Hehe... now, is the bet going to be which is going to be first, Motorola shipping HiP 7 processors or Motorola selling their semiconductor business?
During Motorola's 2001 Q1 conference call, the higher ups said HiP 7 was qualified and about ready to run. If they don't ship the 8540/8560/7457 by Summer, that'll be a whole 2 years in which Moto did nothing with its 130 nm fab. It's amazing that they are still alive. If IBM bought into AltiVec 2 years ago, they would have killed Moto's embedded business by now.
Btw, I think Apple only uses the 7450/7451/7455 in its computers. They have never used the cacheless versions of the chips. Shipped 7450/7451/7455 machines w/o cache, yes, but never the 7440/7441/7445 chips. The ones going into Powerbooks are probably low watt variants of the 7450/7451/7455. And 667 MHz to 1.42 GHz is pretty good. Silicon-on-Insulator saved Apple, and Moto, from ludicrous embarrassment, as opposed to just embarrassing.
That would be more like it. More aggressive. More in keeping with what we're seeing so far this year.
'G5'/970 for 'POWER'Macs. This fall. By then, the range is going to need it. I'm a former Powermac user who is waiting for the 'power' to be redefined in much the same way that the G4 promised to do at its launch.
It's kinda obvious. An iMac range with 1 - 1.4 gig G4 come fall will look okay.
That is, if Apple wants 2003 to be the great leveller it looks like.
Great new software.
Mind blowing Monitors and amazing price cuts.
Price crash on 'power'Macs.
Deep iMac2 cuts pending.
Amazing redefined laptop range from end to end...trouncing many a Wintel laptop along the way?
<strong>Hehe... now, is the bet going to be which is going to be first, Motorola shipping HiP 7 processors or Motorola selling their semiconductor business?
During Motorola's 2001 Q1 conference call, the higher ups said HiP 7 was qualified and about ready to run. If they don't ship the 8540/8560/7457 by Summer, that'll be a whole 2 years in which Moto did nothing with its 130 nm fab. It's amazing that they are still alive. If IBM bought into AltiVec 2 years ago, they would have killed Moto's embedded business by now.
Btw, I think Apple only uses the 7450/7451/7455 in its computers. They have never used the cacheless versions of the chips. Shipped 7450/7451/7455 machines w/o cache, yes, but never the 7440/7441/7445 chips. The ones going into Powerbooks are probably low watt variants of the 7450/7451/7455. And 667 MHz to 1.42 GHz is pretty good. Silicon-on-Insulator saved Apple, and Moto, from ludicrous embarrassment, as opposed to just embarrassing.</strong><hr></blockquote>
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" /> 7440/7441/7445 are not cacheless, they have l2 cache, so rickag is right about his comment and i have still no answer
<strong>7457. Destined for fall iMacs and Powerbooks.
That would be more like it. More aggressive. More in keeping with what we're seeing so far this year.
'G5'/970 for 'POWER'Macs. This fall. By then, the range is going to need it. I'm a former Powermac user who is waiting for the 'power' to be redefined in much the same way that the G4 promised to do at its launch.
[]
It's still only January.
2003 lookin' finger lickin' good.
Lemon Bon Bon </strong><hr></blockquote>
mr jobs likes to have options. so another possibility is 7457 this fall for the powermac ranging from 1.4Ghz-1.8ghz, if ibm got problems ramping up ppc970 production. still a linger fickin' good year.
L3 cache comes and goes in different amounts according to the whims of Apple. That's why you can't look at that and conclusively tell what chip it is.
The L2 cache for the Apollo 7455 is 256K at processor speed and its on the chip, so probably won't change from machine to machine like the L3. The PowerMacs, PowerBooks, iMacs, and eMacs all have the same L2 cache, even apparently these new ones. The 7457 will reportedly have 512K of L2 cache, so it doesn't look like this is the 7457.
It's kinda sad that the G3 in the iBook is more advanced in several ways than the G4 in the PowerMacs. It already has the 200Mhz bus capability, the 512K L2 cache, and the .13 fab process. Too bad they're still selling them at 700 Mhz.
I defer to THT. He is much more knowledgable than I.
I just ass+u+me(d) the versions on Motorola's website that did not list L3 cache were the cpu's Apple used in those computers that did not have L 3 cache. It did not occur to me that Motorola would produce the MPC7450, MPC7451 and MPC7455 processors sans L3 cache. After all the spec's for the MPC7400, MPC7441 and MPC7445 looked identical other than L3 cache
Sorry I couldn't help.
