THT, do you think Motorola is trying to wean it's developers off MPX perminantly by not advancing it's design? Ultimatly they want all developers using RIO with in a specific time frame they have, maybe? That's all fine and good, but where does that leave Apple? The implication is that they want to stop further G4 development and a die shrink is not even planned because they are migrating to the 85XX line. This means soon (and before they discontinue the G4 74XX line) Motorola intends to bring to market an 85XX with powerful floating point units, Altivec. Not if but when. Question is will it have an assortment of embedded-specific features (built-in ethernet, serial ports, etc.) or will it be more taylored to a profession desktop use directed towards customers like Apple?
With all this talk about Motorola now, I'm sorry to further bifurcate this thread!
Question is will it have an assortment of embedded-specific features (built-in ethernet, serial ports, etc.) or will it be more taylored to a profession desktop use directed towards customers like Apple?
I don't think I can take much more of this. One minute we think MOTO stinks, next they are going to be first with 0.09 and out pace Intel. Two weeks ago we were going to get a Power4 derivative, now Apple has turned it down! Earlier today I read another post that they had turned it down because they are working with AMD
If there is just one insider who would just say something re-assuring, please do.
"I don't think I can take much more of this. One minute we think MOTO stinks, next they are going to be first with 0.09 and out pace Intel. Two weeks ago we were going to get a Power4 derivative, now Apple has turned it down! Earlier today I read another post that they had turned it down because they are working with AMD"
It's making you crazy? Everybody else here snapped long ago. We're all nuts.
This is the first time I've been able to see a link that directly lets me put my point across about 'power'Macs.
I've let them know, politely but firmly want I want to pay for. I've also dropped a few hints about why I and maybe other Mac 'faithfuls' aint buying 'power'Macs in droves anymore...
THT, do you think Motorola is trying to wean it's developers off MPX perminantly by not advancing it's design? Ultimatly they want all developers using RIO with in a specific time frame they have, maybe?</strong>
Yes partially. Moto is moving to on-die DDR memory controllers for its embedded processor strategy (as seen in the 8460 and 8540) and recognizes that improving the MPX bus is a waste of time for their main market (and it would be competitor to RapidIO to boot). So I think Moto will consider MPX as legacy tech whence 8460 ships.
I think Moto just doesn't want to invest in system logic controllers anymore and have done what they've needed to stop developing them. Moved the memory controller to the chip and contracted 3rd parties to build RapidIO chips and interfaces. This doesn't leave any room for MPX processor bus development anymore.
<strong>That's all fine and good, but where does that leave Apple? The implication is that they want to stop further G4 development and a die shrink is not even planned because they are migrating to the 85XX line.</strong>
Maybe. I don't why Moto would even consider not producing a 7455 based G4 on 0.13u though. It performs better than either of the 8460 or 8540, and the 7455 would be in terrific die size and power consumption territory at 0.13u.
Obviously, it leaves Apple with their drawers down for awhile. Hey, I said Apple should have taken over processor development 4 years ago. They should have listened to me.
<strong>This means soon (and before they discontinue the G4 74XX line) Motorola intends to bring to market an 85XX with powerful floating point units, Altivec. Not if but when.</strong>
Well, I believe it's still 75xx on the Moto roadmap for host processors. 85xx are monikers for communications processors (8540 et al). And powerful and 85xx don't come to mind as well. These are non-starters for Apple.
<strong>Question is will it have an assortment of embedded-specific features (built-in ethernet, serial ports, etc.) or will it be more taylored to a profession desktop use directed towards customers like Apple?</strong>
I don't know. Apple needs 0.13 micron G4 chips right now. And the news has been scarce for Moto HiP 7 chips.
The only tidbit of real actual information available is this IBM desktop PPC. If IBM can ship a single core Power4 with AltiVec compatible SIMD units, I don't think a 0.13u G4 without modifications has much of chance. A 0.13u single Power4 would be in the 1.3 GHz to 2 GHz range, with more powerful scalar execution resources. SIMD is an unknown right now. Since Peter Sandon of Gekko PPC fame is doing this IBM desktop PPC, I'm not getting good vibes that this is separate SIMD unit; rather, the 2 Power FP units will be used to compute 4 single-precision ops. This could be sweet actually if the single precision ops can be independent.
"If they do that, it wound mean HiP 7 has been a catastrophic failure. At least HiP 5 was reasonably useful, but so far HiP 7 has been a bust for Moto processors. "
But...why? What went wrong? Yeesh. They're a multi-billion dollar company. What the hell are they playing at?
THT, some incisive posts there about some of the peformance issues the G4 faces. Fascinating to read and virtually impossible to rebuke. I like the way you nailed the DDR issue and the many aspects of cpu performance that could improve the G4. Cold hard light of day stuff. I like it!
