But if the hopes are held back again, will the "again" bring this amount of attention in January, or will some be looking elsewhere by that time?</strong><hr></blockquote>
I'll be upgrading my PM450 with a Powerlogix 1ghz chip and then going my merry way. Never to look at another mac.
(wait for it...)
Ahh, I just can't help myself. It's a guilty pleasure. It's like slowing down to see a car wreck. You're not proud of it, but you do it any way.
"Conventional" probably refers to the traditional shared front-side bus (i.e. MPX). Anything else would be unconventional (HyperTransport, RapidIO, a dedicated FSB like the PC, an on-chip memory controller, etc).
I would expect something different than Xserve based on Dorsal's comment. If you believe him, that is. I'll have a lot more faith if he is correct in some real way about the machines at or near MWNY.</strong>
I believe Dorsal was hinting at an on-die memory controller like as planned for the Moto's 8540. I'm not as much of a fan of this design anymore, considering Apple's market that is.
I also disagree that they need to drop Motorola. What they need to do is ensure a future supply of powerful processors for the PowerMac line, and a solid supply of consumer and portable processors. If this means dropping Motorola, then so be it. I suspect the situation is much more complex, however. Ideally both Motorola and IBM would deliver chips that Apple can choose from so that Apple doesn't get stuck back in another single-source situation. From all that has been tossed around lately I could easily see that next year at this time Apple might be swimming in processor options: IBM SIMD equipped low-power G3s for the iBook, Motorola G4s for the consumer line (possibly using RapidIO & on-chip memory controller), and IBM POWER-derived chips for the PowerMac line.</strong><hr></blockquote>
I agree with your disagreement.
The AIM alliance seems to have been damaged mightily by Apple choosing MOT's Altivec and G4 processor - IBM suddenly felt like a third wheel and unappreciated, so they took their toys and went home. It was not so much an alliance as a three-way competition, each trying to get for itself at the expense of the others. Thus, I think Apple's problems with having MOT as its sole supplier of G4s is really its own short-sighted fault. It sounds like the breach with IBM is being healed, but at enormous cost to Apple as it struggles with MOT.
IMHO, the AIM alliance could be revitalised by simply focusing on R&D (the Somerset center), funded by all three and agnostic of company - all ideas are pooled and the on-site management goes with what will work best, without pressure from the parent companies to adopt a "pet" technology. All three companies would then have free license to all technology generated by it. (Yes, I wear rose-colored glasses and live in an ivory tower, but a guy can dream, can't he?) Fabbing would be done on a contractual basis with whatever company (MOT, IBM, AMD, or even Intel) had the capacity to do it. The chips would thus not be a "Motorola" or "IBM" or any other brand - just "from the AIM alliance". This may have been the original intent of it, but it's certainly not the way it's turned out.
I would love to see the AIM research center turn into a "skunkworks" like the Bell labs - cranking out outrageously cool inventions with no immediate practical use, but which revolutionise the world once they're made into real products. It may happen yet.
maybe I'm stupid but is it really difficult to have a dedicated bus for each processor (with the current G4) ? That means in this way each processor can use a full 1 GB/S of bandwidth and still with a FSB of 133 Mhz...
Let's say that Apple has a new mobo and it's been ready for a while now. It's something radically different, and has much higher thruput/bandwidth, maybe task-specific DSPs, FW2, etc. Only problem is...it's been designed for the new CPUs (G5/POWER/whatever). Then along comes IBM/Moto/?, who says....oops, the new CPU won't be ready in quantity 'til December.
Apple knows that's a problem, so they modify the mobo somewhat to use the current G4, and release this "stopgap" at MWNY. Say, a G4 at 1.2/1.4, but with a much better/faster mobo...how much difference would this make?
Then MWSF hits, with a "proven" mobo to drop into the xServe, and a new CPU (maybe competition for upcoming AMD/Intel offerings...the timing sounds right) for another big speed jump for the server/pro lines.
Does this seem reasonable?
Could this bring them in line (or better) with the Wintel world?
Taboo, despite my 'beef' with the current and hugely neglected 'power'Mac specs...I do believe something similar to your scenario could occur. Much as it did as Apple went from G3 to G4. 'Yikes' an' all.
ie I have faith that Apple will address the growing performance gap.
The purchases of 'Shake' and other high end apps like 'Chalice' and co indicate that Apple intend to compete in the hardware space soon. This is a clue to better things to come. But...when?
