Slashdot Mirror
IBM Ships Fastest CPU on Earth
HockeyPuck writes "The 5-billion-instructions-per second Power6 processor from IBM would beat such rivals as the 3.73 gigahertz Pentium Extreme and the 2.4 gigahertz UltraSparc T2 from Sun. 'It's hard to make the average person understand just how fast this is,' said IBM Chief Technology Officer Bernard Meyerson, offering an example meant to explain his company's baby that still leaves the listener awed with the speediness of the two laggards. 'Hold your index finger out in front of your face,' Meyerson said in a telephone interview from IBM headquarters in New York. 'In less time than it would take a beam of light to travel from your knuckle to your fingertip, the new IBM chip would complete one task and start looking for the next, he said.'"
My blog
I'm glad they stopped measuring chip speed in Hertz and are now using the simpler metric fingertip-to-knuckle units.
steampunk web design
and all the water turned into steam!
If you keep throwing chairs, one day you'll break windows....
Too bad Apple no longer uses IBM processors, this would've been a great marketing scheme for Steve Jobs.
'It's hard to make the average person understand just how fast this is,' said IBM Chief Technology Officer Bernard Meyerson, offering an example meant to explain his company's baby that still leaves the listener awed with the speediness of the two laggards.
Made me think of a National Lampoon Radio Hour (SNL before it was on TV) skit about the George Foreman-Muhammed Ali fight. Foreman (John Belushi IIRC) talking about Ali:
"He so fast he can turn off the light and be in bed before the room get dark!"
mcgrew's razor: Never attribute to stupidity that which can be explained by greedy self-interest
IBM just hired some peeps from the Intel marketing division.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
ob disc: I work at sun (but not ON those chips).
I write management software that lets admins turn on/off/standby (etc) the state of the various 'cpus' (threads, as sun calls them). there are 128 and 256 cpus in a regular 2u..4u style rackmount box. these are 'simple' air cooled systems with fans blowing over the whole U-style chassis and over the passive cpu heatsinks. nothing 'scary' at all, really.
it is pretty wild to be able to do the equiv of 'show cpu' and have an ascii output scroll 64, 128 and even 256 times; one for each 'cool thread' which is a real actual processor element.
the down side is that this threading stuff does not automatically get you faster speed on a SINGLE non-threaded traditional task. as I understand it, these T-series sun boxes are meant to process a lot of transactions (think webservers) and not so much number crunching.
how do you define 'fastest chip'? well, one thing is for sure, you do NOT simply go by 'gigahertz' alone. that's really an oversimplification.
--
"It is now safe to switch off your computer."
... it will be in a washing machine controller.
Nice chip. Now what OS and applications run on it?
I'd guess anything that runs on the Power archicture. Here's a list of the various OSs that have been supported on various iterations of the Power architecture at one time or another.
This guy's the limit!
Correct me if I'm wrong, but don't most of the algorithms for scientific simulations run just as well on parallel processors? If this is the case, it makes more sense to have cheaper processors (both to manufacture and run) so that the cap on the IPS is raised just as well (the cap being the ratio of the amount of funds an organization can allocate to these emulators versus the cost of the emulators). Though I'm no computer expert, it seems that making one sequential processor run faster isn't as efficient as making that same processor cost a fraction of the price to get more power from your dollar.
Help fight spam
For multi-core do you have to hold as many fingers up as there are cores to understand the speed over multiple cores?
There are indeed many algorithms that run well in a parallelized environment. IBM even makes the world's fastest supercomputers that take advantage of this fact.
However, there are many other tasks fit for computers that do not parallelize well. In addition, writing massively parallelized software is often quite HARD. It is far easier to design software for a single CPU running very quickly, than a whole boatload of CPU's running slower. There have in fact been quite a few articles in CS journals lately wondering how on earth software is going to be written for all these new bunch-o-cores CPUs. While it can be done, it is tedious, expensive, and error-prone for all but the most trivial tasks.
SirWired
It DOES run Linux!
...>ahem and other OS's too... like Windows NT...
(mod me down if you must - but I just HAD to...)
Seven Days with Ubuntu Unity
That would only be true if the CPU is able to retire a sustained average of one instruction per clock cycle. SFGate's article makes a raw comparison between chips with different number of cores, threads and other factors, considering only GHz...
5 billion THEORICAL instructions per second just mean nothing.
Anyway, the DSP I'm working on, the TI C6416 (1GHz), claims up to 8 billion instructions/s (5 to 6 can be realistically obtained).
1 (short ton / firkin) = 89.1432354 slugs / keg
How many Libraries of Congress can it index per fortnight?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Lyle: I'm gonna get a NAD T-770 digital decoder with a seventy-watt amp and and Burr Brown D.A.C.'s.
Hansome Rob: [at a loss] Yeah...
Lyle: It's a big stereo. Speakers so loud, they blow women's clothes off.
Handsome Rob: Now you're talking!
Set your phasers on "funky"!
Well the big ones are.
Linux
AIX
and i5/OS
Applications?
DB2, Oracle, SAP, and goodness knows how many super advanced and mega expensive packages for specific industries that the average person never knows about.
In other words it isn't wasted on Office, Vista, and other low end applications.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
It's horses for courses. If you're serving web pages and running database queries from a well tuned database, the Sun Niagara chips are fast and very well suited. They serve the pages a little slower, but can serve many more at a time.
If, on the other hand what you're doing is not easily threaded then IBM probably have the upper hand. Say you're doing some mathematical analysis, where you have to do everything in sequence. IBM's faster processor can complete each stage quicker, moving on to the next part and delivering the result faster than a chip with more threads but slower speed.