"Why hints? Just tell them THE TRUTH you bast@rd"
Aha! Leonis! Yeah, I kinda told them the 'truth'...and left them in no doubt what it's gonna take to get the cobwebs off my next POWERMAC fund!!!
Speaking generally, (being one for digression an' all...), I have to confess I almost buckled in my stand off with Apple's 'power'Mac. Indesign 2 free with any 'power'Mac. That's crazy? It costs the best part of £600 here in the UK!!! Buy a low end 'power'Mac and you get a superb piece of software. That's a killer promo' from Apple and Adobe. Watch out, Quark...
Maybe Apple's just beginning the 'fight back'?
(I know as soon as I cave in the G5 will hit soon after...still...Indesign...anybody else tempted? I suppose you could always dump the antiquated tower on ebay...if/when the next revision comes along...)
"Obviously, it leaves Apple with their drawers down for awhile. Hey, I said Apple should have taken over processor development 4 years ago. They should have listened to me."
, I remember you saying it too!
"Well, I believe it's still 75xx on the Moto roadmap for host processors. 85xx are monikers for communications processors (8540 et al). And powerful and 85xx don't come to mind as well. These are non-starters for Apple."
Do you believe the G5 or last iteration of the G4 will be the '7500'? Wasn't this yanked from Motorola's site a while back? I always thought 7500 was the 'G5'. Maybe it's just not enough. And Apple will hopskotch on. But I can't see Apple not using it if Moto's done the work on it. Let's face it. A 7500 would be darn better than what they've got now!
"I don't know. Apple needs 0.13 micron G4 chips right now. And the news has been scarce for Moto HiP 7 chips."
Yes. But what happened? Surely you can't develop something like HiP 7 and do nothing with it?
"The only tidbit of real actual information available is this IBM desktop PPC. If IBM can ship a single core Power4 with AltiVec compatible SIMD units, I don't think a 0.13u G4 without modifications has much of chance."
The Register had an article a while back on this chip. Said it was slated for 2003 Jan'. .13 G4, 7500 with Rio. But other than that, just a G4, no fp improvement or anything extra. Clock speeds of...was it 1.6?
"A 0.13u single Power4 would be in the 1.3 GHz to 2 GHz range, with more powerful scalar execution resources. SIMD is an unknown right now. Since Peter Sandon of Gekko PPC fame is doing this IBM desktop PPC, I'm not getting good vibes that this is separate SIMD unit; rather, the 2 Power FP units will be used to compute 4 single-precision ops. This could be sweet actually if the single precision ops can be independent."
But I'd rather take the latter option you discuss. But the '7500' wouldn't be a bad option to take in Jan' as we're pretty much feel the Power4 lite won't hit until next Summer. But I'd like to be proven wrong of course...
"The combination of unique form, a stubbornly loyal user base, and the simple fact that PowerPC may lag x86 performance but is still sufficient for the vast majority of applications, should allow Apple to keep most of its tiny market share intact over the next few turbulent years. However, Apple should drop silly and misleading ads campaigns like the G4-as-supercomputer or Pentium-as-snail, it just doesn?t stand up to scrutiny. Indeed, the less attention Apple draws to performance comparisons the better.
Conclusion
After five years of keeping PowerPC performance more or less in lockstep with x86, IBM and Motorola?s recent play-it-safe strategy of not straying too far from the design parameters of the high-end embedded control market has backfired. The intense struggle between Intel and AMD for technical supremacy in the x86 market has pushed their processors up into the territory of workstation class RISC processors in terms of sophistication, design aggressiveness, and performance. The lack of urgency on the part of IBM and Motorola has caused the PowerPC to fall significantly behind the top of the line MPUs from Intel and AMD.
The nearly effective lack of applications that require GHz level processors has meant that the damage to the competitiveness of PowerPC and Apple?s Power Mac line may be more of a perception than reality. It is up to Apple Computer, a company that is often more successful managing perception than reality, to overcome the burden of a large and widening performance gap by continuing to concentrate on difficult to quantify but desirable aspects of its products, such as system level packaging aesthetics, low noise and heat output, and freedom from Windows. The key to continued viability is retention of the bulk of the Mac user base while attracting the newcomers to computing who are more repelled by beige boxes than attracted by stratospheric clock rates."
I couldn't resist cutting and pasting this bit.
That about sums up Apple's position about right now. How a G5 from IBM would rock this article to its very core.
First, although, the G4 has fallen behind the high end Intel and AMD chips it is still not slow. The G4 is sufficient for almost any task a computer buyer does.
Second 10.2 is not slow in the least. I have often tried doing the same tasks in 9 as in 10.2 and 10.2 was faster if not the same at 80% of the tasks. I think the reason 10.2 is slower in some applications is not so much the OS its self but the fact that the applications have not been written very well. This is most likely because the programmers are still having to learn how to write good code for this new OS.
