Mslee how do you know there is no 7500?? As far as I know there is one. Did they cancel it or are you just saying this??? Apple may never use the 7500 if they switch to IBM (which I think they should do) but Moto has been working on a 7500. In time.
Well yes and no, you could write software that could take advantage of n processors automatically, but a lot of programmers take the shortcut of writing specifically for 2 processors since that is what the majority of multi-processor systems are.
For example when you have a task with a high degree of parallelism, if you take the shortcut of always spinning off 2 threads, rather than going to the extra trouble of figuring out how many processors are available and spinning off that many threads and adding the logic in each thread to properly divide the work no matter how many threads there are; then you end up with code that will benefit from 2 processors, but gain no additional benefit from more than that.</strong><hr></blockquote>
Running more than one application will use the other processors without them needing to be aware of >2. Depending on what you are doing this can be very useful.
This speed bump took us from 500 MHz to 733 MHz, a 47% speed increase.
Another 47% increaes in MHz would put the G4 at 1470 MHz, roughly 1.5 GHz.
So yes, it is entirely in line with the history of the G4 that such a large speed bump may occur.</strong><hr></blockquote>
JYD,
Did you count how L-O-N-G it took for MOT to go from 500Mhz to 733Mhz (or as you say a 47% bump)? Let me help... SIXTEEN MONTHS! If you wanna bank on MOT's past then to get another 47% pop then, yea... SIXTEEN MONTHS from this past January we'll get to 1.5Ghz... 7 Months down 9 more to go...
we will have up to a 400Mhz increase. I do not know if apple will choose to use it all up in one upgrade or not. They need growth run you know. But since apple has redesigned the case ect I would not be surprised to see a 1.4Ghz machine come august 13th. In time.
Did you count how L-O-N-G it took for MOT to go from 500Mhz to 733Mhz (or as you say a 47% bump)? Let me help... SIXTEEN MONTHS! If you wanna bank on MOT's past then to get another 47% pop then, yea... SIXTEEN MONTHS from this past January we'll get to 1.5Ghz... 7 Months down 9 more to go...
D</strong><hr></blockquote>
Those 16 months were the result of fab problems at Moto, they just couldn't figure out how to scale the damn G4.
IF they can move the G4 to a 130 nm process, then there's no reason why we couldn't see another 50% increase in MHz.
Also, it's important to note that Apple is farther behind NOW than they were when they hit the 500 MHz wall. Powermac sales are trickling, Apple's back is up against the wall. The current situation calls for a dramatic increase in performance.
And Steve Jobs DID say that Apple would be closing the MHz gap. Clearly that has yet to happen, and I hesitate to call Jobs a flat-out liar. He must have based that comment on something, and my guess is that as usual, Moto dragged their ass on it and is late. But late is better than never.
<strong>A. I can't remember a time when processor speeds jumped by more than 100 MHz or so. What gives anyone the idea we will suddenly jump by 2, 3 or even 400 MHz? Use your head ... people who say this are basically "buying their own bullsh*t" because it's what they want, not because there's a logical reason to believe it. </strong><hr></blockquote>
Nonesense. The 500-733 MHz jump about a year ago represented a 47% increase in MHz.
The same increase in mHz now would put the G4 at around 1500 MHz.
Remember that the proportional increase in MHz is more telling than the actual number.
For example, back when CPUs were at 20 Hz, a jump to 40 Hz represented a DOUBLING in speed! Yet on a 1 GHz G4, a 20 Hz speed bump would be laughable. Take that and make it a 100 % increase, and you go from 1 GHz to 2 GHz.
"IF they can move the G4 to a 130 nm process, then there's no reason why we couldn't see another 50% increase in MHz.
Also, it's important to note that Apple is farther behind NOW than they were when they hit the 500 MHz wall. Powermac sales are trickling, Apple's back is up against the wall. The current situation calls for a dramatic increase in performance.
And Steve Jobs DID say that Apple would be closing the MHz gap. Clearly that has yet to happen, and I hesitate to call Jobs a flat-out liar. He must have based that comment on something, and my guess is that as usual, Moto dragged their ass on it and is late. But late is better than never. "
Agreed JD.
Steve Jobs did say that. What timeframe he based that statement on...we can only guess...
For a move to .13, I don't see 1.4 being unreasonable.
This speed bump took us from 500 MHz to 733 MHz, a 47% speed increase.
Another 47% increaes in MHz would put the G4 at 1470 MHz, roughly 1.5 GHz.
</strong><hr></blockquote>
but that Mhz increase came partly because of the move from 4 to 7 pipeline stages, and Mhz for Mhz the 7 pipeline version is a bit slower ( in most cases).. The 7400 and 7450 are quite different. BTW, this was the first time in PPC history, we saw an increase in the number of pipeline stages.
