Slashdot Mirror
Ars Technica's Hannibal on IBM's Cell
endersdouble writes "Ars Technica's Jon "Hannibal" Stokes, known for
his many articles on CPU technology, has posted a new article on IBM's new Cell processor. This one is the first part of a series, and covers the processor's approach to caching and control logic. Good read."
Why do I have the sneaking suspicion that, if successful, this processor will eclipse the PowerPC on the Mac in the next few years?
I want 2 of them, yesterday.
... on cell... likely?
Aside from my own (competent) review of the cell processor, the article possibly the most insightful and technically nicely balanced articles posted on slashdot in a long while!
I'll cover more of the Cell's basic architecture, including the mysterious 64-bit POWERPC core that forms the "brains" of this design.
Looking forward to that... I think that many people will be moving to Mac
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
Part II is up as well.
.. made of risc components.
" Last fall, IBM and Sony said they were developing a workstation based on Cell chips, which is the first product IBM will ship based on Cell."
Regardless if this is the first product shipped or not, a workstation is coming. I can't see it running anything but linux. Given the mass market targeting of the cell, I hope Sony makes a strong go at grabbing the market with cheap hardware, rather than trying to milk the high-end content creation market first.
"A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
Beautiful as it is as a gaming CPU, I still am having trouble seeing how this thing would work in a PC of any sort. The only customers the Cell has so far-- Sony-- are talking about the stream processing units being directly coded for by the programmer. But general PC programmers still haven't accepted Altivec that well despite it being available, easy to use and useful, how are they going to react to "rewrite your program to use SPEs"?. Meanwhile if the programmer does directly code for the SPEs then that's all well and good in a video game system, but it's yet to be explained to me how on earth the SPE works in a time-sharing operating system. Specifically, who gets to use it, when, and what they do with the onboard SPE memory during a context switch? These are not trivial issues.
e.g. 234 M transistors (!) That's why I don't think this will be replacing the G5 any time soon. The die size (at the current prototype's 90nm) is over 200 mm2.
It'll have to get a fair bit smaller/cheaper before the PS3 can use it without major subsidies, and I don't know why they think general consumer devices will want it. God knows how much power it dissipates with all 8 SPEs clocking over at 4 GHz...
Why would anyone engrave "Elbereth"?
If you really want help, try posting an actual link instead of merely quoting a URL. We're all to busy/lazy here to copy and paste that into our browsers.
"I'm not impatient. I just hate waiting." - My Dad
Thank god. I've enjoyed his articles in the past, and if experience is any indication, I will have the false impression that I understand this stuff in a nontrivial way for up to three hours. This is not meant to rag on Hannibal, BTW.
WIth a name like that, I expect to see pictures of him eating those Cell processors, and describing how they taste.
// file: mice.h
#include "frickin_lasers.h"
...clicking on this link also attempts to install a trojan (SARC's name: ByteVerify). I agree: this link should be removed and the poster's IP should be reported to the relevant authorities.
If you really want help, try posting an actual link instead of merely quoting a URL. We're all to busy/lazy here to copy and paste that into our browsers.
Sounds like a pretty effective anti-slashdot effect mechanism to me.
Although the article (which is quite clear) indicates that the AltiVec architecture is closer to G4 than G5, won't the speed increase of having 8 fully-parallel processors (9 if you count the main CPU) more than make up for the issues associated with the loss of the G5's advanced features? It seems to me that this is a natural for Apple - it will give them a 5x - 10x performance boost over anything that's on the drawing boards over at Intel.
Even so, I doubt we'd see Cell-based Macs until at least 2007 - but wouldn't it be great to run PS3 games on your Mac? (As if that'll ever happen.) But then again, given the Cell architecture, your PS3 could use your Mac to make its games run faster! A whole new reason to have an XServe-based supercomputer...
The one thing I don't understand is how I would code for this thing. As best as I understand it, I now have some instructions for controlling the cache (or LAM, whatever) which sounds cool, but are there any details yet of how I'd write code for this? I'm also disappointed that the article didn't explain how one would use their SIMD instructions if they aren't using any of the existing standards. So I load my vectors with the cache control and ask the processors to ever so kindly add them?
Anybody out there with experience on this architecture or even attended the presentation itself can give us mere coders details? Preferably a website.
The bitter lessons of a veteran coder: http://bitterprogrammer.blogspot.com
Is that the 386 instruction set and arcitecture is so non proprietary. What made it so popular certainly wasn't that it was better. If I had the dough, I can literally make one and my own fab without asking a single soul. Alot of times it seems companies try to gather into consortiums to mimic the same effect and gather market momentum, but these are doomed to failure because the more valuable the technology becomes - the greater the pressure to diferentiate and fence off some "teritory" for themselves. We saw this happen first hand with UNIX, where all the flavors would constantly try to group under these unified standards - and they made little progress until Linux came along. The CPU world needs somthing similar to protect people from patent harassment. for design, cores, and fabrication.
Pattnaik said that if IBM were to publish the detailed monitoring information for end users to access, then the company would feel obliged to maintain backwards compatibility in future iterations, and so they'd be limited in the changes they could make to the scheme.