The MPC7441, MPC7445, MPC7451 and MPC7455 all have 64K L1 cache and 256K L2 internal cache. Only the MPC7451 and MPC7455 list L3 cache up to 2MB external(backside cache).
Hence my confusion.
If I never hear the word sandbagging again I'll die happy.
I did not occur to me that Motorola would produce the MPC7450, MPC7451 and MPC7455 processors sans L 3 cache.</strong>
That's the only way Motorola produces them!
The L3 cache is backside cache. "Backside" cache meaning that there is an exclusive, direct bus from CPU to off-chip SRAM cache (as opposed to in the old days when it was on the much slower FSB). The chips themselves ship in a ball grid array package, 25x25 mm in size. Essentially, the 106 sq mm 7455 CPU is set in the middle of a 625 sq mm plate of ceramic. Wires are threaded from the CPU to solder ball points.
I do not know for sure, but Apple probably has to take the chips and solder those ball points onto the CPU module printed circuit board. The backside cache is then soldered on as well. The CPU module is attached to the motherboard through the "Sawtooth" socket they've been using for a long time. I highly doubt Motorola comes into play except for tech support.
It's a lot easier on the Intel side since they use pin grid array packaging with standard sockets. Even when they used slots, Intel/AMD produced the CPU/backside cache modules themselves, so vendors only had to plug them into standard slots.
The problem with the 7440/7441/7445 is that they come in a different package than the 7450/51/55. There's really no use in buying a different chip with different packaging when Motorola produces low voltage (lower watt) variants of the 7450/51/55.
THT takes rickag to school again. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Thanks for the information. Now if I could just remember, Motorola used a different criteria for measuring die size so the 0.18µm process was actually a 0.15µm process so the new method means the 0.10µm process is actually a 0.13µm process except in the case of the frazipan, then all bets are off.
I believe I'm exhibiting the early signs of Alzheimer's. <img src="confused.gif" border="0">
<strong>THT takes rickag to school again. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Thanks for the information. Now if I could just remember, Motorola used a different criteria for measuring die size so the 0.18µm process was actually a 0.15µm process so the new method means the 0.10µm process is actually a 0.13µm process except in the case of the frazipan, then all bets are off.
I believe I'm exhibiting the early signs of Alzheimer's. </strong><hr></blockquote>
Also a new generation of processors coming next year, breaking 2 GHz.
-----------
again on 12/10:
Let's just say I got some info from their vendor.
1.4 is coming in January and next gen (G5) might be ready for Q3 (more likely in Q4).
They always try to meet the MacWorld dates (Jan, Aug) to introduce new products.
--------------
1/28/03:
See, I told you so.
Now, the 2 GHz will come later this year (IBM) and end of 2003 or early 2004 (MOT).
-----------------
1/28/03
I don't know if Apple has decided to use IBM proc for their high-end systems.
Seem like IBM is ahead of MOT 3-6 months in the G5 generation, but MOT will catch up quickly. They will have their 2Ghz+ procs in production in about 9-12 months.
----------------
Ok, so he's not exactly clear but he did say 1.4 towers in Janurary which was right on.
There seems to be an assumption that the 7457 will arrive at a higher clock rate than the 7455. This isn't necessarily true -- it may take time to refine the 7457 production to get the speeds up to the quoted +30%. In the short term this is perfect for the iMac/PowerBook line, low power and still slower than the PowerMacs. When the PowerMac goes to the 970 in Q3/Q4 the 7457 will probably be up to speed and reaching beyond the 1.42 GHz mark.
Originally, I thought that today's release was old tech and my info of a new G4 was incorrect. However, it seems the new pMacs have a smaller heatsink, new fan and new tech based on the Xserve.
[quote] "Cutting-edge Xserve-based system architecture gives the Power Mac G4 a tremendous performance boost.
The new Power Mac G4 systems leverage the highly acclaimed Apple Xserve architecture to deliver the ultimate in computing power. Designed with a clear objective in mind \\ dramatically increased throughput \\ these new desktop workstations combine one or two PowerPC G4 processors, Double Data Rate (DDR) main memory, dedicated L3 cache and integrated I/O to create an extremely responsive system."
<hr></blockquote>
So these may be what I was told about, not the same old G4 clocked up to 1.4ghz (so it's techincally a new design), which I think holds well for the 970 in the fall.
<strong>Originally, I thought that today's release was old tech and my info of a new G4 was incorrect. However, it seems the new pMacs have a smaller heatsink, new fan and new tech based on the Xserve.