Third, Yes motorola has been working on the next PPC chip for apple but so has IBM. Motorola is not as interested in the idea as IBM and apple is not as interested in Motorola. The next chip will most likely come from IBM because of their obvious interest in working with apple. Along with obvious other proof circulating around the Internet. All the information needed to come to this conclusion is out there and there is plenty of it. I suspect it will still be about 8 months though before we will see the G5. If any one has any inside info on a release date before this just post "NO". I don't have a good time for release other than 8 months but sometimes I feel like it will be sooner.
And finally the reason it is so hard too pin down where the G5 will come from is because apple has been working with many companies on the matter and we are all looking for one answer. Apple has everything set up just right. All they need now is a good economy and a good co partner to design their chips. They are being very careful with this decision. Motorola, AMD, and IBM have all been working with apple on this matter. But it looks like IBM came out on top. (AMD would be the least likely because of obvious complications in going x86)
If your talking about an IBM cpu, who knows. But if your talking about a G5 from Motorola, that is based on the MPC 8540/8560 architecture(re: Rapid I/O, Ocean switched web, etc.), I'm inclined to believe the end of 2003 date.
</strong><hr></blockquote>
I'm changing my mind on this. Been thinking about it and seems the rumor generated by the Nasa employee maybe more truth that fiction. The IBM chip is a 64 bit processor, may not implement Altivec and hasn't even been presented. We'll see.
However, if I remember correctly the old Motorola roadmap for the G5 indicated it was to be on a 0.1µ process. With the new alliance between Motorola/Phillips/STMicroelectronics claming they would be using a 0.09µ process to " start production of a high-performance chip by the fourth quarter of this year and a lower performing version in the third quarter of 2003."(<a href="http://www.reuters.com/news_article.jhtml?type=technologynews&StoryID=139 1841" target="_blank">Link</a>) Maybe, just maybe, the delay in sampling the MPC85XX was due to power consumption and heat genterated when manufactured on a 0.13µ process. The press release did not say which facility will produce the processors, it could very well be they will not be produced by Motorola, but by either Phillips and/or STMicroelectronics(re: do they have better fabs??). Motorola has announced they would be shifting production.
The basic architecture of the MPC74XX hasn't really changed in quite some time. Rumors indicated Motorola had problems implementing the 0.13µ process. The extreme size of the heat sinks in the new Powermacs is not the direction Motorola/Apple would normally go.
If most of Motorola's/Apple's efforts have been directed @ a G5, then the G4 would not have any major architectural changes, due to lack of engineering resources. Basically, the G4 needs more pipelines(re: G5 reportedly to have 10 stages), capability of handling DDR frontside bus, better floating point and a smaller process. Too much engineering, when efforts were tied up in G5 development.
Question is,"what did Mr. Chris Belden's comment regarding "a high-performance chip by the fourth quarter of this year " mean?". High end processor for communication, networking, desktops????
Wow. More than 14 pages of opinions. My bet goes towards a PowerPC 7470 done by Motorola. And no G5 either way until Summer 2003. I might know less than the one who will rip me apart from this, but then again it doesn't seem like ANYone on here has 'insider' info. We've all heard that IBM will show a 64bit PowerPC processor for the desktop in October and we all have NO idea whether that chip will be available in 2002 or late 2003.
We also thought Motorolas G5 would be the bliss of 2002. End of rant.
We can count on a 7470 G4, probably fabbed on a 130 nm process, but not for a long time...probably by MWNY 03 at the earliest. Motorola has been trying to figure out the 130 nm process for a while but not long enough, remember the G4 has been fabbed on a 180 nm process since forever.
In parallel, IBM is working on a 64 bit desktop PPC CPU with an Altivec compatible SIMD unit. This chip will be fabbed in IBM's new chip foundry in New York. Full production early 2003, unclear whether it will be fabbed using a 130 nm process or a 90 nm process (only know that production at the foundry will start with 130 nm, but quickly "move to 90-nanometer production".
No, this does not mean that Apple doesn't know what CPU will be used in their next generation Powermacs. People sometimes seem to forget that the G4 has a very long life ahead of it in CONSUMER Macs, i.e., iMacs, iBooks, and of course Powerbooks until IBM can supply Apple with a G5 that doesn't melt the polar ice caps.
Let's say for a moment that IBM cranks out 2 GHz Power 5 CPUs faster than Apple can sell them, beginning in Jan 2003. That would mean a Powermac lineup that finally rocks, and Apple would be selling them like nobody's business. However, the iMac in particular would still need to have a competitive CPU....Apple couldn't very well sell 933 MHz G4 iMacs alongside 2 GHz Powermacs. (well yeah they could and maybe they would, but it would look REALLY bad). So it follows that Apple is probably nagging Motorola to supply them with a G4 that will last another 1-2 years, with the implicit threat that if Moto cannot give Apple a competitive G4 for the iMac, then Apple will simply switch ALL Macs to the Power 5, heat and power consumption be damned. You can bet that Apple has new iMac designs that can handle a Power 5, just in case Motorola fscks up again.