If I were IBM, I'd publish such specs anyway, alongside letting the press know very loudly and clearly that developers should stick to the recommended API if they want any guarantee of future compatibility. OTOH, I do understand their reasoning for doing this, and I don't completely blame them. Even if they did publish the register information with the very loud and clear warning, people would still complain about lack of backwards compatibility, something of which is meant to be incumbent upon the developer, and not IBM.
// file: mice.h
#include "frickin_lasers.h"
This chip seems insanely powerful. With 8 APU's capable of doing DSP, you would think that some countries would impose export restrictions on the thing. If you remember when the G4 came out Apple advertized that the military didn't want that thing leaving the country. But image a chip with the ability to do some serious SIMD operations? The CIA, NSA and others doing signal processing have to love this chip.
The views expressed are mine own and do not express the views of my employer.
that it runs at 30 watts, about like a Pentium M. And it's 64-bit. Can we say....
Dare I say....
Oh the Hell....
PowerBook G5!
I'm guessing it's because the Cell chip is going to be used in the Playstation 3.
"Our opponent is an alien starship packed with atomic bombs. We have a protractor."
I thought it was funny dude, sorry I don't have mod points.
Just then the floating disembodied head of Colonel Sanders started yelling Everything You Know Is Wrong!-Weird Al
As I fully expected, Pattnaik could not discuss a possible workstation-class derivative (read: Apple-oriented derivative) of the POWER5. He also made it clear that he is and has been focused on POWER5 servers only, and any hypothetical workstation-class derivative of the design would be for someone else to discuss.
I'm wondering about the feasibility of such a processor. This design seems to be rather heavily dependent upon the specific design of the OS (namely AIX in this case), and it seems to me that any OS that would want to take advantage of the POWER5 would need some heavy rewrites.
Of course, I could always be wrong on this issue, but I get the impression from the article that server oriented processors (namely Big Iron) and desktop processors are on a diverging path over at IBM. There may be some similarities, but I'm betting that there will be more differences than similarities in future processors.
// file: mice.h
#include "frickin_lasers.h"
What I find interesting is that the vector processor are restricted to single precision floating point calculations.
This isn't terribly useful for scientific computations (there is the same problem with the GPU): currently the IEEE is working on a standard for 128bit precision floating point calculations!
Of course for 3D, video and sound, 32bit precision is good enough and *if* programmers (a big if) manage to overcome the pain of 'parallel programming' then it could be a big success.
But does it run gcc? Or even have a cross-compiler target module? Will gcc become smart enough to emulate some of the SIMD techniques in my regular C++ code, even when I write the same old patterns?
--
make install -not war
Cradle Semiconductor has been working for a while on a similar technology.
Of course, it's all a matter of scale - TI had a 4 DSP, 1 CPU processor a while ago, but it only made 100 MFLOPS. Cradle's first product has 8 DSPs and 6 CPUs - depending on if you can get your data to properly pipeline through the processors, you can achieve up to 3.6 GFLOPs peak with only a 230 MHz clock.
HIV Crosses Species Barrier... into Muppets
Who would conceivably have enough money to build microchip fabrication facilities but not enough money to license the powerpc architecture?
"Reverse engineered implementations exist" is not really much of a meaningful strength if you don't own one such reverse engineered implementation already. You say you can potentially build a 386 chip fab, but the thing is you aren't going to build a 386 chip fab, you're going to just keep on buying Intel and AMD chips, the only noteworthy people currently making x86 chips, because if you built a 386 what would you do with it? It's a 386. The ISA has moved on.
With the low performance PPC cpu, I doubt Apple will want these things. Apple has too much interest in the general purpose computer market to care much about something like the cell processor, that is built for a niche market. For Apple, the cell processor will not only be expensive, but also slow. That goes double, since it is harder for Apple to get developers to suddenly switch from Altivec to the cell architecture than it is for Sony to do the same.
Whoever corrects a mocker invites insult;
whoever rebukes a wicked man incurs abuse.
--Proverbs 9:7
I commented on the wrong article.
// file: mice.h
#include "frickin_lasers.h"
If Sony can fit it in a console and sell a hundred million of them in a year, I'm sure Apple can fit it in a Computer and sell a few million of them as well. If you're going to talk about size/heat dissipation/price, that is.
Besides, those slides show that the 64-bit PPC on it has VMX. That's Altivec, baby. Sure, the SPE's don't have the full functionality of VMX but so what.
The biggest issue I see is that the Cell's design requires the programmer to have full control of the machine. Tell an SPE to do something and don't worry about it. There's a lot of "what if"'s in there. Like "What if some other program tells that SPE to do something else?" etc.
I'm guessing the registration info is fake. If the owner was in the US, why is the phone number in non-US standard? Not to mention using a Russian e-mail address.
Wheel in the sky keeps on turnin'.
Hurd might be an interesting candidate for running on Cell because of the highly threaded design. Hurd servers might be able to swap in and out of cells as they require cycles. It seems a good match; i.e. L4 runs in the main core, and various translators and other processes run on the cells. If a cell could be programmed to run the filesystem, for instance, it would totally free up the core for other business.
Because the PS/3 will have a highly fixed hardware set, implementing a minimal driver set might be feasible given enough reverse-engineering effort.
