User Name
Password

Register Members List Calendar Search FAQ Posting Guidelines
Triple-Core
Thread Tools
mooty
Senior Member
 
Join Date: Jan 2005
Location: London, UK
 
2005-05-13, 03:51

Now that Microsoft are supposedly sticking a 3.2ghz triple-core IBM processor in their new Xbox 360

http://features.teamxbox.com/xbox/11...-Dissected/p2/

Can we expect Apple to be getting the same for their processors? I know I know, an xbox isnt a Powermac, and I dont know much about processors, but I'd like to know thoughts

Last edited by mooty : 2005-05-13 at 03:52. Reason: spelling
  quote
Gizzer
Senior Member
 
Join Date: Jul 2004
Location: Hampshire (the original one)
 
2005-05-13, 05:01

I'm the same as you (don't know much about processors) but if they've got a PowerPC triple core processor running at 3.2Ghz AND it sits in a small enclosure (logical to assume it's running fairly cool), could this be the way forward for Apple CPU's?

Or is it a case of the Xbox CPU is still only comparable to a "Normal" G5 single/dual core G5 running at 2.8Ghz...

There must be a propellor head out there that knows ....
  quote
julesstoop
Veteran Member
 
Join Date: Oct 2004
Location: Leiden, the Netherlands
 
2005-05-13, 05:20

and he is working with IBM or Microsoft...
As far as I know details regarding the specifics of this chip are yet undisclosed.
For all we know it's a much simpler chip than the 970.
  quote
propropro
Member
 
Join Date: Dec 2004
 
2005-05-13, 06:28

I have read form some people in another forums that that "marvellous" triple-core processor is only equivalent to PowerPC G5 1.8Ghz.
Obvisuly it would be completely unreasonable and impossible to put a triple core, 3.2 GHz G5 in a console.
I repeat it again: this processor and the PM G5 are only very distant relatives
  quote
joebells
New Member
 
Join Date: Apr 2005
 
2005-05-13, 09:56

i'm hoping that it gets hacked sooner or later and instead of putting linux on it, or well probably in addition to putting linux on it, they manage to put osx on it.
  quote
wizard69
Member
 
Join Date: Oct 2004
 
2005-05-13, 10:07

Well I hope people are expecting something from Apple. The "MP" could very well be a dual core 970, but I honestly think the 970 series is dead in the water. So you have the possibility of either a Power5 device or something along the lines of a PPE derived device.

Apple would hopefully choose a device that would provide good performance in their systems (assuming that they even have a choice). Now the question is how well would a PPE derived device work in a MAC? For some work loads I would suspect very well. The problem then becomes how well does it work on specific work loads that Apples machines are used extensively in. If the processor performance does revert back to a crawl when running popular MA codes, then PPE/XBOX/whatever processors would be a hard sell in the MAC market place.

The other issue is the need to manage expectations. One can not always expect that a multithreaded proecessors performance can be easily related to the clock rate its running at. From the perspective of the PPE it does appear at this time that wone thread performance can be significantly impacted by the behavior of another thread. So getting good benchmarks may be vary interesting to say the least. On top of all that you will have all those other cores to support the PC, so thread interaction and intterrupt overhead won't be like what we see on current hardware.

What it boils down to is that nobody at this point in time can truely say how a MAC with a three core XBOX processor would perform. My feeling is that it would be a huge win for the MAC but that don't mean much. Hard data and real hardware are the only way to answer the questions.

By the way I don't believe that Apple needs to move completley away from 970 hardware. These new processors (XBOX/PPE/etc) could end up in specific hardware such as portables.

Dave


Quote:
Originally Posted by m0rpheen
......
Can we expect Apple to be getting the same for their processors? I know I know, an xbox isnt a Powermac, and I dont know much about processors, but I'd like to know thoughts
  quote
The Return of the 'nut
BANNED
I am worthless beyond hope.
 
Join Date: Dec 2004
Location: Berkeley
 
2005-05-13, 12:55

Quote:
Originally Posted by propropro
I have read form some people in another forums that that "marvellous" triple-core processor is only equivalent to PowerPC G5 1.8Ghz.
Obvisuly it would be completely unreasonable and impossible to put a triple core, 3.2 GHz G5 in a console.
I repeat it again: this processor and the PM G5 are only very distant relatives
care to provide some proof of that?

By the specs that have been released it appears to be very similar to the G5, even down to each core having its own altivec unit. If its equivalent to a 1.8Ghz G5 i'd imagine they would have saved the costs and just used that.
  quote
hmurchison
Veteran Member
 
Join Date: Jul 2004
Location: LV 426
Send a message via ICQ to hmurchison  
2005-05-13, 13:28

IBM hasn't released the specs. The fact that IBM can fab these with 3 cores, SMT, Altivec and at 3.2Ghz leads me to believe these are low complexity microprocessors designed to do relatively simple tasks. We need to more more about the inflight instruction and the queue/dispatch of instructions before we can tell what "grunt" it has

Warning - Obligatory car analogy coming.