Wht this means is that Apple is finally playing Moto and IBM off against each other, primarily to put the fear of God in Moto. The truth is that Motorola cannot afford to lose Apple's G4 business....they are making a phat profit on Apple now that the G4 is Apple's standard CPU. When Apple migrates the Powermacs to the Power 5, Moto is still counting on G4 sales....but finally Apple is no longer Moto's little bitch. Very soon, Apple will be telling Motorola how things are going to be, and you can bet Steve Jobs is already practicing what he's going to say to Moto executives when he finally tells them to fsck off.
This is the first time I've been able to see a link that directly lets me put my point across about 'power'Macs.
I've let them know, politely but firmly want I want to pay for. I've also dropped a few hints about why I and maybe other Mac 'faithfuls' aint buying 'power'Macs in droves anymore...
Give them your penny's worth if you like...
Lemon Bon Bon </strong><hr></blockquote>
I gave them some feedback, but I didn't leave any hints or penny feedback. I told them that until Powermacs have a competitive performance/price ratio, I'll be upgrading my Sawtooth with a CPU upgrade card.
You know nothing pisses off Steve Jobs more than hearing about Mac users UPGRADING their Powermacs with new CPUs. Maybe the only thing he would get more upset about is someone overclocking a 733 MHz G4 Powermac to 1 GHz.
But more importantly, I think this feedback form means two things:
1. Apple is admitting that they don't know why Powermacs aren't selling, and they are taking it to the people to get an answer.
2. A totally new Powermac revision is on the way. Apple is going to do everything they can to get it right, just like they did with the Powerbook when they used a survey to find out what people want most in their Titaniums.
I gave them some feedback, but I didn't leave any hints or penny feedback. I told them that until Powermacs have a competitive performance/price ratio, I'll be upgrading my Sawtooth with a CPU upgrade card.
You know nothing pisses off Steve Jobs more than hearing about Mac users UPGRADING their Powermacs with new CPUs. Maybe the only thing he would get more upset about is someone overclocking a 733 MHz G4 Powermac to 1 GHz.
But more importantly, I think this feedback form means two things:
1. Apple is admitting that they don't know why Powermacs aren't selling, and they are taking it to the people to get an answer.
2. A totally new Powermac revision is on the way. Apple is going to do everything they can to get it right, just like they did with the Powerbook when they used a survey to find out what people want most in their Titaniums.
3. This confirms that a Power 5 is on the way!
</strong><hr></blockquote>
He's right! Apple has come a long way. Do all of you really think that they have nothing for us. Remember what steve said about waiting for 10.2 before they would have options. Apple can only do so much at one time. Now that 10.2 is out we will see new CPU's soon enough. The Mother Board has been redesigned in the powermac it is obvious that apple will have a new CPU soon. They had to wait for the right time and they had to wait for the CPU to be done. If you look around all the evidence is there to point to a new CPU within the next 6 months or so. All the other times it was just hope that made us think we would have a new CPU but now it makes sense. Stop worrying so much and relax. It will all be fine... I hope.
[quote] In recent days, certain Reports to an Apple developer list purported to report Remarks from IBM Engineering Manager that Apple had rejected a Scheme to replace Motorola?s ill-starred PowerPC G4 with a Desktop Version of IBM?s 64-bit Power4 chip that has been enriched with Altivec Vector-Processing Technology.
With all due respect, I must tell you that this Notion is a Tissue of Lies of the most Unspeakable Brazenness: Despite what untutored Nay-sayers may claim, Apple and IBM are indeed working closely on this excellent Chip.
The heretofore-unspoken name of IBM?s G4 Replacement is GPUL (short for GigaProcessor UltraLite). It is Multi-core, 64-bit Microprocessor that supports Vector/SIMD Multimedia eXtension (VMX), the generic Title for the Technologies comprising Altivec (a k a Velocity Engine).
This Most Worthy CPU also being fine-tuned to work with the Apple Processor Interconnect bus (ApplePI), Apple?s Replacement for the venerable MaxBus. Mac OS X is already booting happily on Prototype Units. And contrary to the scandalous Poppycock mouthed by some Detractors, Emissaries from IBM have visited Apple Headquarters in recent weeks to apprise them of the progress of engineering GPUL for the illustrious Mac Architecture.
As to the Timeline for Delivery of this Powerful new Weapon: Alas, GPUL does not appear to be in the Cards for January?s Generation of Apple Systems. IBM had Mac OS X running on GPUL by November 2001, was delivering Prototypes to Apple by March of this Year, and was testing prototype Mac Hardware Systems and core Apple Application Software by April. Nevertheless, GPUL not scheduled to be delivered ? even to Apple and other Worthy OEM Partners ? until later in 2003, meaning that January?s announced end of the line for Mac OS 9 booting will not coincide with the Brave New World of 64-bit Macs.