I'm not saying that L4/Hurd will kick the nuts off of Linux on an Opteron, I'm just noting that it might be pretty cool to experiment with Hurd on Cell technology. The L4/Hurd team is real close to getting the last peices in place to compile Mach based Hurd under L4, and if you ever tried Debian GNU/Hurd, you know its pretty near feature-complete and a pretty neat system to run. The next task for L4/Hurd is a driver infrastructure, and it might be wise to look at what Cell is bringing to the table before it gets too far along. Know what I mean.
Clickety Click
In part II, he writes:
"Finally, before signing off, I should clarify my earlier remarks to the effect that I don't think that Apple will use this CPU. I originally based this assessment on the fact that I knew that the SPUs would not use VMX/Altivec. However, the PPC core does have a VMX unit. Nonetheless, I expect this VMX to be very simple, and roughly comparable to the Altivec unit o the first G4. Everything on this processor is stripped down to the bare minimum, so don't expect a ton of VMX performance out of it, and definitely not anything comparable to the G5. Furthermore, any Altivec code written for the new G4 or G5 would have to be completely reoptimized due to inorder nature of the PPC core's issue.
So the short answer is, Apple's use of this chip is within the realm of concievability, but it's extremely unlikely in the short- and medium-term. Apple is just too heavily invested in Altivec, and this processor is going to be a relative weakling in that department. Sure, it'll pack a major SIMD punch, but that will not be a double-precision Alitvec-type punch."
Linux on Intel: Think Dead Man Walking and Grid vs. SMP: The Empire Tries Again and Fast, Faster and IBM's PlayStation 3 Processor.
Another article on the Cell design at http://www.theregister.co.uk/2005/02/03/cell_analy sis_part_two/ seems to indicate that there is some sort of DRM built in.
Hannibal doesn't say anything about this (that I noticed) - anyone have more info?
Don't save Windows XP! http://www.petitiononline.com/jjw1xp/petition.html
This RAM functions in the role of the L1 cache, but the fact that it is under the explicit control of the programmer means that it can be simpler than an L1 cache. The burden of managing the cache has been moved into software, with the result that the cache design has been greatly simplified. There is no tag RAM to search on each access, no prefetch, and none of the other overhead that accompanies a normal L1 cache. The SPEs also move the burden of branch prediction and code scheduling into software, much like a VLIW design.
Why? The reason for the instruction window was to simplify software development.
Of course, I like to play devil's advocate with myself, so I'll answer that question.
The purpose of the Cell processor is to enhance home appliances, which have a greater reliance upon low-latency than they do on precision, accuracy, and performane bandwidth. Thus, one can very safely say that the Cell processor will likely have little purpose in scientific calculations.
// file: mice.h
#include "frickin_lasers.h"
The parent (to your post) was joking. Quadras are Motorola 680x0 machines, not PowerPC. You'd have no more luck running OS X on one of those than you would Windows XP.
Hey "never say never", but I don't see Microsoft (xbox2) porting/releasing ANY Windows technology on Sony (ps3) hardware any time soon. The Xbox2/PS3 showdown is going to be the biggest thing since, well, Xbox/PS2.
No, I did not read the f***ing article!
Right now, it has 4x as many transistors as a G5, runs at twice the clock speed, and likely puts out a hell of a lot more heat than a G5 does.
Why would anyone engrave "Elbereth"?
A proposal for Apple
I don't have an account, but this is an honest idea.
Why doesn't Apple include a Playstation 2 support card into their Macintosh line?
Problem: The OSX platform has almost no games. I own several macs, I love my macs, and I sincerely enjoy OSX. But it has no games, and that will never get better, especially as simpler games migrate to the web and the complex ones bail for the console market. The PC gaming market has essentially peaked.
Solution: Embed (or include as a BTO option) a PS2 chipset to a Macintosh. Run the generated display straight through to the graphical overlay plane. Done.
Everything works. The controllers are trivially converted to use USB. The DVD drive is already there. The display is already there. The USB and Firewire is already there. The harddrive is already there. The "memory cards" are already there.
Reason: The Macintosh game library explodes instantly to encompass something like 3,000 PS1 and PS2 games. With no need for emulation, the games are guaranteed to work out of the box and provide the Apple ease of use everyone loves. Sony increases their marketshare, Apple gets a viable expanding game library, and users get a vastly better gaming experience on OSX for maybe $40 of parts and engineering.
Why won't this work?
The difference is that instead of the compiler taking up the slack (as in RISC), a combination of the compiler, the programmer, some very smart scheduling software
Requiring programmers to learn how to write parallel code that makes good use of this processor seems pretty dicey to me. Few programmers have been trained to write parallel code (most struggle with threading). The fact that no popular programming language has a good parallel model is also a big stumbling block.
This problem seems to be looming for all the dual core processors, but I havent seen a big effort to teach programmers how to adapt.
If you read books, or at least watch the history channel... then you would know Hannibal was the man that brought elephants over the Alps and routed and slaughtered Romans by the ten of thousands using but swords and spears. The only thing people dis him about was his one mistake to not go ahead and take Rome -- instead of giving them the chance to surrender. Hell, if lack of perfection is your only flaw that's a hell of a compliment.
He also was a great politician after the Tunic wars.
Stanford professor Dally's stream processor.
It's almost just like Cell but has onchip memory to solve the bandwidth problem.
Dally worked for Cray and mentioned that todays supercomputers are not efficient.