You mean the PPC 601, etc all had 4 pipeline stages?</strong><hr></blockquote>
yes... the 601 is not very different from 603.. The 603 got different cache suppor and lacks some of 601s extra (unoficial/not supported.. Power?) instructions. Thats about it.
The sad fact is that not much beyond cache-support changed from 601 to 750. So, I suspect current R&D cost is much higher than it used to be.
Soooooo, basically the design of the G4 hasn't changed that much since the MPC 7450/MPC 7451 when 1-2 MB L3 cache was added(How complicated would this have been??)
Can anyone tell me if the same engineers would be significantly involved in all the
1.development of SOI
2.increasing the pipeline/adding L3 cache??
3.developing a smaller process.
Just a layman's guess, but I would think not. Different set of core engineers involved in the different facets of development.
Exactly what have the engineers responsible for extending pipelines and adding L3 cache, etc.(re: architectural design changes) been doing all this time???
It DO SEEM like a long time to go without any major redesign.
umm, only thing I'm unsure of in your li'l overvju there is the original 7400, which I think was on a .22µ process, not .20µ
anyways, unless you count the move to SOI a major change, there hasn't been one in 20 months, and I'd say I'm quite optimistic regarding the .13µ+SOI tech.
Anyhoooo, looks like the process guys and the metalurgists have been busy, what have the design guys been doing <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
jeezz, I just had to add, 20 months holy cow 20 months, I can't believe it 20 months. Please, some one with technological knowledge, explain this, there has to be a rational explanation I'm not seeing here. Got to be, great googly moogly, I'm depressed.
Arghhhh, just thought of the 133MHz bus too
I'm going home to sulk, whoe(edit I meant WOE) is me.
[ 08-01-2002: Message edited by: rickag ]
[ 08-01-2002: Message edited by: rickag ]
what did I think, I was stopping a horse <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
Soooooo, basically the design of the G4 hasn't changed that much since the MPC 7450/MPC 7451 when 1-2 MB L3 cache was added(How complicated would this have been??)</strong>
The 7400 to 7450 transition was the redesign. It was much more of a design improvement than from 750 to 7400. Compared to the 7400, the 7450 has 3 more pipeline stages, 2 more simple integer units, 36 bit memory addressing, on-die 256 bit wide L2 cache, DDR backside cache, 3-way instruction dispatch/completion, and all the buffer increases to support the increase instruction throughput. I would not call the 7400 and the 7450 the same chip, much in the way you would not call the 750 and the 7400 the same chip.
The only thing they didn't do that needed to be done was increase the processor bus bandwidth and adding another FPU.
<strong>Can anyone tell me if the same engineers would be significantly involved in all the
1.development of SOI
2.increasing the pipeline/adding L3 cache??
3.developing a smaller process.
Just a layman's guess, but I would think not. Different set of core engineers involved in the different facets of development.</strong>
Yes.
I have a sneaking suspicion that Intel took the Motorola Cu interconnect and SOI guys away from Motorola, so the HiP 6 SOI people might not be there for the HiP 7 ramp, but this is baseless. Only data I have is Moto sued Intel for stealing their engineers, I don't know for which ones.
The CPU design guys are problem concentrating on e500 core chips right now, so it's not surprising the 7450 based chip isn't progress as much as it should.
<strong>It DO SEEM like a long time to go without any major redesign.</strong>
5 years to 6 years for a CPU of the 7450's complexity. It's only been 2 years. After the HiP 7 fab is on the tail end of the tech, the 7450 will wane, that's if, a big if, Motorola is playing the same host processor game. It looks like they are moving in a different direction though with the low power embedded 8540 and 8560 SOC chips.
the 750 was on .22µ though, I mixed them up I suppose
The design guys should have been working on the 7500, overtime and hard. But have they? I doubt it... drinking coffee is less stressing than designing a new chip
Comments
[ 07-31-2002: Message edited by: Algol ]</p>
<strong>
Well yes and no, you could write software that could take advantage of n processors automatically, but a lot of programmers take the shortcut of writing specifically for 2 processors since that is what the majority of multi-processor systems are.
For example when you have a task with a high degree of parallelism, if you take the shortcut of always spinning off 2 threads, rather than going to the extra trouble of figuring out how many processors are available and spinning off that many threads and adding the logic in each thread to properly divide the work no matter how many threads there are; then you end up with code that will benefit from 2 processors, but gain no additional benefit from more than that.</strong><hr></blockquote>
Running more than one application will use the other processors without them needing to be aware of >2. Depending on what you are doing this can be very useful.
Sez who?
Anybody got proof? We hear lots of the 7470.
