Slashdot Mirror
IBM's New Processors To Exceed 5Ghz
Jordin Normisky writes to mention the news, via ZDNet Asia, that IBM's new Power6 processor will be unveiled next month at a conference in San Francisco. They're also planning to announce a second-generation Cell, both of which are expected to run faster than 5GHz. From the article: "In addition, the [Power6] chip 'consumes under 100 watts in power-sensitive applications,' a power range comparable to mainstream 95-watt AMD Opteron chips and 80-watt Intel Xeon chips. Power6 has 700 million transistors and measures 341 square millimeters, according to the program. The smaller that a chip's surface area is, the more that can be carved out of a single silicon wafer, reducing per-chip manufacturing costs and therefore making a computer more competitive. Power6, like the second-generation Cell, is built with a manufacturing process with 65-nanometer circuitry elements, letting more electronics be squeezed onto a given surface area. "
I thought we had finally advanced past the "higher clockspeed = more better" stage...
- Toby
I think that apple will move back to these processors in it's Mac Pro's computers. THey might have signed a contract with Intel forbidding them to do this, but I think that it would be more advantageous and use whatever processor is the best at the moment.
Any EE care to chime in with how IBM might have gone about solving their power consumption issues?
It's seemed that IBM, Intel, and AMD have all had issues dealing with leakage moving to the 65nm process, but Intel and IBM have now apparently solved the issue. As a mere hobbyist, I don't even approach the knowledge level necessary to speculate on such things, but I STILL WANT TO KNOW.
Usually from the bell-end of Apple. I wonder if IBM's fab plants can cash the check their PR department writes.
Do you even lift?
These aren't the 'roids you're looking for.
If I owned an as/400, i/z series server, maybe this would be exciting news as a future upgrade.
But do they achieve a comparable amount of work per cycle?
--
Wi-Fizzle Research
Censorship is obscene. Patriotism is bigotry. Faith is a vice. Slashdot 2.0 sucks.
The smaller that a chip's surface area is, the more that can be carved out of a single silicon wafer, reducing per-chip manufacturing costs and therefore making a computer more competitive. Power6, like the second-generation Cell, is built with a manufacturing process with 65-nanometer circuitry elements, letting more electronics be squeezed onto a given surface area.
The cost of making chips, by far, is the R&D cost. The "first" chip costs hundreds of millions to make. Once the "first chip" is made the margin cost is VERY low. Beyond recovering R&D costs....the rest is just distribution channel costs....then....PROFIT!You know, I've never been happy with Apple's decision to start using Intel's offerings. I don't care if Intel has a roadmap for putting over 100 cores in a chip. IBM has always made a superior chip. If only IBM and Motorola didn't have such dismal shipping problems, maybe this wouldn't be the reality we now know.
No sig for you! Come back one year!
They would get bragging rights with 45nm. 65nm is so old that even AMD has 65nm chips now.
Heck philips/motorola I believe have been producing 65nm microcontrollers, and samsung is producing 50nm flash chips.
And 5GHz should not be difficult considering it doesnt have the x86 overhead, is more RISC and that generally PPC has a simpler core. I'll be interested if it comes with quad cores or more.
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
...undisturbed 802.11a networking!
Tsunami -- You can't bring a good wave down!
They're also planning to announce a second-generation Cell, both of which are expected to run faster than 5GHz.
Why don't they seem to be making any kind of performance comparisons? Talking about physical size, power consumption as compared to intel & AMD are great, but it seems weird that there's no mention of real-world performance against those same competitors. Even a rough estimate would be interesting.
The theory of relativity doesn't work right in Arkansas.
Apple has never used the Power architecture in Macs. It used PowerPC, which isn't the same.
Maybe the BogoMIPS value for it is too large for linux_logo to process (16-bit number if i am right)
WulframII - Free Online Mutiplayer 3D Tank Shooting Game
There should be some really interesting stuff this year on how they kept the power down.
Of course, a chip nearly 2 cm on a side is going to be a beast no matter what. This is going to be fun!
Lacking <sarcasm> tags,
In your face, Steve Jobs!
Boy, Howdy! are you out of the loop. I work on those suckers and believe you me, the chip cost is not trivial.
Do the math: the cost of a 300 mm wafer in a 65 nm process runs well over $5000 (how much is a Deep Dark Secret.) Ignoring geometric yield loss, that's about 70,000 mm of potential dice per. If one chip is 350 square mm, you're getting about 200 per wafer, or $25 per chip fab cost. Yield drops off steeply with size (think in terms of losing ten to twenty dice per wafer, regardless of die size) and that adds into the fab cost too.