An 8 cylinder motor can derive max horsepower and torque at as little as 5500rpm. A Honda Vtec engine gets maximum horsepower at 8000rpm. If you believed that engine speed was the end-all-be-all of performance then you would cash out for the vtec and be miffed when the 8 cylinder left you in the dust. Cycles don't tell us anything...what we can do with "each" cycle is of paramount importance.

omgwtfbbq
  quote
The Return of the 'nut
BANNED
I am worthless beyond hope.
 
Join Date: Dec 2004
Location: Berkeley
 
2005-05-13, 13:32

as usual (and unfortunately) AppleInsider's Future Hardware forum has a much better and technical discussion about this processor for those who are interested.
  quote
oldmacfan
Veteran Member
 
Join Date: Jun 2004
Location: Mile 1
 
2005-05-13, 18:11

Quote:
Originally Posted by The Return of the 'nut
as usual (and unfortunately) AppleInsider's Future Hardware forum has a much better and technical discussion about this processor for those who are interested.
I read that thread and most of those people who posted anything remotely intelignet are also members of this site.

So why the choose to post the crap here and not there is beyond me, I can't stand the adds.

Mile 1
  quote
The Return of the 'nut
BANNED
I am worthless beyond hope.
 
Join Date: Dec 2004
Location: Berkeley
 
2005-05-13, 19:09

Quote:
Originally Posted by oldmacfan
I read that thread and most of those people who posted anything remotely intelignet are also members of this site.

So why the choose to post the crap here and not there is beyond me, I can't stand the adds.
yea, that's why its unfortunate. we like tech geeks here as well!
  quote
JK47
Senior Member
 
Join Date: Aug 2004
 
2005-05-13, 22:13

Paul Thurrott thinks its the same, so it must be true......

It features a fast 3.2 GHz PowerPC processor with 3 processor cores, a feature Apple should be feeling pretty jealous about, since its fastest PowerPC chips are single core and quite a bit slower.

http://www.windowsitpro.com/windowsp...ott_46417.html
  quote
Ocelot
Member
 
Join Date: Mar 2005
Location: The Ragged Mountains
 
2005-05-14, 02:20

Arstechnica has a MacAch forum with a really good "perpetual future CPU" thread tht covers this in great technical detail. I don't understand all of it but it makes for good reading if you're interested.

The short answer is that the multi-core gamebox PPCs are very simple compared to the PPC 970, but they bode well for IBM's fabbing abilities. But you reall must read the thread yourself since there's no way I can summarize it well.

"I will tremble for my country when I reflect that God is just." - Thomas Jefferson
  quote
Snoopy
Member
 
Join Date: Jul 2004
Location: Portland, OR
 
2005-05-14, 09:21

It's likely that the Cell and Waternoose (Trinity?) have similar cores, with a long pipeline but very limited logic depth per stage. Overall, it has lower leakage and runs fast, but performs poorly compared with the Power series cores. Yet it is a tradeoff. The design allows IBM to add more cores to the chip. At this point, nobody may know where the future of CPU design is headed for sure. Once Apple has all the chips to try out, the design that performs best will likely win. Here is a quote from David Wang's article:

Jerry

" . . . the paper published in ISSCC 2000 described a processor that supported the complete POWERPC instruction set and operated at 1 GHz on a 0.25µm process technology. The microarchitecture of the research processor was disclosed in some detail in the ISSCC 2000 paper. However, that processor was a single issue processor whose design goal was to reach high operating frequency by limiting pipestage delay to 13 FO4, and power consumption limitations were not considered. For the PPE, several major changes in the design goal dictated changes in the microarchitecture from the research processor disclosed at ISSCC in 2000. Firstly, to further increase frequency, the per stage circuit delay design target was lowered from 13 FO4 to 11 FO4. Secondly, limiting power consumption and minimize leakage current were added as high priority design goals for the PPE. Collectively, these changes limited the per stage logic depth, and the pipeline was lengthened as a result. The addition of SMT and the two issue design goal completed the metamorphosis of the research processor to the PPE. The result is a processing core that operates at a high frequency with relatively low power consumption, and perhaps relatively poorer scalar performance compared to the beefy POWER5 processor core."

http://www.realworldtech.com/page.cf...WT021005084318
  quote
Programmer
Member
 
Join Date: Nov 2004
 
2005-05-14, 11:15

Quote:
Originally Posted by Ocelot
The short answer is that the multi-core gamebox PPCs are very simple compared to the PPC 970, but they bode well for IBM's fabbing abilities. But you reall must read the thread yourself since there's no way I can summarize it well.
Looking at the limited information we have naively, the 970 is a 4-5 way issue processor (but probably averages closer to 3 instructions/clock) and the new core (in XB2 and Cell, similar if not identical) are only 2-way issue (probably 1.5 instructions/clock on average). That means in the same number of clocks the 970 will do double the work. A 3.2 GHz XB2 core is probably roughly equivalent to a 1.6 GHz 970, and a 4 GHz Cell's Power core is probably roughly equivalent to a 2 GHz 970. The 2.7 GHz G5 in Apple's latest would roughly equate to a 5.4 GHz version of one of these new cores.