If Apple want their 'workstation' to compete with dual x86 computers running Maya then the chances of seeing some type of 'G5'processor is becoming more realistic as x86 processors march on. Apple can't afford to be on dual 1.2 gig G4s as x86 go 64 bit and 3 gig.
Apple needs their pro' sales...and their healthy mark ups. Sure, Apple isn't immune from the current industry slow down...and are more exposed than most. But if they want their already hurting pro sales not to get squeezed further then they are simply going to have to offer something far more compelling for Apple users and WINTEL 'Switchers' to go Mac. A mere 'bump' aint going to cut it this time.
It won't be the end of the world if we don't get what we want. But I'll be highly surprised if Apple don't finally begin to address the performance issue at Mac New York.
If they want consumers money for £3,000 'power'Macs...they're going to have to innovate their way into pockets.
They got Jagwire (nearly), they got the looks, the sexy drop down case, they got the superdrive...now give us the mobo and G5.
Aint just me either. In Macworld.co.uk poll, majority of mac fans also want the G5...we'll see, eh?
Fact is chaps even if we got a G5 4ghz next month it isn't going to make any difference to Apples market share. People just look at windows do a cost/benifit analysis and buy the fastest they can afford at the time.
Hmmm, slight oversimplification? It's already been proven many times than when Apple releases a product with better "bang for the buck," it's more popular. Sure, a 4ghz G5 wouldn't give them 95% market share, but they'd sure sell a lot of them and their market share would creep upward.
<strong>Hmmm, slight oversimplification? It's already been proven many times than when Apple releases a product with better "bang for the buck," it's more popular. Sure, a 4ghz G5 wouldn't give them 95% market share, but they'd sure sell a lot of them and their market share would creep upward.</strong><hr></blockquote>
I would buy one
ps.. where's dorsal? It's sad but I feel like a crack addict who needs a fix (although I'm sure that's worse than what I'm going though) :eek:
What confuses me is the talk about the towers at MWNY being the 'stop gap' release. Excuse me, um but I was under the impression that the stop gap was the 1ghz towers. That's why a ho hum release as Apple knew they sucked and had something better just not ready. To me, that means MWNY will have something new and awesome otherwise we're looking at 2 stop gap releases this year for towers? And this after the pro sales suck already? Sorry, but I don't see Apple doing that. Not sure what I see them doing however, lol
Well you know if there is a stop-gap, hack, whatever.... how the hell NOW is Apple going to do their famous Photoshop bake-offs.... Makes me think that we'll have something fast enough to win a PhotoShop bake-off. Otherwise, he'll just go, "Oh and we have dual 1.xghz towers now!" But its interesting to think about. They ALWAYS have bake offs with the PowerMacs at Expos, so they almost have to release something fast enough to win a bake off with a Pentium IV.
What I think all of us would love to see is a new top of the line PMac beating the crap out of a dual top of the line Athlon at a quick 3D render. Now that would be a bake off to remember!
<strong>What I think all of us would love to see is a new top of the line PMac beating the crap out of a dual top of the line Athlon at a quick 3D render. Now that would be a bake off to remember!
</strong><hr></blockquote>
Introducing the new PowerMac G5 with Bionic arm! Watch it sneak up on the unsuspecting Athlon whilst it is distracted rendering a 3D scene gratiutous violence ensues, ending with the Athlon box scattered in iPod sized pieces on the floor and just wait till you see the new iMac with bionic leg! It kicks the Duron's a$$! :eek:
<strong>Well you know if there is a stop-gap, hack, whatever.... how the hell NOW is Apple going to do their famous Photoshop bake-offs.... Makes me think that we'll have something fast enough to win a PhotoShop bake-off. </strong><hr></blockquote>
Just an interesting tidbit -- I was mucking around with Photoshop 7 on my LCD iMac, when I noticed that any time I ran a filter, scaled an image, or did anything else terribly CPU intensive, the audio and/or video would stutter rather badly in iTunes or the DVD player.
This says to me that Adobe is using a little hack to boost the priority of their CPU intensive threads so that they will get more cycles. I just found it a little interesting; it appears Adobe is doing something a bit special that few other OS X apps do.
I haven't confirmed this empirically, but I'm fairly sure it is accurate given the symptoms I'm seeing. Normal threads should not have a higher priority the audio threads that are used to play mp3s, CDs, etc.
