<strong>Just wondering what impact these announcements on any timelines Motorola and/or IBM have on implementing this technology into processors like the MPC7457-RM and the IBM 75X
If I read this correctly, we may have quite a wait to see any Rapid I/O in any desktop processors. Any one with more knowledge have a better insight??
</strong><hr></blockquote>
Neither IBM nor Moto have these chips on their roadmaps before 2004. I can't get to the URLs you posted but the titles look like they are talking about the serial physical layer, whereas an Apple machine and these desktop processors would almost certainly use the parallel physical layer for performance reasons.
milestone for high-performance interconnect architectures, ECMA International today
announced its new interconnect standard ECMA-342, which includes both parallel and serial
versions of the RapidIO interconnect architecture developed by the members of the RapidIO
Trade Association. This marks the first acceptance by an international standards organization of
the next-generation of open interconnect standards designed to ensure reliability and systemlevel
compatibility in high-performance networking, communications and embedded systems."<hr></blockquote>
Programmer, thanks for the response.
Like I said I'm not sure what this means.
Yes the MPC 7457-RM on the leaked roadmap is projected for 2004, I guess I should have been more general in my statement. It wasn't that long ago the MPC8540 was supposed to be sampling in the second half of 2002. Even though the MPC8540 isn't designed for desktops, at the beginning of 2002 some people in another forum(who shall remain nameless) felt Motorola's desktop version would appear before the MPC8540.
Any way I thought these press releases were interesting.
Yes the MPC 7457-RM on the leaked roadmap is projected for 2004, I guess I should have been more general in my statement. It wasn't that long ago the MPC8540 was supposed to be sampling in the second half of 2002. Even though the MPC8540 isn't designed for desktops, at the beginning of 2002 some people in another forum(who shall remain nameless) felt Motorola's desktop version would appear before the MPC8540.
Any way I thought these press releases were interesting.
</strong><hr></blockquote>
Yes, they are. Its good to see RIO making progress to becoming reality.
I don't think we're going to see a desktop 8xxx series processor from Motorola. In 2004 we may see a RapidIO G4 which would be a significant improvement, especially if it reaches 1.8-2.0 GHz as rumoured. If they take it to 0.09 microns that would be a spectacular chip for notebooks and other places where low power / low heat is needed. RIO will allow fast NUMA architectures so we could see things like 4 or 8 way Xserves where each G4 is at least 2-4 times faster than a current 1 GHz G4, possibly more. Imagine a rack full of those things -- it would be quite the server farm. The 970 will still be the crown jewel, but that doesn't make it ideal for all uses.
Gregg Bartlett was named Deputy Operations Manager for the Crolles2 Project in May 2002. Crolles2 is the name of Motorola?s partnership with STMicroelectronics and Philips for breakthrough technology development and 300mm pilot manufacturing. Gregg has responsibility for all aspects of the Crolles project for Motorola as well as a functional role within the Crolles joint organizational structure.
Prior to this, Gregg was the director of the Advanced Products Research & Development Laboratory (APRDL) in Austin, Texas. APRDL is part of DigitalDNA? Laboratories (DDL), the research and development organization of SPS.
From 1996 to 2002, he held leadership roles in technology and manufacturing partnership environments. First, he worked with Infineon for the White Oak joint venture in Richmond, and most recently with AMD for the HiPERMOS technology development agreement in Austin. Gregg has more than 18-years experience at Motorola with the last three-years in technology development roles in DDL. His focus has been process technology including the CDR1 SiGe:C process running in MOS-11, and HiP-7 currently in production in MOS-13. His experience spans technology development, fab design and construction as well as start-up and ramp.
Gregg attended Kansas State University where he received his B.S. degree in Chemical Engineering in 1983.
December 2002<hr></blockquote>
*bold is mine
what product is Motorola using HiP7 on? I wasn't aware of any.
According to <a href="http://e-www.motorola.com/brdata/PDFDB/docs/SG1001.pdf"; target="_blank">this PDF</a> witch outlines the availibility of 32-bit microprocessors from Motorola Q1-03, the only processors manufactured on HiP7 is the communications processors PowerQUICC II 8270/75/80 witch are expected to be sampled Q2-03.
No sightings of MPC 7457 just yet, but they are expected to be using HiP7 as well.
The PDF is dated Dec 5 2002 and is modified Jan 9 2003.
No new information is included, it just confirms its existence.