That's bare minimum, assuming there aren't any bad lots etc. It adds up fast.
Lacking <sarcasm> tags,
It would be ludicrous, but Kutaragi's talked before about never reducing the price of the PS3 but instead upgrading it with more memory, bigger hard drives, etc. It would be pretty damned amusing if, a year and a half after PS3 launch, instead of cutting prices with a new easier to produce Cell and Blu-ray they upgraded the PS3 with the Cell2(and hosed everyone who'd already bought one). This would be so stupid and arrogant that it's only plausible because it's Sony.
Apple has been doing this for years
So what if IBM's new chips run at 5GHz or more? What about gigaflops?
I had hoped the majority of slashdotters would be able to see past the megahertz myth by now.
Apparently not.
Comment removed based on user account deletion
Power5+ cpus have been king of the tpc hill since they were first introduced. Power6 will only increase that lead i suspect.
pSeries, iSeries, and zSeries, are still hard at work doing same they have always done, running banks, distribution centers, and the like. The difference is that mini's and mainframes don't need glossy magazines so that people know they get work done, they just do it.
If you look at the direction AMD is going you will see the archietecture so common in the mini/mainframe areana is coming down to the home.
It was always hilarious to hear the network guys brag about their 4-way network tower with its 8gb plus of ram, they loved to lord those numbers over the iseries folks, too bad that the 512mb uniprocessor iseries box served more people. It took a while to educate the pc-nuts (and I am a pc nut as much as an i/p/zseries nut) that processor and memory do not make a computer. It is design and integration that makes a computer.
plus having a real OS doesn't hurt either.
Besides IBM never needed Apple's marketshare, they conquered the world of home entertainment without ever exposing themselves to much risk
* Winners compare their achievements to their goals, losers compare theirs to that of others.
In the world of technology a promise of more/better performance counts as much as a drunken "I love you." One reason why Apple jumped from PPC is that IBM failed to deliver a 3.0 Ghz chip within a reasonable time frame (in the PPC970 series) and completely failed on delivering a laptop chip. Believe it when you actually see shipping servers.
Leave the gun, take the cannoli -- Clemenza, The Godfather
Anyone have an answer?
This also puts Apple in a good situation, AT NO POINT do they have computers that are inferior to their competition. Before, if Motorola or IBM outmatched Intel, Apple had bragging rights, if Intel beat them, then they were at a disadvantage. Now, if there is a supply issue in x86 land, then Dell, HP, and Apple are all in the same boat. Apple now competes on its software, not on Motorola/IBM's interest in beating Intel.
At PPC was often a disadvantage and only occasionally an advantage for Apple, they get chips out of the equation all together, and now fight on software, a much better boat to be in for them. They have the same suppliers as HP and Dell, a decent size economies of scale situation (they are the 4th or 5th biggest hardware manufacturer, so while Dell and HP are bigger, they are still a HUGELY lucrative account), and have their software advantage over other PC makers. The different CPU issue just confused the matter...
They'd have had a BIG edge if they went with x86-64 off the bat, and never had any legacy x86-32 code to deal with, giving a performance edge over windows, but I just don't understand migrating to x86 when x86-32 only had 6 months of life left in it instead of waiting for x86-64 and having a performance edge for 5-10 years, but that's just me.
There are clock speeds and there are operations. I know what an operation is, but how are cpu clockspeeds rated? Is it just something as silly as their clock source? By defination it is "cycles per second", but what exactly is cycling? I've always been confused by this and I think I just don't understand how digital processors work enough.
If an officer ever threatens to taze you, say you have a pacemaker.
If not, who cares? Let the flames begin.
I did a search with my threshold turned down to 1, and no one asked the most important question yet. But does it run Linux?
Experience teaches only the teachable. -AH
The biggest effect the Intel switch has had is to put a stake in the heart of the horrid old OS 9 vampire. I'm more than half convinced that the reason Jobs timed the switch when he did was because he'd just - about six or so months earlier - been able to pull the last G4 Powermac that could boot into OS 9 off the Apple store without the usual storm of protests. With the Intel switch, the new Macs don't include Classic and won't even run the old OS 9 software.
Even if IBM had a 5 GHz quad-core Power PC that was plug-compatible with the G5 *right now*, Apple wouldn't ship it. The last thing he wants is to have some beggar pull that stake out of Classic's ribcage and have it come lurching back to life.
Apple "Switched" not because Intel was faster or better or had a "better roadmap" as they claim. It's because IBM couldnt get the heat down on the chips. They had no G5 PowerBook and the towers had to be made like giant wind tunnles.