New consoles are sold at a loss, but there's a limit to how muc of a loss companies can take. If the CPU itself ends up costing Sony $300+, they'd be looking at a massive loss on the consoles, probably larger than they are willing to take. That was actually a noted problem with the X-box, the loss per unit was large so they had to sell quite a few games per unit to make it up. I'm not even sure if they made any money on it.
Well, in MS's case, they can pull shit like that. Microsoft makes loads of cash off their software division, and has loads already in the bank. They can afford to operate a new division at a loss, even a pretty substanital loss (if the X-box division did lose money, it wasn't a large amount).
Sony, not to much. Their Playstation divison is their biggest money maker these days. So they can afford to take a loss on console hardware, but only so much that they know they'll make it back on games. They can't risk operating the division at a loss because it'd spell serious trouble for the company. They also aren't flush with cash. They've about $10 Billion, but have $12 Billion or so in debt (Microsoft has $34 Billion and no debt to speak of). They have to keep the money rolling in or things get ugly.
Also we know from history that having the fastest processor or shinest graphics isn't what wins a given round of the console wars. It's all about games, and perception.
Now who knows on pricing at this point, but the grandparent has a good point. That is a massive god damn die, like P4EE sized or so. Hot and expensive. As die size goes up, so do failure rates and thus cost, espically at high clock speeds. Hence why the EEs cost so damn much. I'd say it's a safe bet that this cell processor isn't going to be cheap.
From the sounds of it, it's not going to need to be. Sounds like it's a high end calculation chip for badass number crunchers. Given that Power4/5s and Itanium 2s are popular for that sort of thing, people in those apps won't bat an eye at a $1000+ price tag.
So how is this architecture so revolutionary and amazing when compared to the processor in the Cray X1? (Which has a MIPS [like PPC, but without broken IO] core, and multiple vector units configured in much the same way as this seems to be).
I do not represent myself.
Thanks.
Wheel in the sky keeps on turnin'.
A budget-class PC laptop of that time might have been about 900 MHz to 1.1 GHz. I wouldn't consider such a laptop anything near useable. They tended to have poor quality sound systems that bottlenecked the processor and atrociously short battery times. The ibook was legendary for its excellent battery performance
Get off what you 'assume', assumption is just intuition for idiots.
We have test 200mhz laptops with 80mb of ram 5gb hard drives, released 1997 all running WindowsXP Professional (yes even the themes turned on) and they benchmark faster than they did when they shipped with Windows 95.
Secondly, they can do full 30fps video as long as it is uncompressed AVI or even WMA 9. QuickTime (MPEG4), MPEG2, and real stutter horribly on video playback unfortunately.
As for battery, don't know, these laptops hold for 3hrs with a single charge, and yes techs are REQUIRED and have no problems using them daily in test scenarios.
Now if you really want to compare laptops to laptops, why don't I show you our 900mhz AMD Compaq laptops, they have JBL sound systems in them, and there isn't a single feature the cannot perform with the exception of running a T&L based video game, as the integrated video doesn't handle it, oh wait, the 900mhz PowerBook video didn't support such features either. (BTW, This is not to say that there are not several 900-1000mhz class laptops that have upper end video features), I am just using what we have in our test labs for comparison.
The 900mhz laptop has a DVD/CDRW, came out late 2000 early 2001 (trying to remember if we got them before holidays or not). They do full software DVD decoding with less than 20% CPU utilization and pretty much do anything fairly fast that we through at them. We even have a beta version of Windows 2003 server running on one with 256mb of RAM. (Yes we are always pushing the limits, but it works as fast as the WindowsXP pro version of the machine sitting next to it.)
Now off my rant... Macs truly are great, and the PowerBooks of the time were great, but that DOES NOT MEAN they were the BEST, WILL ALWAYS BE THE BEST, or you should be complacent listening to Apple tell you what you are getting is the best when it might not be. It is time for us as MAC users to stand up and DEMAND that technology becomes as much a part of what a MAC is as the EASE of USE in the Interface.
The time is now, we need to STOP accepting what they tell us and give us and force them to truly give us the LATEST technological concepts, not just the above average concepts when compared to the PC world. These are Macs, they SHOULD BE BETTER. IT shouldn't even be subjected to a debate they should be so far advanced a debate should not be possible. PERIOD.
Sadly, it just isn't true now, and has not been for many years. OSX has giving the Mac world some credibility backing OS technology, but not Apple needs to take Macs to the next level.
Even if my comment inspires one Mac user to say hey Apple, we want better, then maybe we all can be the symbolic person with the hammer from their 1984 video and WAKE THEM UP this time.
you can keep the malware off.
But I can run Appleworks and Apple's mail on my 300 MHz, 192 MByte iBook while it serves apache to the web and compiles things like tomcat.
And I don't have to fight with stupid VB macros! (Do have to fight with stupid Applescript scripts.)
I haven't attempted to tune it to handle a slashdotting, so I won't offer a pointer.
will be using this "SoC".
However, after having RTFA(s), the Cell processor
would look like a very good candidate for a F/OSS
VIDEO BOARD - fast multicore processors, a large
local memory, simplified RISC with most control
in software, and a 64-bit PPC "traffic cop".
One additional area (at least) that I would
expect the Cell processor to be incorporated
into would be next generation radar and sonar
systems, due to vector processing capabilities.