<a href="http://www.mdronline.com/events/dinner/"; target="_blank">This URL</a> witch is an invitation to Microprocessor Report 4th Annual Analysts' Choice Awards. The nominee in the category High Performance Embedded Processor 2002 is among others:
Motorola MPC7457
Quite a feat to be nominated for an award in a year in witch it didn't make any appearance..
<strong>Not to change the subject but what is going to happen with Altivec and Motorola? I can't see them not making processors anymore without Altivec.
If I were to speculate, I would guess that they are re-designing the G5 and when it comes out, will be the G6.</strong><hr></blockquote>
I would guess that they aren't designing a new high-end core because it doesn't fit with their business plans. They'll leverage the existing G4 core by adding some of their system-on-a-chip functionality to it (RIO & memory controller), and by ramping its clock rate and shrinking the process its made on. Beyond that I wouldn't count on much...
Quite a feat to be nominated for an award in a year in witch it didn't make any appearance.<hr></blockquote>
I enjoyed that quote. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
It is interesting that the nominations are "recognizing excellence in technology innovation, design, and implementation."
This would imply that there are working samples of the MPC7457 out there(re: implementation). Question, if the MPC7457 is only a MPC7455 manufactured on a 0.13µm process, where does "innovation" and "design" enter into the evaluation?
Maybe more knowledgeable people than myself can surmise a guess as to the possibility there have been any significant design modifications to the MPC7455 included into the MPC7457. Bear in mind the MPC7457 is nominated in the "High-Performance
Embedded Processors" category.
Extended pipelines? Additional FP unit? Increased FSB capabilities? - naww can't be, this is Motorola and Motorola = embedded.
Well, it appears the latest upgrades don't include the MPC7457.
If you go to tech spec's on Apple's website it lists the L2 cache for all 3 processors @ " 256K on-chip L2 cache running at processor speed".
If I were a betting man, this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004.
<strong>Well, it appears the latest upgrades don't include the MPC7457.
If you go to tech spec's on Apple's website it lists the L2 cache for all 3 processors @ " 256K on-chip L2 cache running at processor speed".
If I were a betting man, this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004. </strong><hr></blockquote>
Apple doesn't have to use the 7457 in the PowerMacs when it comes out. They still need a lower power chip for portables and iMacs.
Apple doesn't have to use the 7457 in the PowerMacs when it comes out. They still need a lower power chip for portables and iMacs.</strong><hr></blockquote>
No they don't. Let's see, Apple introduces a 1.6GHz MPC7457 w/ a 200MHz FSB in the iMac and Powerbooks - Professional users descend on Cupertino and burn it to the ground.
Or Apple introduces underclocked MPC7457's @ 1.0 - 1.2GHz w/ a 167MHz FSB in the iMac and Powerbooks and Professional users descend on Cupertino and burn it to the ground.
<strong>No they don't. Let's see, Apple introduces a 1.6GHz MPC7457 w/ a 200MHz FSB in the iMac and Powerbooks - Professional users descend on Cupertino and burn it to the ground.
Or Apple introduces underclocked MPC7457's @ 1.0 - 1.2GHz w/ a 167MHz FSB in the iMac and Powerbooks and Professional users descend on Cupertino and burn it to the ground.
I think I need therapy. Now where's that Prozac.</strong><hr></blockquote>
And if Apple introduces 1.8 GHz 970-based PowerMacs at the same time as the 7457 appears in the low end? Wow, then everybody is happy.
The bad part (to me) is delivery for the 1.42 GHz is 4-6 weeks. What has SLAPple been doing all this time. More interested in secrecy than selling machines. Guess I'll wait to see if the boxes reach Hurricane Force-5 or not.
<strong>Well, it appears the latest upgrades don't include the MPC7457. </strong><hr></blockquote>
No, it doesn't, but that's not all bad news: Did anyone here think the 7455 would get anywhere near 1.4GHz? I was surprised when they got it to 1.25GHz.
This, to me, is an indication that Mot's starting to beat their fabrication problems, and that's good news.
Think about it: The 7455 took us from 667MHz to 1.42GHz. Not shabby at all. How far will the '57 take us, even if the 970 ends up pushing it into the consumer line?
<strong>Think about it: The 7455 took us from 667MHz to 1.42GHz.</strong><hr></blockquote>The 7455 Apollo started with the 1Ghz machines. The 667-733 were 7450s, without SOI.