This says to me that Adobe is using a little hack to boost the priority of their CPU intensive threads so that they will get more cycles. [ 06-20-2002: Message edited by: moki ]</strong><hr></blockquote>
You can raise/lower the priority of a process using:
int setpriority(int which, int who, int prio);
It may also be that on your specific setup, Photoshop generates substantial memory swaping and as the priority of a page fault is higher than reading from a disk file, the audio skips. The only way to test that would be to play the audio from memory.
<strong>Just an interesting tidbit -- I was mucking around with Photoshop 7 on my LCD iMac, when I noticed that any time I ran a filter, scaled an image, or did anything else terribly CPU intensive, the audio and/or video would stutter rather badly in iTunes or the DVD player.</strong><hr></blockquote>
I wonder if they do that in Windows too? That's just lame... I hope they at least make it a user preference.
<strong>The AIM alliance seems to have been damaged mightily by Apple choosing MOT's Altivec and G4 processor - IBM suddenly felt like a third wheel and unappreciated, so they took their toys and went home. It was not so much an alliance as a three-way competition, each trying to get for itself at the expense of the others. Thus, I think Apple's problems with having MOT as its sole supplier of G4s is really its own short-sighted fault.</strong><hr></blockquote>
I remember hearing at the time that IBM's engineering group felt that a SIMD vector unit was counter to the philosophy of RISC and therefore shouldn't be added to the PowerPC. Once the G4 was delivered and they had a chance to play with it the were "impressed" and started to change their tune. Since then more and more news of IBM SIMD units has been cropping up. So I lay the blame for this rift largely at IBM's feet for not having the vision to see that alternative computing models can deliver serious performance improvements. Now, of course, IBM has come around 180 degrees and is working on things like the "Cell" project, their own SIMD, and they have done the Gekko for Nintendo.
You can raise/lower the priority of a process using:
int setpriority(int which, int who, int prio);</strong><hr></blockquote>
Unfortunately, there are about a half dozen different APIs for doing this with Mach, and they don't all work as expected. I've done something similar to what Adobe seems to be doing... I'm fairly certain that if they are doing what I suspect, it isn't just setpriority().
[quote]Originally posted by *l++:
<strong>It may also be that on your specific setup, Photoshop generates substantial memory swaping and as the priority of a page fault is higher than reading from a disk file, the audio skips. The only way to test that would be to play the audio from memory.</strong><hr></blockquote>
That certainly could be, but I'm fairly certain it isn't. I'll do some real investigation tomorrow.
<strong>I wonder if they do that in Windows too? That's just lame... I hope they at least make it a user preference.</strong><hr></blockquote>
They most likely coult do it in Windows as well, at least NT/2000/XP, but I am not sure whether Win95 has something like that. As far as I know the Windows APIs support priority settings, but usually the user can set the priority from the task manager, anywhere from "low" to "critical". Which I always do for rendering or handling big photoshop files anyway..
Comments
<strong>*snip*
But if the hopes are held back again, will the "again" bring this amount of attention in January, or will some be looking elsewhere by that time?</strong><hr></blockquote>
I'll be upgrading my PM450 with a Powerlogix 1ghz chip and then going my merry way. Never to look at another mac.
(wait for it...)
Ahh, I just can't help myself. It's a guilty pleasure. It's like slowing down to see a car wreck. You're not proud of it, but you do it any way.
[ 06-20-2002: Message edited by: tsukurite ]</p>
"Conventional" probably refers to the traditional shared front-side bus (i.e. MPX). Anything else would be unconventional (HyperTransport, RapidIO, a dedicated FSB like the PC, an on-chip memory controller, etc).
I would expect something different than Xserve based on Dorsal's comment. If you believe him, that is. I'll have a lot more faith if he is correct in some real way about the machines at or near MWNY.</strong>
I believe Dorsal was hinting at an on-die memory controller like as planned for the Moto's 8540. I'm not as much of a fan of this design anymore, considering Apple's market that is.
[ 06-20-2002: Message edited by: THT ]</p>
<strong>
I also disagree that they need to drop Motorola. What they need to do is ensure a future supply of powerful processors for the PowerMac line, and a solid supply of consumer and portable processors. If this means dropping Motorola, then so be it. I suspect the situation is much more complex, however. Ideally both Motorola and IBM would deliver chips that Apple can choose from so that Apple doesn't get stuck back in another single-source situation. From all that has been tossed around lately I could easily see that next year at this time Apple might be swimming in processor options: IBM SIMD equipped low-power G3s for the iBook, Motorola G4s for the consumer line (possibly using RapidIO & on-chip memory controller), and IBM POWER-derived chips for the PowerMac line.</strong><hr></blockquote>
I agree with your disagreement.