Comments
With all this talk about Motorola now, I'm sorry to further bifurcate this thread!
Question is will it have an assortment of embedded-specific features (built-in ethernet, serial ports, etc.) or will it be more taylored to a profession desktop use directed towards customers like Apple?
<hr></blockquote>
<a href="http://e-www.motorola.com/webapp/sps/site/prod_summary.jsp?code=MPC8560&nodeId=03M9430304504 67M98657" target="_blank">MPC8460</a>
Three FCCs supporting:
*\tUp to 155 Mbps ATM SAR-AAL0, AAL1, AAL2, AAL3/4, AAL5
*\t10/100 Mbps Ethernet (up to three) IEEE 802.3X
*\t45 Mbps HDLC/transparent (up to three)
Two triple-speed Ethernet controllers (TSECs) supporting 10/100/1000 Mbps Ethernet (IEEE 802.3, 802.3u, 802.3x, 802.3z, and 802.3ac compliant) with two GMII/TBI/RGMII interfaces
Serial Communications Controllers\t4\t--
If there is just one insider who would just say something re-assuring, please do.
It's making you crazy? Everybody else here snapped long ago. We're all nuts.
Lemon Bon Bon <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
This is the first time I've been able to see a link that directly lets me put my point across about 'power'Macs.
I've let them know, politely but firmly want I want to pay for. I've also dropped a few hints about why I and maybe other Mac 'faithfuls' aint buying 'power'Macs in droves anymore...
Give them your penny's worth if you like...
Lemon Bon Bon
<strong>I've also dropped a few hints about why I and maybe other Mac 'faithfuls' aint buying 'power'Macs in droves anymore...
</strong><hr></blockquote>
Why hints? Just tell them THE TRUTH you bast@rd
[ 09-12-2002: Message edited by: Leonis ]</p>
take a look at
<a href="http://www.realworldtech.com/listing.cfm?section=rwtlabs&subject=tech"; target="_blank">http://www.realworldtech.com/listing.cfm?section=rwtlabs&subject=tech</a>;
and they show that the DDR advantage is on the order of 10-20% at the most.
and while there there be sure to read the article "Apple's Power Failure"
<a href="http://www.realworldtech.com/page.cfm?AID=RWT051400000000"; target="_blank">http://www.realworldtech.com/page.cfm?AID=RWT051400000000</a>;
The content is still current and more urgent than it was then in 2000.
THT, do you think Motorola is trying to wean it's developers off MPX perminantly by not advancing it's design? Ultimatly they want all developers using RIO with in a specific time frame they have, maybe?</strong>
Yes partially. Moto is moving to on-die DDR memory controllers for its embedded processor strategy (as seen in the 8460 and 8540) and recognizes that improving the MPX bus is a waste of time for their main market (and it would be competitor to RapidIO to boot). So I think Moto will consider MPX as legacy tech whence 8460 ships.
I think Moto just doesn't want to invest in system logic controllers anymore and have done what they've needed to stop developing them. Moved the memory controller to the chip and contracted 3rd parties to build RapidIO chips and interfaces. This doesn't leave any room for MPX processor bus development anymore.
<strong>That's all fine and good, but where does that leave Apple? The implication is that they want to stop further G4 development and a die shrink is not even planned because they are migrating to the 85XX line.</strong>
Maybe. I don't why Moto would even consider not producing a 7455 based G4 on 0.13u though. It performs better than either of the 8460 or 8540, and the 7455 would be in terrific die size and power consumption territory at 0.13u.
Obviously, it leaves Apple with their drawers down for awhile. Hey, I said Apple should have taken over processor development 4 years ago. They should have listened to me.
<strong>This means soon (and before they discontinue the G4 74XX line) Motorola intends to bring to market an 85XX with powerful floating point units, Altivec. Not if but when.</strong>
Well, I believe it's still 75xx on the Moto roadmap for host processors. 85xx are monikers for communications processors (8540 et al). And powerful and 85xx don't come to mind as well. These are non-starters for Apple.
<strong>Question is will it have an assortment of embedded-specific features (built-in ethernet, serial ports, etc.) or will it be more taylored to a profession desktop use directed towards customers like Apple?</strong>
I don't know. Apple needs 0.13 micron G4 chips right now. And the news has been scarce for Moto HiP 7 chips.
The only tidbit of real actual information available is this IBM desktop PPC. If IBM can ship a single core Power4 with AltiVec compatible SIMD units, I don't think a 0.13u G4 without modifications has much of chance. A 0.13u single Power4 would be in the 1.3 GHz to 2 GHz range, with more powerful scalar execution resources. SIMD is an unknown right now. Since Peter Sandon of Gekko PPC fame is doing this IBM desktop PPC, I'm not getting good vibes that this is separate SIMD unit; rather, the 2 Power FP units will be used to compute 4 single-precision ops. This could be sweet actually if the single precision ops can be independent.