I would love to see an IBM development system
for this architecture, but wouldn't expect to
buy a PS3 and Sony "game" SDK, due to closed
source and NDA incompatabilities with GPL.
I want an OS I can use.
Not quite. The Cell is 9 complete yet simple CPU's in one. Each handles its own tasks with its own memory. Imagine 9 computers each with a really fast network connection to the other 8. You could problably treat them as extra vector processors, but you'd then miss out on a lot of potential applications. For instance, the small processors can talk to each other rather than work with the PowerPC at all.
Hardly. Sony is following the same game plan as they did with their Emotion Engine in the PS2. Everyone thought that they were losing 1-200 bucks per machine at launch, but financial records have shown that besides the initial R&D (the cost of which is hard to figure out), they were only selling the PS2 at a small loss initially, and were breaking even by the end of the first year. By fabbing their own units, they took a huge risk, but they reaped huge benefits. Their risk and reward is roughly the same now as it was then.
Doubtful. The problem is that though the main CPU is PowerPC-based like current Apple chips, it is stripped down, and the Altivec support will be much lower than in current G5s. Unoptomized, Apple code would run like a G4 on this hardware. They would have to commit to a lot of R&D for their OS to use the additional 8 processors on the chip, and redesign all their tweaked Altivec code. It would not be a simple port. A couple of years to complete, at least.
This is half-true. While it will be hard, most game logic will be performed on the traditional PowerPC part of the Cell, and thus normal to program. The difficult part will be concentrated in specific algorithms, like a physics engine, or certain AI. The modular nature of this code will mean that you could buy a physics engine already designed to fit into the 128k limitation of the subprocessor, and add the hooks into your code. Easy as pie.
Bwahahaha! No way. This is a delicate bit of coding that is going to need to be tweaked by highly-paid coders for every single game. Letting on OS predictively determine what code needs to get sent to what processor to run is insane in this case. The cost of switching out instructions is going to be very high, so any switch will need to be carefully considered by the designer, or the frame-rate will hit rock-bottom.
This is one myth that could be correct. The Cell is huge (relatively), and given IBM's problems in the recent past with making large, fast PowerPC chips, it's a huge gamble on the part of all parties involved that they can fab enough of these things.
Now off my rant... Macs truly are great, and the PowerBooks of the time were great, but that DOES NOT MEAN they were the BEST
At the time there was no sony vaio, so the powerbook titanium was the smallest laptop around. It also had optional wireless and standard firewire and gigabit ethernet built in. Os 10.1 was a bit lacking but i'd take it over whatever windows version any day (i tried 98 2000 and xp home)
I'd say it was the best.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
The SPARC V8 spec is open, there's also an open source implementation: the Leon and it's supported by Linux.
What VCR?
Last VCR we owned died while I was out of the country. Probably has something to do with the significant other unplugging everything when we leave the house, so, yes, I gave up resetting the clock well before we started dating.
But even then before it died, it didn't "blick" 12:00 when we were watching because all that was on-screen. No LED/LCD panel. We never saw the "blicking" 12:00 unless we deliberately selected the setup screen.
Now, the microwave does "blick" 12:00 while it's in use, except when we put something in to cook while we're out. (When I say she unplugs everything, I mean everything. Except the fridge. And the iBook which is hosting my personal web page.)
So, what's your point?
be parallellizeable to some extent?
Granted, gcc would need som reprogramming to work this, but I think with a decent compiler-rewrite any standard c-code should be able to run on chips like this with at least some parallellization-benefits.
Not Buzzword 2.0 compliant. Please speak english.
As much as the idea of massive parallelism appeals to me (closer to reality), the more I read about Cell, the more I wonder if this is the right direction to be heading in. Sony is pushing the low level housekeeping back up the stack into software, while Microsoft is making things as easy as possible for coders by letting them write solely to high level APIs like DirectX.
Cell will no doubt be super fast when driven correctly, but that could prove to be a significant additional cost to the development budget.
But what would I know..
No, I did not read the f***ing article!
Fast and easy for:
:(
Highly prallelizable single precision floating point calculations.
Slow and difficult for:
Complex integer logic.
Double precision floating point math.
From the looks of it, Cell will be a decent gaming engine. Expect highly impressive (not necessarily realistic) gaming physics. AI developers, though, may not be able to squeeze enough performance from Cell for leap ahead AI complexity unless they figure out a way to do AI with floating point matrices. We may therefore start seeing neural networks becoming more popular for PS3 AI's.
The question is, other than gaming development, what will IBM's Cell workstations and servers be good for? Possibly movie CG, military training simulations, imagery processing, etc.
Those hoping for ideal scientific computing platforms will have to look elsewhere unless IBM comes out with double precision SPUs.
I priced 'em. I bought my first iBook and a $500 developer's subscription. I didn't need to pay $500 for the developer's subscription, to get Mac OS X beta, but I did, and I still paid well less than what I'd have had to pay for an equivalently configured MSWxxx lapbox.
Admittedly, Linux on iNTEL would have cost about the same or less money, but it would have well made that up in the cost of time to bring up the dev environment.
There's also a factor of requirements that shouldn't be ignored.
Huh?
You believe in magic? Should we ask then why the Linux kernel is still improving? Why the Gnome desktop is still improving?