Also, this new core is 2-way SMT which typically means that you get two threads, each running at half speed (well, a little more because they are using up the wasted cycles of the other thread). A 3.2 GHz XB2 core, therefore, is something like having 2 x 1 GHz 970. If your code is properly multithreaded and can take advantage of lots of threads then the XB2 is promising about 6 GHz worth of 970, vaguely compareable to a 3 GHz 970MP.

Note that there are many many variables in performance so these can only be very rough rules of thumb and ought to be taken with a grain of salt until some form of real benchmark is available. An example of other factors is that both XB2 and Cell claim to have 22-25 GB/sec of memory bandwidth, compared to something like 6 GB/sec in Apple's G5 (although that isn't strictly fair since the GPU in the game consoles typically uses main memory bandwidth, but on a Mac it has its own VRAM for many things... and that VRAM is 20+ GB/sec). An example favoring the other direction is that the 970 is an out-of-order processor with pretty good branch prediction, so on crappy code it probably does quite a bit better than this new core... which is in-order so it likely spends a good deal of time spinning its wheels while struggling through crappy code. This means that properly optimized code is going to sing on the new cores, but poorly optimized code will choke them far worse than it does the 970 (which is a big part of why IBM originally designed the 970 core the way they did [as POWER4] for their server line).
  quote
Gargoyle
http://ga.rgoyle.com
 
Join Date: May 2004
Location: In your dock hiding behind your finder icon!
 
2005-05-15, 10:11

That all sounds very logical, Programmer. Also, you should all remember that this xBox chip is a "Custom IBM processor based on the PowerPC range". This mean's that IBM could have take the G5 (or any other PowerPC) as a base and then removed any instructions/features that the XB2 is not going to use in order to simplify the chip. What you will end up with is a CPU that is very good and a much smaller number of things. (Bit like the GPU's from nVidia and ATI, etc)

OK, I have given up keeping this sig up to date. Lets just say I'm the guy that installs every latest version as soon as its available!
  quote
tokamac
New Member
 
Join Date: Mar 2005
Location: Old Europe...
 
2005-05-16, 08:32

The Xbox 360 CPU has 3 x VMX units. And there are "128 registers per hardware thread" (see Xbox 360 Fact Sheet).
I recall that SPE implementation in Cell were originally stated to have VMX ISA (am I correct?), but Altivec was in fact too limited with 32 registers. So IBM went further and designed the SPE wider, with 128 registers.
Is there any difference between "128 registers" and "128 registers per hardware thread"? How much registers per hardware thread does the G5 has?
  quote
Programmer
Member
 
Join Date: Nov 2004
 
2005-05-17, 00:17

Quote:
Originally Posted by tokamac
The Xbox 360 CPU has 3 x VMX units. And there are "128 registers per hardware thread" (see Xbox 360 Fact Sheet).
I recall that SPE implementation in Cell were originally stated to have VMX ISA (am I correct?), but Altivec was in fact too limited with 32 registers. So IBM went further and designed the SPE wider, with 128 registers.
Is there any difference between "128 registers" and "128 registers per hardware thread"? How much registers per hardware thread does the G5 has?
In both cases programs have 128 registers available to use (actually VMX128 also has the normal PowerPC integer and float registers -- 32 of each). The "per hardware thread" thing means that the execution units of the core are shared between two threads... if each of those threads could fully use the processor then doing that would cut their speeds in half. Since most threads actually spending a lot of their time waiting, however, their resources aren't always busy anyhow so sharing them doesn't cost anything much of the time. In practice this means that you have two hardware threads running slower than one would, but faster than half the speed of just one.

The SPE, on the other hand, is designed from the ground up to have 128 vector registers and it has local very high speed memory of its very own to play in (plus a DMA controller to bring things in and out while the SPE core is busy working). The SPE is also smaller, which is how Sony/IBM has managed to fit 1 PPE + 8 SPEs onto a single chip, compared to Microsoft/IBM's 3 enhanced PPEs (or something so similar as to be equivalent). That's 9 cores to 3, and now it looks like they'll be at the same clock rate.
  quote
rickag
Member
 
Join Date: Jul 2004
 
2005-05-17, 08:48

Comparing Microsoft's 3 core PPC vs Sony's Cell PPC chip, would the following statement have any validity?(considering these are specialized cpu's for gaming.)

May the best programmers win.

Just waiting to be included in one of Apple's target markets.
  quote
Posting Rules Navigation
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Post Reply

Forum Jump
Thread Tools
Similar Threads
Thread Thread Starter Forum Replies Last Post
Core Audio, Core Video piwozniak Apple Products 19 2005-07-23 12:04
reason why you saw no dual core. jkloos Speculation and Rumors 42 2005-05-02 10:46
12" i/power book and Core Image jbloodwo Purchasing Advice 2 2005-05-01 22:22
Graphulator + Core 3d !! NaMo4184 Speculation and Rumors 3 2005-03-18 02:38
Core Image and Core Video- Requirements ASZ993 Speculation and Rumors 13 2005-02-18 14:14


« Previous Thread | Next Thread »

All times are GMT -5. The time now is 01:40.


Powered by vBulletin®
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004 - 2024, AppleNova