Is that on the road map?
Just wondering...
Lemon BOn BOn
<strong>
This speed bump took us from 500 MHz to 733 MHz, a 47% speed increase.
Another 47% increaes in MHz would put the G4 at 1470 MHz, roughly 1.5 GHz.
So yes, it is entirely in line with the history of the G4 that such a large speed bump may occur.</strong><hr></blockquote>
JYD,
Did you count how L-O-N-G it took for MOT to go from 500Mhz to 733Mhz (or as you say a 47% bump)? Let me help... SIXTEEN MONTHS! If you wanna bank on MOT's past then to get another 47% pop then, yea... SIXTEEN MONTHS from this past January we'll get to 1.5Ghz... 7 Months down 9 more to go...
D
Motorola redesigned the G4 with the 7450 chip, and go a speed increase of almost 50%.
Another chip redesign (and it is over 18 months since the last one), and you could very well see the same speed increase again.
Barto
[ 07-31-2002: Message edited by: Barto ]</p>
<strong>
JYD,
Did you count how L-O-N-G it took for MOT to go from 500Mhz to 733Mhz (or as you say a 47% bump)? Let me help... SIXTEEN MONTHS! If you wanna bank on MOT's past then to get another 47% pop then, yea... SIXTEEN MONTHS from this past January we'll get to 1.5Ghz... 7 Months down 9 more to go...
D</strong><hr></blockquote>
Those 16 months were the result of fab problems at Moto, they just couldn't figure out how to scale the damn G4.
IF they can move the G4 to a 130 nm process, then there's no reason why we couldn't see another 50% increase in MHz.
Also, it's important to note that Apple is farther behind NOW than they were when they hit the 500 MHz wall. Powermac sales are trickling, Apple's back is up against the wall. The current situation calls for a dramatic increase in performance.
And Steve Jobs DID say that Apple would be closing the MHz gap. Clearly that has yet to happen, and I hesitate to call Jobs a flat-out liar. He must have based that comment on something, and my guess is that as usual, Moto dragged their ass on it and is late. But late is better than never.
<strong>A. I can't remember a time when processor speeds jumped by more than 100 MHz or so. What gives anyone the idea we will suddenly jump by 2, 3 or even 400 MHz? Use your head ... people who say this are basically "buying their own bullsh*t" because it's what they want, not because there's a logical reason to believe it. </strong><hr></blockquote>
Nonesense. The 500-733 MHz jump about a year ago represented a 47% increase in MHz.
The same increase in mHz now would put the G4 at around 1500 MHz.
Remember that the proportional increase in MHz is more telling than the actual number.
For example, back when CPUs were at 20 Hz, a jump to 40 Hz represented a DOUBLING in speed! Yet on a 1 GHz G4, a 20 Hz speed bump would be laughable. Take that and make it a 100 % increase, and you go from 1 GHz to 2 GHz.
I hope his optimisim in Mr. Jobs is reasonable, as I happen to share it.
Also, it's important to note that Apple is farther behind NOW than they were when they hit the 500 MHz wall. Powermac sales are trickling, Apple's back is up against the wall. The current situation calls for a dramatic increase in performance.
And Steve Jobs DID say that Apple would be closing the MHz gap. Clearly that has yet to happen, and I hesitate to call Jobs a flat-out liar. He must have based that comment on something, and my guess is that as usual, Moto dragged their ass on it and is late. But late is better than never. "
Agreed JD.
Steve Jobs did say that. What timeframe he based that statement on...we can only guess...
For a move to .13, I don't see 1.4 being unreasonable.
Lemon Bon Bon
Lemon Bon Bon
<strong>
This speed bump took us from 500 MHz to 733 MHz, a 47% speed increase.
Another 47% increaes in MHz would put the G4 at 1470 MHz, roughly 1.5 GHz.
</strong><hr></blockquote>
but that Mhz increase came partly because of the move from 4 to 7 pipeline stages, and Mhz for Mhz the 7 pipeline version is a bit slower ( in most cases).. The 7400 and 7450 are quite different. BTW, this was the first time in PPC history, we saw an increase in the number of pipeline stages.
[ 08-01-2002: Message edited by: blabla ]</p>
<hr></blockquote>
You mean the PPC 601, etc all had 4 pipeline stages?
<strong>
You mean the PPC 601, etc all had 4 pipeline stages?</strong><hr></blockquote>
yes... the 601 is not very different from 603.. The 603 got different cache suppor and lacks some of 601s extra (unoficial/not supported.. Power?) instructions. Thats about it.