You might criticize the choice of branch-merging the BSDs with a Mach kernel, and you might question the wisdom of objective-C in system-level code, but criticizing a company for cleaning up their code after they have the product out the door and making money seems a little unreasonable to me.
Give us all a break. We all love Linux, and/or *BSD, and we all enjoy compiling our apps (yes we do) and kernels and writing our own firewall and mucking about in /etc and /var. What's that got to do with the price of tee in Redmond?
Well, anyway, bragging on price of hardware and spending days bringing the system up so you can use it seems a little confused. You buy the system with your money and your time, and you choose it according to what you need to do with it. (Which is why I need to get a Mac mini and put Yellow Dog on it, and get a single-board PPC and put openBSD on it. Three different sets of requirements there.)
Personally, I think the war of words about whether Macs or PCs are better is a waste of time. MSWxxx vs. the real world, yes, that's worth discussing. But Linux vs. BSD vs. Mac OS X? iNTEL vs. PC? Praise your deity (or the viscissitudes, as you choose) for variety, and, in the case of PPC/Mac OS X, be glad it's close to price parity.
So basically this processor is an attempt to address what this guy calls 'Memory to Execution' latency, for a very specific set of applications - ones with lots of 4-D vector transforms. General purpose code execution is probably going to suck, but it could shine in video games and perhaps some types of robotic and embedded applications, but that will depend upon power-usage and heat. Maybe the programmers will have Super Mario 12 ready for it by 2013.
Do your test machines that you have running these more recent OS's have AntiVirus protection on them?
I talk about stuff.
Dude, you're a riot. This is the third one I've noticed. How many more as I scroll down? =) Oh, and I made you a friend, since nebbishes like you need as many friends as they can get. I kid! I kid! You're not really a nebbish.
It's not offtopic, dumbass. It's orthogonal.
Ummm they did put PS2 chips in workstations. They called it the GS Cube.
I suppose that if Saddam had wanted to cluster a bunch of PS2's he'd have had his techs find out the information from the NCSA:
Want a PS2 dev kit? (Sorry all sold out for NTSC U/C territory)
Sorry, messed up on the URL's
The cluster
http://arrakis.ncsa.uiuc.edu/ps2/
The dev kit
http://playstation2-linux.com
GCC should be able to compile binaries for the cell cpu in 1 year of the cell's release, meaning someone should be able to compile netbsd and/or linux for it, both the kernel and userland. I would imagine it would not be optimised much first, but threading will suddenly be important on both platforms, and we can expect much more attention on libpthreads, as well as the SMP scheduler of the linux kernel. It should be able to recognise different-strength CPUs and assign tasks, even reserve a cpu for graphics functions, reserve another for running certain drivers etc. (I dont know if its currently capable of all that).
So in a few years time its possible the development of at least one of the unixen, will be focused on highly threaded distributed applications, new scheduler and libc designs to help in that, and better levels of performance than a linear monolithic kernel with none or bad threading and the simple scheduler that linux had a while ago.
Would it kill Palm to release opensource BeOS at a time like this, when a BeOS compiled for the cell CPU will be the killer desktop OS?
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
Well, it certainly might seem that he is being a hypocrite. See:
"In another part of the article, Blachford claims that the cell processing units have no "cache." Instead, they each have a "local memory" that fetches data from main memory in 1024-bit blocks. Well, that's sort of like saying that an iMac doesn't have a "monitor," but it does have a surface on which visual output is displayed. In other words, the Cell "local memories," which are roughly analogous to the vector units' "scratchpad RAM" on the PS2's Emotion Engine, function as caches for the PUs. What has thrown the author for a loop is that they're small, and the fact that they're tied to each cellular processing unit means that they don't function in the memory heirarchy in the exact same way that an L1 does in a traditional processor design. They do, however, cache things. But maybe I'm being nitpicky with this."
and
"Finally, to address something more specific to the Cell architecture itself, on page 1 we find this claim:
It has been speculated that the vector units are the same as the AltiVec units found in the PowerPC G4 and G5 processors. I consider this highly unlikely as there are several differences. Firstly the number of registers is 128 instead of AltiVec's 32, secondly the APUs use a local memory whereas AltiVec does not, thirdly Altivec is an add-on to the existing PowerPC instruction set and operates as part of a PowerPC processor, the APUs are completely independent processors.
The author appears to be confusing an instruction set with an implementation. The 128-register detail is a problem, because, as the author correctly points out, conventional Altivec has only 32 vector registers. So obviously it's a given that Cell won't be using straight-up Altivec. But it's entirely possible that it'll use some kind of 128-register derivative of the Altivec instruction set. The fact that the individual processing units have a local cache has little to do with whether or not the PUs themselves implement some hypothetical Altivec derivative. Finally, the statement, "Altivec is an add-on to the existing PowerPC instruction set," is correct, but the rest of that sentence--"and operates as part of a PowerPC processor"--doesn't make a whole lot of sense to me in this context. Altivec is an ISA extension that is implemented in different ways on different PowerPC processors. The Cell processor's PUs could very well implement a hypothetical 128-register Altivec2 ISA extension, or they could implement some other SIMD ISA extension. The fact that SIMD code, written to whatever ISA, is farmed out to individual PUs has nothing to do with it. (If what I just said confuses you, you might check out this article.) "
compared to
"The main differences between an individual SPE and an early RISC machine are twofold. First, and most obvious, is the fact that the Cell SPE is geared for single-precision SIMD computation. Most of its arithmetic instructions operate on 128-bit vectors of four 32-bit elements. So the execution core is packed with vector ALUs, instead of the traditional fixed-point ALUs. The second difference, and this is perhaps the most important, is that the L1 cache has been replaced by 256K of locally addressable memory. The SPE's ISA, which is not VMX/Altivec-derivative (more on this below), includes instructions for using the DMA controller to move data between main memory and local storage. The end result is that each SPE is like a very small vector computer, with its own "CPU" and RAM."