So these may be what I was told about, not the same old G4 clocked up to 1.4ghz (so it's techincally a new design), which I think holds well for the 970 in the fall.</strong><hr></blockquote>
Weren't the first MD PowerMac's based off Xserve technology? I thought they were.
Motorola's desktop processor division is dying a slow, painful death, and the death knoll should speed up the 970 release. I would be very surprised if Apple had ANY product with a Motorola chip inside come 2004. They just need to go away. They change their collective minds more often than a fat chick in a Pizza Hut (sorry, in advance if that offends anyone <img src="embarrassed.gif" border="0"> )
<strong>Weren't the first MD PowerMac's based off Xserve technology?</strong><hr></blockquote>Yes they were. The Xserve technology Apple's talking about is the ad hoc DDR-solution I think.. and perhaps the blue mobos
<strong>Way back when Don Carlos (another board) said:
12/10/02
1.4 GHz Mac coming early next year
Motorola is reading a new PowerPC proc.
Also a new generation of processors coming next year, breaking 2 GHz.
-----------
again on 12/10:
Let's just say I got some info from their vendor.
1.4 is coming in January and next gen (G5) might be ready for Q3 (more likely in Q4).
They always try to meet the MacWorld dates (Jan, Aug) to introduce new products.
--------------
1/28/03:
See, I told you so.
Now, the 2 GHz will come later this year (IBM) and end of 2003 or early 2004 (MOT).
-----------------
1/28/03
I don't know if Apple has decided to use IBM proc for their high-end systems.
Seem like IBM is ahead of MOT 3-6 months in the G5 generation, but MOT will catch up quickly. They will have their 2Ghz+ procs in production in about 9-12 months.
----------------
Ok, so he's not exactly clear but he did say 1.4 towers in Janurary which was right on.
Interesting.
Chas</strong><hr></blockquote>
His followup to my questions last night:
>>are you saying that the IBM will go in the desktops and the MOTs in laptops/consumers<<
Yes.
>>Also, is there going to be a big performance jump going to G5 regardless of the MHz rating?<<
More cache, faster bus with new topology, 32 & 64 bit cores, new pipeline ...
It might be a quantum leap for Apple computers (2x-3x in performance initially!).
Comments
<strong>
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?</strong><hr></blockquote>
look at the level 3 cache for the models, it should answer the questions.
<strong>
look at the level 3 cache for the models, it should answer the questions.</strong><hr></blockquote>
okay, 7445/7455 then?
<strong>look at the level 3 cache for the models, it should answer the questions.</strong><hr></blockquote>level 2
During Motorola's 2001 Q1 conference call, the higher ups said HiP 7 was qualified and about ready to run. If they don't ship the 8540/8560/7457 by Summer, that'll be a whole 2 years in which Moto did nothing with its 130 nm fab. It's amazing that they are still alive. If IBM bought into AltiVec 2 years ago, they would have killed Moto's embedded business by now.
Btw, I think Apple only uses the 7450/7451/7455 in its computers. They have never used the cacheless versions of the chips. Shipped 7450/7451/7455 machines w/o cache, yes, but never the 7440/7441/7445 chips. The ones going into Powerbooks are probably low watt variants of the 7450/7451/7455. And 667 MHz to 1.42 GHz is pretty good. Silicon-on-Insulator saved Apple, and Moto, from ludicrous embarrassment, as opposed to just embarrassing.
That would be more like it. More aggressive. More in keeping with what we're seeing so far this year.
'G5'/970 for 'POWER'Macs. This fall. By then, the range is going to need it. I'm a former Powermac user who is waiting for the 'power' to be redefined in much the same way that the G4 promised to do at its launch.
It's kinda obvious. An iMac range with 1 - 1.4 gig G4 come fall will look okay.
That is, if Apple wants 2003 to be the great leveller it looks like.
Great new software.
Mind blowing Monitors and amazing price cuts.
Price crash on 'power'Macs.
Deep iMac2 cuts pending.
Amazing redefined laptop range from end to end...trouncing many a Wintel laptop along the way?
It's still only January.
2003 lookin' finger lickin' good.
Lemon Bon Bon
<strong>Hehe... now, is the bet going to be which is going to be first, Motorola shipping HiP 7 processors or Motorola selling their semiconductor business?
During Motorola's 2001 Q1 conference call, the higher ups said HiP 7 was qualified and about ready to run. If they don't ship the 8540/8560/7457 by Summer, that'll be a whole 2 years in which Moto did nothing with its 130 nm fab. It's amazing that they are still alive. If IBM bought into AltiVec 2 years ago, they would have killed Moto's embedded business by now.