The sad fact is that not much beyond cache-support changed from 601 to 750. So, I suspect current R&D cost is much higher than it used to be.
lets hope so at this rate we'll be at 2 ghz 8 montsh from now
doubt it though
i just hope for a lowend tower of 1ghz with ddr
MPC 7400 4 stages 0.20µ
MPC 7410 4 stages 0.18µ
MPC 7441 7 stages 0.18µ
MPC 7450 7 stages 0.18µ 1-2MB L3 cache
MPC 7451 7 stages 0.18µ 1-2MB L3 cache
MPC 7445 7 stages 0.18µ SOI
MPC 7455 7 stages 0.18µ SOI 1-2MB L3 cache
Soooooo, basically the design of the G4 hasn't changed that much since the MPC 7450/MPC 7451 when 1-2 MB L3 cache was added(How complicated would this have been??)
Can anyone tell me if the same engineers would be significantly involved in all the
1.development of SOI
2.increasing the pipeline/adding L3 cache??
3.developing a smaller process.
Just a layman's guess, but I would think not. Different set of core engineers involved in the different facets of development.
Exactly what have the engineers responsible for extending pipelines and adding L3 cache, etc.(re: architectural design changes) been doing all this time???
It DO SEEM like a long time to go without any major redesign.
Just asking
anyways, unless you count the move to SOI a major change, there hasn't been one in 20 months, and I'd say I'm quite optimistic regarding the .13µ+SOI tech.
Thanks for the information, I got the information from <a href="http://e-www.motorola.com/webapp/sps/site/taxonomy.jsp?nodeId=01M0ylsbTdG" target="_blank">here</a>, I need to not trust Motorola.
Anyhoooo, looks like the process guys and the metalurgists have been busy, what have the design guys been doing <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
jeezz, I just had to add, 20 months holy cow 20 months, I can't believe it 20 months. Please, some one with technological knowledge, explain this, there has to be a rational explanation I'm not seeing here. Got to be, great googly moogly, I'm depressed.
Arghhhh, just thought of the 133MHz bus too
I'm going home to sulk, whoe(edit I meant WOE) is me.
[ 08-01-2002: Message edited by: rickag ]
[ 08-01-2002: Message edited by: rickag ]
what did I think, I was stopping a horse <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
[ 08-01-2002: Message edited by: rickag ]</p>
Lets see if I got this right
MPC 7400 4 stages 0.20µ [edit: 0.5, 1, 2 MB L2 cache]
MPC 7410 4 stages 0.18µ [edit: 0.5, 1, 2 MB L2 cache]
MPC 7441 7 stages 0.18µ
MPC 7450 7 stages 0.18µ 1-2MB L3 cache
MPC 7451 7 stages 0.18µ 1-2MB L3 cache
MPC 7445 7 stages 0.18µ SOI
MPC 7455 7 stages 0.18µ SOI 1-2MB L3 cache
Soooooo, basically the design of the G4 hasn't changed that much since the MPC 7450/MPC 7451 when 1-2 MB L3 cache was added(How complicated would this have been??)</strong>
The 7400 to 7450 transition was the redesign. It was much more of a design improvement than from 750 to 7400. Compared to the 7400, the 7450 has 3 more pipeline stages, 2 more simple integer units, 36 bit memory addressing, on-die 256 bit wide L2 cache, DDR backside cache, 3-way instruction dispatch/completion, and all the buffer increases to support the increase instruction throughput. I would not call the 7400 and the 7450 the same chip, much in the way you would not call the 750 and the 7400 the same chip.
The only thing they didn't do that needed to be done was increase the processor bus bandwidth and adding another FPU.
<strong>Can anyone tell me if the same engineers would be significantly involved in all the
1.development of SOI
2.increasing the pipeline/adding L3 cache??
3.developing a smaller process.
Just a layman's guess, but I would think not. Different set of core engineers involved in the different facets of development.</strong>
Yes.
I have a sneaking suspicion that Intel took the Motorola Cu interconnect and SOI guys away from Motorola, so the HiP 6 SOI people might not be there for the HiP 7 ramp, but this is baseless. Only data I have is Moto sued Intel for stealing their engineers, I don't know for which ones.
The CPU design guys are problem concentrating on e500 core chips right now, so it's not surprising the 7450 based chip isn't progress as much as it should.
<strong>It DO SEEM like a long time to go without any major redesign.</strong>
5 years to 6 years for a CPU of the 7450's complexity. It's only been 2 years. After the HiP 7 fab is on the tail end of the tech, the 7450 will wane, that's if, a big if, Motorola is playing the same host processor game. It looks like they are moving in a different direction though with the low power embedded 8540 and 8560 SOC chips.
the 750 was on .22µ though, I mixed them up I suppose
The design guys should have been working on the 7500, overtime and hard. But have they? I doubt it... drinking coffee is less stressing than designing a new chip