But if you read closely you will see that Blachford, to generalize, was "right" (e.g. local memory and no AltiVec on SPE) for the wrong reasons, and even then some of the info was factually incorrect (e.g. SPE fetches blocks of 1024 bits). I do think that Hannibal was too hard on the guy (probably because of his completely unsubstantied claims about performance) and I think Hannibal should've cut Blachford some slack based on the source material that Blachford had available to him (although Blachford's
Reading the article, it reminds me of the typical mainframe architecture, where you have a central supervisory CPU, but most of the specialized work is done by the channel processors.
In the Cell, the main PPC CPU appears to identify a piece of work that needs to be done, schedules it to run on a SPE, uploads the code snippet to the SPE's LS via DMA transfer, and then goes off and does something else worthwhile while the SPE munches on it. I presume there's an interrupt mechanism to let the PPC know that a SPE has some results to return.
Compiler writers ought to be able to handle this new architecture well enough -- it's sort of like the current CPU/GPU split, where you've got the main program running on the system CPU, and specialized graphical transform programlets running on the GPU. There may need to be macros or code section identifiers in the source to let the compiler know which to target for that bit of code.
Obviously, this is just the first iteration of the Cell processor. I can see them widening the SPE from single precision to double precision (for the scientific market -- the game market probably doesn't need it), and going to a multi-core design to reduce the die size.
Chip H.
Actually the latest graphics processors from ATI and NVidia have about 200 Million transistors already and an Athlon has about 50 Million.
So a cell is probably going to be faster, smaller, cheaper and runnig cooler than the usual CPU+GPU system we have in most highend PCs today.
You're forgetting that a machine with the Cell processor still needs seperate GPU. The PS3 will use the Cell, but it is also using an Nvidia GPU.
So you can't compare the Athlon + GPU to a Cell, you need to compare the Athlon + GPU to a Cell + GPU.
Wow. Rated as redundant, even though I was the first to post this particular observation (I posted when there were only 32 replies). The joys of slashdot.
:)
Troll me down!
Anyway, regarding the reply to my response, thank you for taking the time. I was just too lazy to start digging out the passages. Hypocrite may have been too harsh a word. I think that Hannibal was probably just dismissing an author who had jumped to conclusions based off of vague and overly complex source material. Especially since that author has had a history of doing so in the past. I think it would have been nice, however, for Hannibal to actually read the patent himself and see if he came to different conclusions. But anyway, the past is the past.
It seems to me that this is a natural for Apple - it will give them a 5x - 10x performance boost over anything that's on the drawing boards over at Intel.
Don't believe the hype. Sony is using the same pattern of press releases and hype that they used regarding the Emotion Engine before the PS2 came out. They claimed it would be more than twice as fast as the highest end Intel chips.
When it actually came out, the performance it delivered wasn't even up to par with the mid range Celeron/P3 that the Xbox used. I'm sure in a few hand-selected calculations it excelled, and gave them the legal ability to claim it was more than twice as fast as the highest end PC chips. But real world performance paled in comparison to the claims.
Sometimes I wonder how many times you can fool a dog with the same trick. Most dogs learn after a couple times, but some never seem to figure it out.
In this case, we have Sony releasing all sorts of big claims and hype just like they did with the Emotion Engine. Anybody remember that? It's a tried and true tactic meant to generate interest in a product. It leads people to believe that something very exciting is about to occur, a revolution is just around the corner... this time maybe their wildest dreams will come true.
But as usually is the case, and the case with the Emotion Engine, it turns out to be "just another chip". Regardless of all the claims and details on paper, the actual silicon has limitations. Rarely does something revolutionary occur; usually it's more of a modest evolutionary improvement.
But modest improvements don't generate the hype that "amazing, revolutionary" product announcements do, so the PR departments do what works and they try to fool the unsuspecting public.
And here we are, falling for the same tricks.
Trolls are people too.
It's been 15 seconds since you hit 'reply'.
That was classic intercourse!
And there's nothing wrong with that.
That was classic intercourse!
Your post: "OK, if you want to include emulation then technically it might be doable with a 68k port of PearPC. But it's not going to happen natively."
Do you see any mention of native code running in the original post? Are you not a native speaker, and you confused the word "nicely" with the word "natively"?
And wtf is a "68k port of PearPC"? Do you even know what PearPC is? From sourceforge:
Maybe you should ask someone that is knowledgeable about computers what "architecture-independent" means.
Maybe I'm being too rough on you, but what sort of an ass tries to get in the last word and counter a funny example from danamania with a stupid qualification like "But it's not going to happen natively." Of course it's not going to happen natively! Did you go to school in Kansas or something?
