bus that is in any way alike with what is required by the 970. The Athlon uses a DDR processor-to-chipset bus. It currently maxes out at 166MHz x 2. The 970, by all accounts, will run something like a 225MHz x 4 (quad-pumped) processor bus. The 970 bus is also comprised of 2 one-way 32bit buses. The buses used by the two PROCESSORS is totally different.

Hypertransport is on the Athlon64 and Opteron and likely on the 970. Its like PCI bus in that it doesn't matter about processor. Big-Endian , little-endian just has to do with how you interpret the data. Some PPC processors have the ability to byte swap to partially deal with the endian issue.
If nVidia is working on Hypertransport, then they have the capability to do a chipset for Apple.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.