But...why? What went wrong? Yeesh. They're a multi-billion dollar company. What the hell are they playing at?
THT, some incisive posts there about some of the peformance issues the G4 faces. Fascinating to read and virtually impossible to rebuke. I like the way you nailed the DDR issue and the many aspects of cpu performance that could improve the G4. Cold hard light of day stuff. I like it!
"Why hints? Just tell them THE TRUTH you bast@rd"
Aha! Leonis!
Speaking generally, (being one for digression an' all...), I have to confess I almost buckled in my stand off with Apple's 'power'Mac. Indesign 2 free with any 'power'Mac. That's crazy? It costs the best part of £600 here in the UK!!! Buy a low end 'power'Mac and you get a superb piece of software. That's a killer promo' from Apple and Adobe. Watch out, Quark...
Maybe Apple's just beginning the 'fight back'?
(I know as soon as I cave in the G5 will hit soon after...still...Indesign...anybody else tempted? I suppose you could always dump the antiquated tower on ebay...if/when the next revision comes along...)
(Keep the pics coming Leonis!)
Lemon Bon Bon
"Well, I believe it's still 75xx on the Moto roadmap for host processors. 85xx are monikers for communications processors (8540 et al). And powerful and 85xx don't come to mind as well. These are non-starters for Apple."
Do you believe the G5 or last iteration of the G4 will be the '7500'? Wasn't this yanked from Motorola's site a while back? I always thought 7500 was the 'G5'. Maybe it's just not enough. And Apple will hopskotch on. But I can't see Apple not using it if Moto's done the work on it. Let's face it. A 7500 would be darn better than what they've got now!
"I don't know. Apple needs 0.13 micron G4 chips right now. And the news has been scarce for Moto HiP 7 chips."
Yes. But what happened? Surely you can't develop something like HiP 7 and do nothing with it?
"The only tidbit of real actual information available is this IBM desktop PPC. If IBM can ship a single core Power4 with AltiVec compatible SIMD units, I don't think a 0.13u G4 without modifications has much of chance."
The Register had an article a while back on this chip. Said it was slated for 2003 Jan'. .13 G4, 7500 with Rio. But other than that, just a G4, no fp improvement or anything extra. Clock speeds of...was it 1.6?
"A 0.13u single Power4 would be in the 1.3 GHz to 2 GHz range, with more powerful scalar execution resources. SIMD is an unknown right now. Since Peter Sandon of Gekko PPC fame is doing this IBM desktop PPC, I'm not getting good vibes that this is separate SIMD unit; rather, the 2 Power FP units will be used to compute 4 single-precision ops. This could be sweet actually if the single precision ops can be independent."
But I'd rather take the latter option you discuss. But the '7500' wouldn't be a bad option to take in Jan' as we're pretty much feel the Power4 lite won't hit until next Summer. But I'd like to be proven wrong of course...
Lemon Bon Bon
Conclusion
After five years of keeping PowerPC performance more or less in lockstep with x86, IBM and Motorola?s recent play-it-safe strategy of not straying too far from the design parameters of the high-end embedded control market has backfired. The intense struggle between Intel and AMD for technical supremacy in the x86 market has pushed their processors up into the territory of workstation class RISC processors in terms of sophistication, design aggressiveness, and performance. The lack of urgency on the part of IBM and Motorola has caused the PowerPC to fall significantly behind the top of the line MPUs from Intel and AMD.
The nearly effective lack of applications that require GHz level processors has meant that the damage to the competitiveness of PowerPC and Apple?s Power Mac line may be more of a perception than reality. It is up to Apple Computer, a company that is often more successful managing perception than reality, to overcome the burden of a large and widening performance gap by continuing to concentrate on difficult to quantify but desirable aspects of its products, such as system level packaging aesthetics, low noise and heat output, and freedom from Windows. The key to continued viability is retention of the bulk of the Mac user base while attracting the newcomers to computing who are more repelled by beige boxes than attracted by stratospheric clock rates."
I couldn't resist cutting and pasting this bit.
That about sums up Apple's position about right now. How a G5 from IBM would rock this article to its very core.
Thanks DrBoar. Good link.
Lemon Bon Bon
First, although, the G4 has fallen behind the high end Intel and AMD chips it is still not slow. The G4 is sufficient for almost any task a computer buyer does.
Second 10.2 is not slow in the least. I have often tried doing the same tasks in 9 as in 10.2 and 10.2 was faster if not the same at 80% of the tasks. I think the reason 10.2 is slower in some applications is not so much the OS its self but the fact that the applications have not been written very well. This is most likely because the programmers are still having to learn how to write good code for this new OS.