It's not offtopic, dumbass. It's orthogonal.
Because you didn't read the article?
The only cores they currently stack are memory chips; this is because memory logic runs comparatively cool. Stacked Cell processors would melt.
I read some of the other responses regarding Blachford's other inane and idiotic theories. Just because he has has said stupid things in the past doesn't mean everything he says should be immediately dismissed.
While technically that's true, it does mean that you should take the things he has to say with a grain (of a 50 lb bag) of salt.
Someone's past performance is a pretty good indicator of their future performance. This guy has a history of being a loon, so it's helpful to keep that in mind when listening to what he has to say.
At the time there was no sony vaio, so the powerbook titanium was the smallest laptop around
Actually, no...
Sony has hand cross handheld size laptops since at least 1997. They had built in Camera, full size keyboards, and were full computers that ran about 600 or 700mhz. The are also tons of other lesser name brands that have comparable computers with even more features. Check out Fuijitsu and countless others.
If you think the Apple PowerBook is extraordinary, they you NEED to pay attention to the REST of the world a bit more, then GO BACK TO APPLE AND SAY - "WHY DON'T WE HAVE THESE FEATURES, ESPECIALLY THE ONES GEEKS IN THE NON MAC WORLD HAVE HAD FOR YEARS AND YEARS"
As long as you eat the dog food Apple gives you and you NEVER question it, you will NEVER get the cutting edge or leading technology from Apple. Why do they need to invest in R&D, their marketing department can just simply fool the customer base, as they are CONTINUALLY DOING.
Turn them around and instead of being fooled; get them to give us the BEST TECHNOLOGY THERE IS, PERIOD.
Please wake up Mac users and stop justifying what the 'great' Apple is giving us, and demand true technology leading equipment in EVERY aspect.
There are so many fringe concepts, theories and technologies that they could implement today and they don't because the users don't know any better and are easily conned with the technology Apple keeps giving them.
Do your test machines that you have running these more recent OS's have AntiVirus protection on them?
Long term test units do, as well as units that are directly involved in beta testing of anti-viral software. Install and wipe system usually don't get it, and most techs don't run real time virus software, as they are monitoring what is coming in and out on the system and nightly or weekly if the are away from the shop, scans are plenty to keep the systems virus free.
Now I will bite... why?
If this is going to turn into so debate over the performance degradation of real-time scanning or something like that, you can skip it. There is a performance hit for some anti-virus real time protection based on the heuristics turned on at the time, however, considering most anti-virus software (like Norton) was running real time virus protection with little to no performance loss on 386 and 486 class systems, the numbers in today's comparisons are not very big at all.
Doh!
- jon
Ganymede, a GPL'ed metadirectory for UNIX
I stand corrected on the vaio line being already available around that time.
All the rest of your comment seems not very informative. I haven't found any laptops who had the same array of features of powerbooks at that time - i repeat: wireless, gigabit ethernet, firewire. If you think an integrated cam is the same, we are obviously requiring different things from our machines. More specs and less schooling, if you can.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
Actually I didn't think your 'incredible' specs were worth even mentioning. The only 'spec' that would not have been a standard feature in the PC world would have been the gigabit ethernet. So Apple gets a point there.
And the camera on the sony wasn't a 'feature' it was one of the few things that helped people remember them. If you want a list of 20 features of from PC laptops that you were standard or could get that Apple didn't offer, just ask. Geesh.
I can give you a list of stuff from my 2002 laptop that Apple STILL DOES NOT OFFER. Try a 1600x1200 LCD (Darn nice for us graphic artists - too bad Apple don't cater to them anymore). Shall I go on, or do you get the point?
Quit DEFENDING WHAT THEY ARE SPOON FEEDING YOU AND DEMAND MORE FROM THEM. By reponding to my posts are you doing nothing but proving my point that you are brainwashed/have no idea what Apple has compared to other products/and do not have the guts to stand up to Apple and EXPECT MORE!!!
My rant is get Apple fans to be REAL FANS and not just take what Apple gives them and instead question what Apple is giving them.
I know when people mention the Xbox-2 a group always mention backwards compatability like a crazed console fan boy and then the rest of us tell them why it would be very hard due to architectural differences. We take it as a given that the ps3 will be backwards compatable and it seems the design fo the cell seems to have backwards compatability in mind. It's veyr loosly structured to be similiar (to my primitive knowlege of the chip layouts in the ps2) to the EE chip.
"There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
Well, artists got their 17" monitor, anyway you are talking to the wrong person here. Sure, I think OSX is the best desktop OS out there, and have been using Apples since //c, but i'm no Apple or Jobs fanboy, I go for what I think is the most rewarding computing experience. That is, debian (see my journal).
Apple fanboys might sure ask for better hardware. Microsoft fanboys might sure ask for a better OS. Linux fanboys might sure start STFU, boycotting closed hardware manufacturers, and contributing code and docs.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
Well, artists got their 17" monitor, anyway you are talking to the wrong person here
Yeah, at 1440x900... Great for geeks watching movies, not very high resolution for Graphic Designers...
And at 17" the pixelization is very very noticeable and annoying.
And sorry for harping on you directly, that wasn't my intent.. My posts tend to be more for the general consensus than just usually directed at a single person in a thread.