The AIM alliance seems to have been damaged mightily by Apple choosing MOT's Altivec and G4 processor - IBM suddenly felt like a third wheel and unappreciated, so they took their toys and went home. It was not so much an alliance as a three-way competition, each trying to get for itself at the expense of the others. Thus, I think Apple's problems with having MOT as its sole supplier of G4s is really its own short-sighted fault. It sounds like the breach with IBM is being healed, but at enormous cost to Apple as it struggles with MOT.
IMHO, the AIM alliance could be revitalised by simply focusing on R&D (the Somerset center), funded by all three and agnostic of company - all ideas are pooled and the on-site management goes with what will work best, without pressure from the parent companies to adopt a "pet" technology. All three companies would then have free license to all technology generated by it. (Yes, I wear rose-colored glasses and live in an ivory tower, but a guy can dream, can't he?) Fabbing would be done on a contractual basis with whatever company (MOT, IBM, AMD, or even Intel) had the capacity to do it. The chips would thus not be a "Motorola" or "IBM" or any other brand - just "from the AIM alliance". This may have been the original intent of it, but it's certainly not the way it's turned out.
I would love to see the AIM research center turn into a "skunkworks" like the Bell labs - cranking out outrageously cool inventions with no immediate practical use, but which revolutionise the world once they're made into real products. It may happen yet.
My $0.02, YMMV.
TJM
It will be an improvement, no?
Let's say that Apple has a new mobo and it's been ready for a while now. It's something radically different, and has much higher thruput/bandwidth, maybe task-specific DSPs, FW2, etc. Only problem is...it's been designed for the new CPUs (G5/POWER/whatever). Then along comes IBM/Moto/?, who says....oops, the new CPU won't be ready in quantity 'til December.
Apple knows that's a problem, so they modify the mobo somewhat to use the current G4, and release this "stopgap" at MWNY. Say, a G4 at 1.2/1.4, but with a much better/faster mobo...how much difference would this make?
Then MWSF hits, with a "proven" mobo to drop into the xServe, and a new CPU (maybe competition for upcoming AMD/Intel offerings...the timing sounds right) for another big speed jump for the server/pro lines.
Does this seem reasonable?
Could this bring them in line (or better) with the Wintel world?
Just idle speculation....
ie I have faith that Apple will address the growing performance gap.
The purchases of 'Shake' and other high end apps like 'Chalice' and co indicate that Apple intend to compete in the hardware space soon. This is a clue to better things to come. But...when?
If Apple want their 'workstation' to compete with dual x86 computers running Maya then the chances of seeing some type of 'G5'processor is becoming more realistic as x86 processors march on. Apple can't afford to be on dual 1.2 gig G4s as x86 go 64 bit and 3 gig.
Apple needs their pro' sales...and their healthy mark ups. Sure, Apple isn't immune from the current industry slow down...and are more exposed than most. But if they want their already hurting pro sales not to get squeezed further then they are simply going to have to offer something far more compelling for Apple users and WINTEL 'Switchers' to go Mac. A mere 'bump' aint going to cut it this time.
It won't be the end of the world if we don't get what we want. But I'll be highly surprised if Apple don't finally begin to address the performance issue at Mac New York.
If they want consumers money for £3,000 'power'Macs...they're going to have to innovate their way into pockets.
They got Jagwire (nearly), they got the looks, the sexy drop down case, they got the superdrive...now give us the mobo and G5.
Aint just me either. In Macworld.co.uk poll, majority of mac fans also want the G5...we'll see, eh?
Lemon Bon Bon <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
<strong>Hmmm, slight oversimplification? It's already been proven many times than when Apple releases a product with better "bang for the buck," it's more popular. Sure, a 4ghz G5 wouldn't give them 95% market share, but they'd sure sell a lot of them and their market share would creep upward.</strong><hr></blockquote>
I would buy one
ps.. where's dorsal? It's sad but I feel like a crack addict who needs a fix (although I'm sure that's worse than what I'm going though) :eek:
[ 06-20-2002: Message edited by: Blizaine ]</p>
What I think all of us would love to see is a new top of the line PMac beating the crap out of a dual top of the line Athlon at a quick 3D render. Now that would be a bake off to remember!