Power6 is a big change from Power4 & Power5 series. The key factor is: it gains clock and SMT at the expense of OoOE. In-order execution means its performance is deeply dependent on perfectly tuned compilers.
Other than the lack of out-of-order, on paper it looks pretty strong. Dual core, lots of bandwidth, up to 7 IPC (5 in one thread, 2 in the other), big GHz, voltage & frequency slewing, and yes it has AltiVec.
p.s. No, it would not be good for Macs. POWER chips are all made for big iron.
Apple's development tools continue to generate binaries that support both Intel and PowerPC, and they continue to support the operating system on both architectures, so why not? Yes, I know, there's a difference between POWER and PowerPC, but it's not a big difference -- after all, you run the same version of Linux on an IBM pSeries that you do on a Power Mac.
Apple would have buying leverage against both IBM and Intel by being able to shift portions of their manufacturing from one architecture to another with each model. And they'd have access to some of the fastest processors on earth. Can you imagine one of these things powering Photoshop, or even rendering the next Pixar movie?
Tired of FB/Google censorship? Visit UNCENSORED!
http://en.wikipedia.org/wiki/POWER6 Each core has two integer units, two binary floating-point units, and a decimal floating-point unit, and is capable of two way SMT. The binary floating-point unit incorporates âoemany microarchitectures, logic, circuit, latch and integration techniques to achieve [a] 6-cycle, 13-FO4 pipeline,â according to a company paper.[6] Unlike the servers from IBM's competitors, the POWER6 has hardware support for decimal arithmetic and will include the first decimal floating-point unit integrated in silicon. More than 50 new floating point instructions handle the decimal math and conversions between binary and decimal.[7] This is a feature being added to the processors powering IBM's System z.[8] So it has a 5Ghz clock rate but can actually manage a bit more than 5 Bips peak. But A notable difference from POWER5 is that IBM moved from an out-of-order design to an in-order design, a drastic change which should require software recompilation for top performance. However, the processor still achieves significant performance improvements even with unmodified software, according to the lead engineer on the POWER6 project.[2] Hmmph. I'd bet it's got a really long pipeline to reach that clock speed. The POWER6 has approximately 790 million transistors and 341 mm large fabricated on an 65 nm process. It was released on the 8th June 2007, at speeds of 3.5 GHz, 4.2 GHz and 4.7 GHz[2], but the company has noted prototypes have reached 6 GHz.[3] POWER6 reached first silicon in the middle of 2005[4]. Wow it's huge, almost twice the size of a Core 2 Duo.
I think IBM is doing taking the NetBurst approach - a long pipeline to get to high frequencies. Plus it's a server chip only used in their servers so they can design for a much higher TDP than Intel or AMD and rely on water cooling.
I think this guy is spot on
http://aceshardware.freeforums.org/praising-the-power-6-design-t426.html Later this year Intel will release the 65 nm bulk CMOS Tukwila and
it will likely easily outperform the 65 nm SOI CMOS Power6 on the
benchmarks of most interest to buyers of business critical servers
despite running at less than half its clock frequency and having
less than half its socket level bandwidth. IBM might have created
a better product and closer competitor to Tukwila better if Power6
had been a quad design based on a Power5 core worked over to
improve performance/power but then its wouldn't have the mega-
giga for headlines in the WSJ and given IBM Micro a measure of
bragging rights to help justify its continued existence.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
I always said rating CPU's in Hertz is like rating engines by cubic inches. Bigger *can* get more performance, but it's no guarantee. The compression, carbeuration, transmission, fuel flow, exhaust, all add up to final performance, same as cycles per instruction, the amount of work each instruction can do, the memory bandwidth and the IO system all add up to system performance in a computer.
Karma Whoring for Fun and Profit.
...but it's not my index finger....
He's never had a help desk job....
"We are all geniuses when we dream"
- E.M. Cioran
and I thought he was going to finish that with "and it goes THIS fast!!!", as he waves his finger across his face as fast as he can.
That's how my brother and I used to measure seconds when we were 5 years old. Accurate to within 500% (your mileage may vary).
It's very useful to remember that the speed of light is about a billion feet per second, or a foot in a billionth of a second. He was just looking for a measure that is 1/5 of a foot long.
But how many BogoMIPS is that? After all, that is most computers do, most of the time.
Think about it: This new chip can do nothing up to 2 to 3 times faster than any other chip on the market! We are talking about incredible productivity gains during idle times!
(why yes, I do work in marketing...)
Tequila: It's not just for breakfast anymore!
They're overclocking the POWER6 chip, is that right? Sure, IBM, Cray, DEC et. al. used to do this routinely on big iron back in the day when computer technology was still a science. I still remember seeing a beautiful oil "waterfall" on the front of some mainframes. It wasn't called overclocking back then - it was just how things were done. Now, with computing being a commodity, most companies don't bother with this - too unreliable, too bulky, too power-hungry. Remember, the weakest part of any electronic device is the mechanical aspect and water cooling involves a lot of mechanical processes. You've really got to have a need for speed to bother with this (and, yes, some big environments have such a need - but not many).
For those few environments which need this much speed per processor, this is an important development. Just don't count on it ever impacting the average desktop (commodity) system - the technology won't "trickle down" (unlike the coolant?).
Financial institutions are required by law to perform financial calculations on a computer as they would on paper, so a chip that can do these calculations natively have a built-in market that is willing to pay the extra for the features. This is a special-purpose processor. There have been and will continue to be purpose-built calculating machines, so it's really not fair to say that IBM is simply trying to dazzle us to justify their existence. As long as no one else makes these machines, their existence is justified.