Third, Yes motorola has been working on the next PPC chip for apple but so has IBM. Motorola is not as interested in the idea as IBM and apple is not as interested in Motorola. The next chip will most likely come from IBM because of their obvious interest in working with apple. Along with obvious other proof circulating around the Internet. All the information needed to come to this conclusion is out there and there is plenty of it. I suspect it will still be about 8 months though before we will see the G5. If any one has any inside info on a release date before this just post "NO". I don't have a good time for release other than 8 months but sometimes I feel like it will be sooner.
And finally the reason it is so hard too pin down where the G5 will come from is because apple has been working with many companies on the matter and we are all looking for one answer. Apple has everything set up just right. All they need now is a good economy and a good co partner to design their chips. They are being very careful with this decision. Motorola, AMD, and IBM have all been working with apple on this matter. But it looks like IBM came out on top. (AMD would be the least likely because of obvious complications in going x86)
<strong>Yes. But what happened? Surely you can't develop something like HiP 7 and do nothing with it? </strong><hr></blockquote>
G5s are being fabbed using 130nm. Obviously, Motorola has decided that Apple is less important to keep around than having the best in embedded CPUs.
Brato
<strong>
If your talking about an IBM cpu, who knows. But if your talking about a G5 from Motorola, that is based on the MPC 8540/8560 architecture(re: Rapid I/O, Ocean switched web, etc.), I'm inclined to believe the end of 2003 date.
</strong><hr></blockquote>
I'm changing my mind on this. Been thinking about it and seems the rumor generated by the Nasa employee maybe more truth that fiction. The IBM chip is a 64 bit processor, may not implement Altivec and hasn't even been presented. We'll see.
However, if I remember correctly the old Motorola roadmap for the G5 indicated it was to be on a 0.1µ process. With the new alliance between Motorola/Phillips/STMicroelectronics claming they would be using a 0.09µ process to " start production of a high-performance chip by the fourth quarter of this year and a lower performing version in the third quarter of 2003."(<a href="http://www.reuters.com/news_article.jhtml?type=technologynews&StoryID=139 1841" target="_blank">Link</a>) Maybe, just maybe, the delay in sampling the MPC85XX was due to power consumption and heat genterated when manufactured on a 0.13µ process. The press release did not say which facility will produce the processors, it could very well be they will not be produced by Motorola, but by either Phillips and/or STMicroelectronics(re: do they have better fabs??). Motorola has announced they would be shifting production.
The basic architecture of the MPC74XX hasn't really changed in quite some time. Rumors indicated Motorola had problems implementing the 0.13µ process. The extreme size of the heat sinks in the new Powermacs is not the direction Motorola/Apple would normally go.
If most of Motorola's/Apple's efforts have been directed @ a G5, then the G4 would not have any major architectural changes, due to lack of engineering resources. Basically, the G4 needs more pipelines(re: G5 reportedly to have 10 stages), capability of handling DDR frontside bus, better floating point and a smaller process. Too much engineering, when efforts were tied up in G5 development.
Question is,"what did Mr. Chris Belden's comment regarding "a high-performance chip by the fourth quarter of this year " mean?". High end processor for communication, networking, desktops????
I know the feeling.
G5 priority. A quarter 4 for San Fran 2003? You could be onto something there Rick'...
Lemon Bon Bon
We also thought Motorolas G5 would be the bliss of 2002. End of rant.
In parallel, IBM is working on a 64 bit desktop PPC CPU with an Altivec compatible SIMD unit. This chip will be fabbed in IBM's new chip foundry in New York. Full production early 2003, unclear whether it will be fabbed using a 130 nm process or a 90 nm process (only know that production at the foundry will start with 130 nm, but quickly "move to 90-nanometer production".
No, this does not mean that Apple doesn't know what CPU will be used in their next generation Powermacs. People sometimes seem to forget that the G4 has a very long life ahead of it in CONSUMER Macs, i.e., iMacs, iBooks, and of course Powerbooks until IBM can supply Apple with a G5 that doesn't melt the polar ice caps.
Let's say for a moment that IBM cranks out 2 GHz Power 5 CPUs faster than Apple can sell them, beginning in Jan 2003. That would mean a Powermac lineup that finally rocks, and Apple would be selling them like nobody's business. However, the iMac in particular would still need to have a competitive CPU....Apple couldn't very well sell 933 MHz G4 iMacs alongside 2 GHz Powermacs. (well yeah they could and maybe they would, but it would look REALLY bad). So it follows that Apple is probably nagging Motorola to supply them with a G4 that will last another 1-2 years, with the implicit threat that if Moto cannot give Apple a competitive G4 for the iMac, then Apple will simply switch ALL Macs to the Power 5, heat and power consumption be damned. You can bet that Apple has new iMac designs that can handle a Power 5, just in case Motorola fscks up again.