[ 06-20-2002: Message edited by: ouroboros ]</p>
<strong>What I think all of us would love to see is a new top of the line PMac beating the crap out of a dual top of the line Athlon at a quick 3D render. Now that would be a bake off to remember!
</strong><hr></blockquote>
Introducing the new PowerMac G5 with Bionic arm! Watch it sneak up on the unsuspecting Athlon whilst it is distracted rendering a 3D scene gratiutous violence ensues, ending with the Athlon box scattered in iPod sized pieces on the floor and just wait till you see the new iMac with bionic leg! It kicks the Duron's a$$! :eek:
<strong>Well you know if there is a stop-gap, hack, whatever.... how the hell NOW is Apple going to do their famous Photoshop bake-offs.... Makes me think that we'll have something fast enough to win a PhotoShop bake-off. </strong><hr></blockquote>
Just an interesting tidbit -- I was mucking around with Photoshop 7 on my LCD iMac, when I noticed that any time I ran a filter, scaled an image, or did anything else terribly CPU intensive, the audio and/or video would stutter rather badly in iTunes or the DVD player.
This says to me that Adobe is using a little hack to boost the priority of their CPU intensive threads so that they will get more cycles. I just found it a little interesting; it appears Adobe is doing something a bit special that few other OS X apps do.
I haven't confirmed this empirically, but I'm fairly sure it is accurate given the symptoms I'm seeing. Normal threads should not have a higher priority the audio threads that are used to play mp3s, CDs, etc.
[ 06-20-2002: Message edited by: moki ]</p>
<strong>
This says to me that Adobe is using a little hack to boost the priority of their CPU intensive threads so that they will get more cycles. [ 06-20-2002: Message edited by: moki ]</strong><hr></blockquote>
You can raise/lower the priority of a process using:
int setpriority(int which, int who, int prio);
It may also be that on your specific setup, Photoshop generates substantial memory swaping and as the priority of a page fault is higher than reading from a disk file, the audio skips. The only way to test that would be to play the audio from memory.
<strong>Just an interesting tidbit -- I was mucking around with Photoshop 7 on my LCD iMac, when I noticed that any time I ran a filter, scaled an image, or did anything else terribly CPU intensive, the audio and/or video would stutter rather badly in iTunes or the DVD player.</strong><hr></blockquote>
I wonder if they do that in Windows too? That's just lame... I hope they at least make it a user preference.
<strong>The AIM alliance seems to have been damaged mightily by Apple choosing MOT's Altivec and G4 processor - IBM suddenly felt like a third wheel and unappreciated, so they took their toys and went home. It was not so much an alliance as a three-way competition, each trying to get for itself at the expense of the others. Thus, I think Apple's problems with having MOT as its sole supplier of G4s is really its own short-sighted fault.</strong><hr></blockquote>
I remember hearing at the time that IBM's engineering group felt that a SIMD vector unit was counter to the philosophy of RISC and therefore shouldn't be added to the PowerPC. Once the G4 was delivered and they had a chance to play with it the were "impressed" and started to change their tune. Since then more and more news of IBM SIMD units has been cropping up. So I lay the blame for this rift largely at IBM's feet for not having the vision to see that alternative computing models can deliver serious performance improvements. Now, of course, IBM has come around 180 degrees and is working on things like the "Cell" project, their own SIMD, and they have done the Gekko for Nintendo.
<strong>
You can raise/lower the priority of a process using:
int setpriority(int which, int who, int prio);</strong><hr></blockquote>
Unfortunately, there are about a half dozen different APIs for doing this with Mach, and they don't all work as expected. I've done something similar to what Adobe seems to be doing... I'm fairly certain that if they are doing what I suspect, it isn't just setpriority().
[quote]Originally posted by *l++:
<strong>It may also be that on your specific setup, Photoshop generates substantial memory swaping and as the priority of a page fault is higher than reading from a disk file, the audio skips. The only way to test that would be to play the audio from memory.</strong><hr></blockquote>
That certainly could be, but I'm fairly certain it isn't. I'll do some real investigation tomorrow.
[ 06-21-2002: Message edited by: Programmer ]</p>
<strong>I wonder if they do that in Windows too? That's just lame... I hope they at least make it a user preference.</strong><hr></blockquote>
They most likely coult do it in Windows as well, at least NT/2000/XP, but I am not sure whether Win95 has something like that. As far as I know the Windows APIs support priority settings, but usually the user can set the priority from the task manager, anywhere from "low" to "critical". Which I always do for rendering or handling big photoshop files anyway..