I'm confused. I was told new G4 was coming but it didn't. Like noted above, I hope it'snot further delayed meaning the new G4 would debut in the fall in place of the 970. Apple could however just leave the new G4 for the consumer models. That would give those models a few bumps which would allow IBM to get it's new G3/Altivec going or to make a smaller 970 version. I'm stuck trying to figure that out.
I'm also still shocked fom the display prices, those beat even the most naively positive guesses.
<strong>And if Apple introduces 1.8 GHz 970-based PowerMacs at the same time as the 7457 appears in the low end? Wow, then everybody is happy.</strong><hr></blockquote>
Yes this is true, but your using logic, I wasn't, I was trying to be humorous.
But, a lot of people were expecting the MPC7457 to be the next upgrade. Soooo, I should have added that the iMac/Powerbooks were to be updated, like say, next week with the MPC757, leaving the Powermacs to fend with the MPC7455's until the IBM 970 appears very very late 2003 or 2004.
Amorph makes very important point concerning the MPC7455 scaling(terminology?) to 1.42GHz. Pretty amazing when you think about it(re: 7 stages and all). Let's see, 0.13µm process adds 30% to the clock speed. That's 1.3 X 1.42GHz = 1.85GHz.
So, next week Apple introduces the new iMacs @ 1.85GHz w/ a 200MHz FSB and Powerbooks @ 1.2GHz w/ a 167MHz FSB. Then, profesional users descend on Cupertino and burn it to the ground.
My feeble attempt @ logic was, [quote]"Well, it appears the latest upgrades don't include the MPC7457...
... this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004."<hr></blockquote>
Comments
<strong>Just wondering what impact these announcements on any timelines Motorola and/or IBM have on implementing this technology into processors like the MPC7457-RM and the IBM 75X
If I read this correctly, we may have quite a wait to see any Rapid I/O in any desktop processors. Any one with more knowledge have a better insight??
</strong><hr></blockquote>
Neither IBM nor Moto have these chips on their roadmaps before 2004. I can't get to the URLs you posted but the titles look like they are talking about the serial physical layer, whereas an Apple machine and these desktop processors would almost certainly use the parallel physical layer for performance reasons.
<strong>I can't get to the URLs you posted.</strong><hr></blockquote>
They can both be found @
<a href="http://www.rapidio.org/home"; target="_blank">http://www.rapidio.org/home</a>;
[quote]"In a significant standardization
milestone for high-performance interconnect architectures, ECMA International today
announced its new interconnect standard ECMA-342, which includes both parallel and serial
versions of the RapidIO interconnect architecture developed by the members of the RapidIO
Trade Association. This marks the first acceptance by an international standards organization of
the next-generation of open interconnect standards designed to ensure reliability and systemlevel
compatibility in high-performance networking, communications and embedded systems."<hr></blockquote>
Programmer, thanks for the response.
Like I said I'm not sure what this means.
Yes the MPC 7457-RM on the leaked roadmap is projected for 2004, I guess I should have been more general in my statement. It wasn't that long ago the MPC8540 was supposed to be sampling in the second half of 2002. Even though the MPC8540 isn't designed for desktops, at the beginning of 2002 some people in another forum(who shall remain nameless) felt Motorola's desktop version would appear before the MPC8540.
Any way I thought these press releases were interesting.
[ 01-25-2003: Message edited by: rickag ]</p>
<strong>
Programmer, thanks for the response.
Like I said I'm not sure what this means.
Yes the MPC 7457-RM on the leaked roadmap is projected for 2004, I guess I should have been more general in my statement. It wasn't that long ago the MPC8540 was supposed to be sampling in the second half of 2002. Even though the MPC8540 isn't designed for desktops, at the beginning of 2002 some people in another forum(who shall remain nameless) felt Motorola's desktop version would appear before the MPC8540.
Any way I thought these press releases were interesting.
</strong><hr></blockquote>
Yes, they are. Its good to see RIO making progress to becoming reality.
I don't think we're going to see a desktop 8xxx series processor from Motorola. In 2004 we may see a RapidIO G4 which would be a significant improvement, especially if it reaches 1.8-2.0 GHz as rumoured. If they take it to 0.09 microns that would be a spectacular chip for notebooks and other places where low power / low heat is needed. RIO will allow fast NUMA architectures so we could see things like 4 or 8 way Xserves where each G4 is at least 2-4 times faster than a current 1 GHz G4, possibly more. Imagine a rack full of those things -- it would be quite the server farm. The 970 will still be the crown jewel, but that doesn't make it ideal for all uses.