IBM just couldnt make a cool + powerful chip like Intel could.. but.. that looks like thats in the past now..
Bringing liberty to the masses. - http://freetalklive.com/
IBM does not give a heck to Desktop market unless you are calling them about 10.000 terminal running Enterprise Big Iron monster and they may even suggest you buy Dell terminals/PCs if it fits their project better. What matters to them is the mainframe, technologies used, software used and the entire consulting to keep such business up.
Motorola/Freescale lives happily in embedded processor market and telecoms market too.
I guess such stories should have "power-not-powerPC department" tag.
Also, yes , our great leader/prophet whatever was right switching to Intel/x86 because of above reasons. Both companies tries to stay away from Desktop market and they won't be bothered by ridiculous 3Ghz PPC G5 (a STRIPPED DOWN POWER4) Apple fanboys. Apple can't effect those decisions by their current market share. If it goes back to great 50% 50% marketshare values, they can demand anything of course.
(Happily written from a 33C/92F running Quad G5)
The chips are already in production, this is the very end of the cycle on these. Keep in mind that while many seem to relate this to PPC, this is really the POWER line targeting servers and IBM has been traditionally pretty accurate with their statements regarding POWER4, POWER5, POWER5+ in the past.
Low power consumption, PPC, fast. Can we have this in a laptop please?
I for one welcome our new 6 Gigahertz masters.
Have Tardis, will travel.
Yup. And your point is ... ?
And the next million wafers cost how much?
Nobody builds stuff like that with a run rate of a few measly thousand. No way to recover the NRE. If the variable costs don't dominate the bugetary numbers, then the project doesn't get authorized.
By the way:
That's just the ante to get in the game.
Unless you're doing some massive reuse, that's not even a pimple on the actual project cost. A major CPU design team runs to the hundreds of engineers. My team, which is doing microcontrollers, is about twenty.
Is way, way out of date. 65 nm mask sets for SOI run well over a million. Last I looked, much closer to two million.
Lacking <sarcasm> tags,
The original IBM PC clocked at 4.7 MHz.
Now they can go 4.7 GHz.
About frickin' time.
My other car is a 1984 Nark Avenger.
4 core POWER5+ pServer running Linux is $5,500
${YEAR+1} is going to be the year of Linux on the desktop!
The reason why Intel/AMD had to reduce speeds was that they were achieving the greater speeds through very long pipelines - which is fine, but from an instruction point of view, it has to wait through the pipeline before it will get executed. IBM have kept their pipeline at the same length as the previous generation chips and have increased the speed, in part, through a combination of 65nm and 90nm parts - so in this case it is like taking the Core 2 and increasing its clock speed beyond the NetBurst Pentium: The performance is multiplied.
There's a reason IBM supercomputers number the most in the most powerful computers in the world beyond any other manufacturer. Check out top500.org.
I haven't checked the information yet, but here's an abstract on the rest, found through google:
c essor-Trashes-Competition-with-6-GHz-17765.shtml1 606194731
The Power6 processor will run between 4GHz and 5GHz and it has been proven to chew away data at a speed of 6GHz in the lab.
IBM see things a little differently and they decided to raise the frequency in both cores of the processor.
For high-end models, four POWER6 MPUs will be packaged in a single multi-chip module, along with four L3 victim caches, each 32MB.
On the management side, IBM is also improving their virtualization capabilities in the POWER6. In particular products, a single processor may be able to host 2-300 virtual instances, although theoretically up to 1024 VMs are possible. Memory partitioning and migration have been added as well, which reduces system down time for repairs.
IBM is claiming a factor of two performance increase, which would be consistent with the vastly higher clockspeeds and increases in raw system bandwidth.
IBM's roadmaps currently include the POWER6+, which is presumably a 45nm derivative product. Judging by past practices, the POWER6+ will debut in the second half of 2008, probably just in time to dash the hopes of rivals.
The Power and PowerPC lines will grow one step closer together with Power6, which incorporates the AltiVec instruction set that speeds up many multimedia tasks. AltiVec, also known as VMX, increases efficiency by letting a single processing instruction be applied to multiple data elements. That's helpful for video and audio tasks on desktop machines, but servers will benefit as well in, for example, high-performance computing tasks such as genetic data processing, McCredie said
Where Power5 can transfer data on and off the chip at a rate of 150 gigabytes per second, Power6 can do so at 300GBps, McCredie said.
Oh, and it is also good for BCD's (binary coded decimals) which obviously points to the expected customers (high end financial firms, presumably).