Btw, I think Apple only uses the 7450/7451/7455 in its computers. They have never used the cacheless versions of the chips. Shipped 7450/7451/7455 machines w/o cache, yes, but never the 7440/7441/7445 chips. The ones going into Powerbooks are probably low watt variants of the 7450/7451/7455. And 667 MHz to 1.42 GHz is pretty good. Silicon-on-Insulator saved Apple, and Moto, from ludicrous embarrassment, as opposed to just embarrassing.</strong><hr></blockquote>
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" /> 7440/7441/7445 are not cacheless, they have l2 cache, so rickag is right about his comment and i have still no answer
<strong>7457. Destined for fall iMacs and Powerbooks.
That would be more like it. More aggressive. More in keeping with what we're seeing so far this year.
'G5'/970 for 'POWER'Macs. This fall. By then, the range is going to need it. I'm a former Powermac user who is waiting for the 'power' to be redefined in much the same way that the G4 promised to do at its launch.
[]
It's still only January.
2003 lookin' finger lickin' good.
Lemon Bon Bon
mr jobs likes to have options. so another possibility is 7457 this fall for the powermac ranging from 1.4Ghz-1.8ghz, if ibm got problems ramping up ppc970 production. still a linger fickin' good year.
The L2 cache for the Apollo 7455 is 256K at processor speed and its on the chip, so probably won't change from machine to machine like the L3. The PowerMacs, PowerBooks, iMacs, and eMacs all have the same L2 cache, even apparently these new ones. The 7457 will reportedly have 512K of L2 cache, so it doesn't look like this is the 7457.
It's kinda sad that the G3 in the iBook is more advanced in several ways than the G4 in the PowerMacs. It already has the 200Mhz bus capability, the 512K L2 cache, and the .13 fab process. Too bad they're still selling them at 700 Mhz.
<img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
I just ass+u+me(d) the versions on Motorola's website that did not list L3 cache were the cpu's Apple used in those computers that did not have L 3 cache. It did not occur to me that Motorola would produce the MPC7450, MPC7451 and MPC7455 processors sans L3 cache. After all the spec's for the MPC7400, MPC7441 and MPC7445 looked identical other than L3 cache
Sorry I couldn't help.
The MPC7441, MPC7445, MPC7451 and MPC7455 all have 64K L1 cache and 256K L2 internal cache. Only the MPC7451 and MPC7455 list L3 cache up to 2MB external(backside cache).
Hence my confusion.
If I never hear the word sandbagging again I'll die happy.
[ 01-28-2003: Message edited by: rickag ]</p>
I did not occur to me that Motorola would produce the MPC7450, MPC7451 and MPC7455 processors sans L 3 cache.</strong>
That's the only way Motorola produces them!
The L3 cache is backside cache. "Backside" cache meaning that there is an exclusive, direct bus from CPU to off-chip SRAM cache (as opposed to in the old days when it was on the much slower FSB). The chips themselves ship in a ball grid array package, 25x25 mm in size. Essentially, the 106 sq mm 7455 CPU is set in the middle of a 625 sq mm plate of ceramic. Wires are threaded from the CPU to solder ball points.
I do not know for sure, but Apple probably has to take the chips and solder those ball points onto the CPU module printed circuit board. The backside cache is then soldered on as well. The CPU module is attached to the motherboard through the "Sawtooth" socket they've been using for a long time. I highly doubt Motorola comes into play except for tech support.
It's a lot easier on the Intel side since they use pin grid array packaging with standard sockets. Even when they used slots, Intel/AMD produced the CPU/backside cache modules themselves, so vendors only had to plug them into standard slots.
The problem with the 7440/7441/7445 is that they come in a different package than the 7450/51/55. There's really no use in buying a different chip with different packaging when Motorola produces low voltage (lower watt) variants of the 7450/51/55.
[ 01-28-2003: Message edited by: THT ]</p>
Thanks for the information. Now if I could just remember, Motorola used a different criteria for measuring die size so the 0.18µm process was actually a 0.15µm process so the new method means the 0.10µm process is actually a 0.13µm process except in the case of the frazipan, then all bets are off.
I believe I'm exhibiting the early signs of Alzheimer's. <img src="confused.gif" border="0">
<strong>THT takes rickag to school again. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Thanks for the information. Now if I could just remember, Motorola used a different criteria for measuring die size so the 0.18µm process was actually a 0.15µm process so the new method means the 0.10µm process is actually a 0.13µm process except in the case of the frazipan, then all bets are off.