Wht this means is that Apple is finally playing Moto and IBM off against each other, primarily to put the fear of God in Moto. The truth is that Motorola cannot afford to lose Apple's G4 business....they are making a phat profit on Apple now that the G4 is Apple's standard CPU. When Apple migrates the Powermacs to the Power 5, Moto is still counting on G4 sales....but finally Apple is no longer Moto's little bitch. Very soon, Apple will be telling Motorola how things are going to be, and you can bet Steve Jobs is already practicing what he's going to say to Moto executives when he finally tells them to fsck off.
<strong><a href="http://www.apple.com/feedback/powermac.html"; target="_blank">http://www.apple.com/feedback/powermac.html</a>;
This is the first time I've been able to see a link that directly lets me put my point across about 'power'Macs.
I've let them know, politely but firmly want I want to pay for. I've also dropped a few hints about why I and maybe other Mac 'faithfuls' aint buying 'power'Macs in droves anymore...
Give them your penny's worth if you like...
Lemon Bon Bon
I gave them some feedback, but I didn't leave any hints or penny feedback. I told them that until Powermacs have a competitive performance/price ratio, I'll be upgrading my Sawtooth with a CPU upgrade card.
You know nothing pisses off Steve Jobs more than hearing about Mac users UPGRADING their Powermacs with new CPUs. Maybe the only thing he would get more upset about is someone overclocking a 733 MHz G4 Powermac to 1 GHz.
But more importantly, I think this feedback form means two things:
1. Apple is admitting that they don't know why Powermacs aren't selling, and they are taking it to the people to get an answer.
2. A totally new Powermac revision is on the way. Apple is going to do everything they can to get it right, just like they did with the Powerbook when they used a survey to find out what people want most in their Titaniums.
3. This confirms that a Power 5 is on the way!
<strong>
I gave them some feedback, but I didn't leave any hints or penny feedback. I told them that until Powermacs have a competitive performance/price ratio, I'll be upgrading my Sawtooth with a CPU upgrade card.
You know nothing pisses off Steve Jobs more than hearing about Mac users UPGRADING their Powermacs with new CPUs. Maybe the only thing he would get more upset about is someone overclocking a 733 MHz G4 Powermac to 1 GHz.
But more importantly, I think this feedback form means two things:
1. Apple is admitting that they don't know why Powermacs aren't selling, and they are taking it to the people to get an answer.
2. A totally new Powermac revision is on the way. Apple is going to do everything they can to get it right, just like they did with the Powerbook when they used a survey to find out what people want most in their Titaniums.
3. This confirms that a Power 5 is on the way!
He's right! Apple has come a long way. Do all of you really think that they have nothing for us. Remember what steve said about waiting for 10.2 before they would have options. Apple can only do so much at one time. Now that 10.2 is out we will see new CPU's soon enough. The Mother Board has been redesigned in the powermac it is obvious that apple will have a new CPU soon. They had to wait for the right time and they had to wait for the CPU to be done. If you look around all the evidence is there to point to a new CPU within the next 6 months or so. All the other times it was just hope that made us think we would have a new CPU but now it makes sense. Stop worrying so much and relax. It will all be fine... I hope.
From <a href="http://www.macedition.com/nmr/nmr_20020914.php"; target="_blank">MacEdition</a>
[quote] In recent days, certain Reports to an Apple developer list purported to report Remarks from IBM Engineering Manager that Apple had rejected a Scheme to replace Motorola?s ill-starred PowerPC G4 with a Desktop Version of IBM?s 64-bit Power4 chip that has been enriched with Altivec Vector-Processing Technology.
With all due respect, I must tell you that this Notion is a Tissue of Lies of the most Unspeakable Brazenness: Despite what untutored Nay-sayers may claim, Apple and IBM are indeed working closely on this excellent Chip.
The heretofore-unspoken name of IBM?s G4 Replacement is GPUL (short for GigaProcessor UltraLite). It is Multi-core, 64-bit Microprocessor that supports Vector/SIMD Multimedia eXtension (VMX), the generic Title for the Technologies comprising Altivec (a k a Velocity Engine).
This Most Worthy CPU also being fine-tuned to work with the Apple Processor Interconnect bus (ApplePI), Apple?s Replacement for the venerable MaxBus. Mac OS X is already booting happily on Prototype Units. And contrary to the scandalous Poppycock mouthed by some Detractors, Emissaries from IBM have visited Apple Headquarters in recent weeks to apprise them of the progress of engineering GPUL for the illustrious Mac Architecture.
As to the Timeline for Delivery of this Powerful new Weapon: Alas, GPUL does not appear to be in the Cards for January?s Generation of Apple Systems. IBM had Mac OS X running on GPUL by November 2001, was delivering Prototypes to Apple by March of this Year, and was testing prototype Mac Hardware Systems and core Apple Application Software by April. Nevertheless, GPUL not scheduled to be delivered ? even to Apple and other Worthy OEM Partners ? until later in 2003, meaning that January?s announced end of the line for Mac OS 9 booting will not coincide with the Brave New World of 64-bit Macs.
<hr></blockquote>