<a href="http://www.motorola.com/mediacenter/bios/0,1071,281,00.html"; target="_blank">http://www.motorola.com/mediacenter/bios/0,1071,281,00.html</a>;
[quote]GREGG BARTLETT
Deputy Operations Manager, Crolles2 Project
Technology & Manufacturing
Semiconductor Products Sector, Motorola, Inc.
Gregg Bartlett was named Deputy Operations Manager for the Crolles2 Project in May 2002. Crolles2 is the name of Motorola?s partnership with STMicroelectronics and Philips for breakthrough technology development and 300mm pilot manufacturing. Gregg has responsibility for all aspects of the Crolles project for Motorola as well as a functional role within the Crolles joint organizational structure.
Prior to this, Gregg was the director of the Advanced Products Research & Development Laboratory (APRDL) in Austin, Texas. APRDL is part of DigitalDNA? Laboratories (DDL), the research and development organization of SPS.
From 1996 to 2002, he held leadership roles in technology and manufacturing partnership environments. First, he worked with Infineon for the White Oak joint venture in Richmond, and most recently with AMD for the HiPERMOS technology development agreement in Austin. Gregg has more than 18-years experience at Motorola with the last three-years in technology development roles in DDL. His focus has been process technology including the CDR1 SiGe:C process running in MOS-11, and HiP-7 currently in production in MOS-13. His experience spans technology development, fab design and construction as well as start-up and ramp.
Gregg attended Kansas State University where he received his B.S. degree in Chemical Engineering in 1983.
December 2002<hr></blockquote>
*bold is mine
what product is Motorola using HiP7 on? I wasn't aware of any.
[ 01-26-2003: Message edited by: rickag ]
[ 01-26-2003: Message edited by: rickag ]</p>
No sightings of MPC 7457 just yet, but they are expected to be using HiP7 as well.
I hope these documents haven't been posted elsewere yet.
<a href="http://e-www.motorola.com/brdata/PDFDB/docs/ALTIVECFACT.pdf"; target="_blank">This PDF</a> witch is a facts sheet of AltiVec. It mentions MPC7457 as a member of the MPC74xx-family witch includes AltiVec.
The PDF is dated Dec 5 2002 and is modified Jan 9 2003.
No new information is included, it just confirms its existence.
<a href="http://www.mdronline.com/events/dinner/"; target="_blank">This URL</a> witch is an invitation to Microprocessor Report 4th Annual Analysts' Choice Awards. The nominee in the category High Performance Embedded Processor 2002 is among others:
Motorola MPC7457
Quite a feat to be nominated for an award in a year in witch it didn't make any appearance..
If I were to speculate, I would guess that they are re-designing the G5 and when it comes out, will be the G6.
<strong>Not to change the subject but what is going to happen with Altivec and Motorola? I can't see them not making processors anymore without Altivec.
If I were to speculate, I would guess that they are re-designing the G5 and when it comes out, will be the G6.</strong><hr></blockquote>
I would guess that they aren't designing a new high-end core because it doesn't fit with their business plans. They'll leverage the existing G4 core by adding some of their system-on-a-chip functionality to it (RIO & memory controller), and by ramping its clock rate and shrinking the process its made on. Beyond that I wouldn't count on much...
Quite a feat to be nominated for an award in a year in witch it didn't make any appearance.<hr></blockquote>
I enjoyed that quote. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
It is interesting that the nominations are "recognizing excellence in technology innovation, design, and implementation."
This would imply that there are working samples of the MPC7457 out there(re: implementation). Question, if the MPC7457 is only a MPC7455 manufactured on a 0.13µm process, where does "innovation" and "design" enter into the evaluation?
Maybe more knowledgeable people than myself can surmise a guess as to the possibility there have been any significant design modifications to the MPC7455 included into the MPC7457. Bear in mind the MPC7457 is nominated in the "High-Performance
Embedded Processors" category.
Extended pipelines? Additional FP unit? Increased FSB capabilities? - naww can't be, this is Motorola and Motorola = embedded.
But then again hope springs enternal.
If you go to tech spec's on Apple's website it lists the L2 cache for all 3 processors @ " 256K on-chip L2 cache running at processor speed".
If I were a betting man, this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004.
<strong>Well, it appears the latest upgrades don't include the MPC7457.
If you go to tech spec's on Apple's website it lists the L2 cache for all 3 processors @ " 256K on-chip L2 cache running at processor speed".