Sources:
http://news.softpedia.com/news/New-Power6-IBM-Pro
http://realworldtech.com/page.cfm?ArticleID=RWT10
http://news.zdnet.com/2100-9584_22-6124451.html
You can find it in the "clocking" session from ISSCC two years back. Slick stuff, doesn't use as much power and gets much less clock skew than a clock tree does.
Lacking <sarcasm> tags,
This Power6 chip the one they were talking about in the movie Hackers? http://www.imdb.com/title/tt0113243/
Now that was a movie way ahead of its time!
"It has a killer refresh rate.
P6 chip. Triple the speed of the Pentium.
Yeah. It's not just the chip, it has a PCI bus. But you knew that.
Indeed. RISC architecture is gonna change everything.
Yeah. RISC is good."
Yeah, they'd have taken a hit for 6 months, especially in the laptop arena. I was still sitting on a 1 Ghz G4 Powerbook, because replacing the machine for a 1.42 Ghz G4 Powerbook seemed like a pointless upgrade. It didn't look like a low power G5 (or 2 Ghz G4) was in the works, and Apple would have had an AWFUL 1H2006. However, there was an expectation of a week 1H2006 because of the transition.
I think that the long term edge of having ALL software running on x86-64 instead of x86-32 would have given them a performance edge over Windows for 3-5 years, and should have been considered. Supporting 32 and 64 bit PPC is different, they don't. They support PPC-32, with some libraries for mathematical software in PPC-64. The core OS is PPC-32, and that will never change.
I think that making the OS, and all apps x86-64 would be nice.
Higher clockspeed is a VERY good thing. While parallel execution units/multiple cores is nice for doing several things at once, sometime you just want the single application you are using to get done faster. So you can either run the clock faster or make it accomplish more per clock tick.
I gues maybe if they had a gazillion parallel cores or execution pipelines they could try to guess the most likely actions you are going to ask the CPU to do before you actaully ask it and calculate the reults and then guess the next one after that, etc (kind of like calculating every possible move and counter move in a chess game), but I don't think we'll have enough parallel processing capacity to do that very "deep" for a while (although I think most modern CPU's do a little bit of predictive work, but nothing on that scale.)
I personally think Netburst was a very good idea (optimize the CPU to make it easier to clock to "ludicrous speed"), but they simply ran into some unexepcted barriers and had to go back to a different architecture until they figured a way past those issues. Maybe the next Intel (core 3?) architecture can incorporate all the cool features that let core 2 do more per clock AND see significant clock speed increases.
Could it complete "sleep 1" any sooner? no? sigh.
I'm still personally waiting for the day when we judge processors by how many instructions per cycle they can process. Hey, a 1Hz processor that does 5,000,000 instructions per cycle *SHOULD* be on the equivalent of a 10Hz processor that does 500,000 instructions per cycle, right? (I'm hoping I got the basic math down, here.)
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Perhaps IBM is trying to win back Apple with this announcement? I'm half-joking. Somethign tells me that this is rather well-timed to compete with Intel. "Look! You switched just as we put out something new! Look at the power in high-end servers! You could have THIS in your desktop or laptop!" Of course, I'm probably dreaming.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Here's your answer: yes
:-)
Insert
Motorola hasn't made ANY chips since 2004 when they spun off their semiconductor products group and it became Freescale Semiconductor.
Only one poster (poopdeville) in this whole thread seems to have gotten this correct.
If you're talking about chips, and you think of Motorola, you should really be thinking of Freescale now.
5 Ghz is enough speed for everyone.
My understanding is that the current-generation "CISC" chips have a preprocessor (on the die) that turns some of the more complex instructions into several less complex ones, which are what the processor's logic pipelines work on. Thus while the compiler might still produce CISC microcode, it's really being cross-compiled one further time before it's actually executed. So you can sort of argue it either way. The processor as a unit does execute more instructions than RISC chips; but in a way they're not executing them natively.
The main reason I've heard why CISC+preprocessor designs won out over pure RISC is because the die area required for the CISC-to-RISC pre-processor shrunk steadily as new manufacturing processes were developed. The logic required to translate CISC to RISC is basically fixed, and as you go to smaller and smaller processes, it represents a smaller and smaller portion of the total processor die area. Thus, every year the advantage of RISC over CISC shrinks. It's sort of an unintended consequence of Moore's law: the ability to pack more transistors onto a die has also made radical changes to the architecture's less attractive, since it makes preprocessor logic "cheaper" in terms of die space.
The x86 instruction set has become the lingua franca of processors, even though most of them immediately break those x86 instructions down into RISC-ish sub-instructions for actual processing. Whether you can still call this hybrid architecture CISC or RISC is mostly semantics; it's a little of both.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."