I believe I'm exhibiting the early signs of Alzheimer's.
...early signs :cool:
12/10/02
1.4 GHz Mac coming early next year
Motorola is reading a new PowerPC proc.
Also a new generation of processors coming next year, breaking 2 GHz.
-----------
again on 12/10:
Let's just say I got some info from their vendor.
1.4 is coming in January and next gen (G5) might be ready for Q3 (more likely in Q4).
They always try to meet the MacWorld dates (Jan, Aug) to introduce new products.
--------------
1/28/03:
See, I told you so.
Now, the 2 GHz will come later this year (IBM) and end of 2003 or early 2004 (MOT).
-----------------
1/28/03
I don't know if Apple has decided to use IBM proc for their high-end systems.
Seem like IBM is ahead of MOT 3-6 months in the G5 generation, but MOT will catch up quickly. They will have their 2Ghz+ procs in production in about 9-12 months.
----------------
Ok, so he's not exactly clear but he did say 1.4 towers in Janurary which was right on.
Interesting.
Chas
<strong>...early signs :cool: </strong><hr></blockquote>
Thweeeeeeeett!!! referee throws flag. Unsportmanslike penalty assessed to Bigc for unnecessary piling on.
[quote] "Cutting-edge Xserve-based system architecture gives the Power Mac G4 a tremendous performance boost.
The new Power Mac G4 systems leverage the highly acclaimed Apple Xserve architecture to deliver the ultimate in computing power. Designed with a clear objective in mind \\ dramatically increased throughput \\ these new desktop workstations combine one or two PowerPC G4 processors, Double Data Rate (DDR) main memory, dedicated L3 cache and integrated I/O to create an extremely responsive system."
<hr></blockquote>
So these may be what I was told about, not the same old G4 clocked up to 1.4ghz (so it's techincally a new design), which I think holds well for the 970 in the fall.
<strong>Originally, I thought that today's release was old tech and my info of a new G4 was incorrect. However, it seems the new pMacs have a smaller heatsink, new fan and new tech based on the Xserve.
So these may be what I was told about, not the same old G4 clocked up to 1.4ghz (so it's techincally a new design), which I think holds well for the 970 in the fall.</strong><hr></blockquote>
Weren't the first MD PowerMac's based off Xserve technology? I thought they were.
Motorola's desktop processor division is dying a slow, painful death, and the death knoll should speed up the 970 release. I would be very surprised if Apple had ANY product with a Motorola chip inside come 2004. They just need to go away. They change their collective minds more often than a fat chick in a Pizza Hut (sorry, in advance if that offends anyone <img src="embarrassed.gif" border="0"> )
[ 01-29-2003: Message edited by: Rhumgod ]</p>
<strong>Weren't the first MD PowerMac's based off Xserve technology?</strong><hr></blockquote>Yes they were. The Xserve technology Apple's talking about is the ad hoc DDR-solution I think.. and perhaps the blue mobos
<strong>Way back when Don Carlos (another board) said:
12/10/02
1.4 GHz Mac coming early next year
Motorola is reading a new PowerPC proc.
Also a new generation of processors coming next year, breaking 2 GHz.
-----------
again on 12/10:
Let's just say I got some info from their vendor.
1.4 is coming in January and next gen (G5) might be ready for Q3 (more likely in Q4).
They always try to meet the MacWorld dates (Jan, Aug) to introduce new products.
--------------
1/28/03:
See, I told you so.
Now, the 2 GHz will come later this year (IBM) and end of 2003 or early 2004 (MOT).
-----------------
1/28/03
I don't know if Apple has decided to use IBM proc for their high-end systems.
Seem like IBM is ahead of MOT 3-6 months in the G5 generation, but MOT will catch up quickly. They will have their 2Ghz+ procs in production in about 9-12 months.
----------------
Ok, so he's not exactly clear but he did say 1.4 towers in Janurary which was right on.
Interesting.
Chas</strong><hr></blockquote>
His followup to my questions last night:
>>are you saying that the IBM will go in the desktops and the MOTs in laptops/consumers<<
Yes.
>>Also, is there going to be a big performance jump going to G5 regardless of the MHz rating?<<
More cache, faster bus with new topology, 32 & 64 bit cores, new pipeline ...
It might be a quantum leap for Apple computers (2x-3x in performance initially!).
Chas