If I were a betting man, this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004.
Apple doesn't have to use the 7457 in the PowerMacs when it comes out. They still need a lower power chip for portables and iMacs.
<strong>
Apple doesn't have to use the 7457 in the PowerMacs when it comes out. They still need a lower power chip for portables and iMacs.</strong><hr></blockquote>
No they don't. Let's see, Apple introduces a 1.6GHz MPC7457 w/ a 200MHz FSB in the iMac and Powerbooks - Professional users descend on Cupertino and burn it to the ground.
Or Apple introduces underclocked MPC7457's @ 1.0 - 1.2GHz w/ a 167MHz FSB in the iMac and Powerbooks and Professional users descend on Cupertino and burn it to the ground.
I think I need therapy. Now where's that Prozac.
<strong>No they don't. Let's see, Apple introduces a 1.6GHz MPC7457 w/ a 200MHz FSB in the iMac and Powerbooks - Professional users descend on Cupertino and burn it to the ground.
Or Apple introduces underclocked MPC7457's @ 1.0 - 1.2GHz w/ a 167MHz FSB in the iMac and Powerbooks and Professional users descend on Cupertino and burn it to the ground.
I think I need therapy. Now where's that Prozac.</strong><hr></blockquote>
And if Apple introduces 1.8 GHz 970-based PowerMacs at the same time as the 7457 appears in the low end? Wow, then everybody is happy.
[ 01-28-2003: Message edited by: Programmer ]</p>
<strong>Well, it appears the latest upgrades don't include the MPC7457.
No, it doesn't, but that's not all bad news: Did anyone here think the 7455 would get anywhere near 1.4GHz? I was surprised when they got it to 1.25GHz.
This, to me, is an indication that Mot's starting to beat their fabrication problems, and that's good news.
Think about it: The 7455 took us from 667MHz to 1.42GHz. Not shabby at all. How far will the '57 take us, even if the 970 ends up pushing it into the consumer line?
<strong>Think about it: The 7455 took us from 667MHz to 1.42GHz.</strong><hr></blockquote>The 7455 Apollo started with the 1Ghz machines. The 667-733 were 7450s, without SOI.
I'm also still shocked fom the display prices, those beat even the most naively positive guesses.
<strong>And if Apple introduces 1.8 GHz 970-based PowerMacs at the same time as the 7457 appears in the low end? Wow, then everybody is happy.</strong><hr></blockquote>
Yes this is true, but your using logic, I wasn't, I was trying to be humorous.
But, a lot of people were expecting the MPC7457 to be the next upgrade. Soooo, I should have added that the iMac/Powerbooks were to be updated, like say, next week with the MPC757, leaving the Powermacs to fend with the MPC7455's until the IBM 970 appears very very late 2003 or 2004.
Amorph makes very important point concerning the MPC7455 scaling(terminology?) to 1.42GHz. Pretty amazing when you think about it(re: 7 stages and all). Let's see, 0.13µm process adds 30% to the clock speed. That's 1.3 X 1.42GHz = 1.85GHz.
So, next week Apple introduces the new iMacs @ 1.85GHz w/ a 200MHz FSB and Powerbooks @ 1.2GHz w/ a 167MHz FSB. Then, profesional users descend on Cupertino and burn it to the ground.
My feeble attempt @ logic was, [quote]"Well, it appears the latest upgrades don't include the MPC7457...
... this latest upgrade means the MPC7457 to arrive midsummer, with the IBM 970(re: if it ever makes it to a Mac) to arrive very very late in 2003 or 2004."<hr></blockquote>
[ 01-28-2003: Message edited by: rickag ]</p>
<strong>The 7455 Apollo started with the 1Ghz machines. The 667-733 were 7450s, without SOI.</strong><hr></blockquote>
OK, so 867 to 1.42GHz. Still, not bad.
<strong>The 7455 Apollo started with the 1Ghz machines. The 667-733 were 7450s, without SOI.</strong><hr></blockquote>
according this link: <a href="http://www.apple-history.com/g4_quicksilver.html"; target="_blank">http://www.apple-history.com/g4_quicksilver.html</a>; you're right about the processors apple used but according to motorola the 7455 starts @ 600Mhz and tops out @ 1Ghz so this means the 7455 is more scalable than motorola says.
now, i'm currious about the processors in the imac/emac/powerbook, are these also 7455's or just 7450's?