IBM today simultaneously launched the fastest microprocessor ever built and an ultra-powerful new computer server that leverages the chip’s many breakthroughs in energy conservation and virtualization technology. The new server is the first ever to hold all four major benchmark speed records for business and technical performance.
At 4.7 GHz, the dual-core POWER6 processor doubles the speed of the previous generation POWER5™ while using nearly the same amount of electricity to run and cool it. This means customers can use the new processor to either increase their performance by 100 percent or cut their power consumption virtually in half.
IBM’s new 2- to 16-core server also offers three times the performance per core of the HP Superdome machine, based on the key TPC-C benchmark. The processor speed of the POWER6 chip is nearly three times faster than the latest HP Itanium processor that runs HP’s server line. Even more impressive, the processor bandwidth of the POWER6 chip – 300 gigabytes per second — could download the entire iTunes catalog in about 60 seconds – 30 times faster than HP’s Itanium.
But the new server offers more than just raw performance – it is the world’s most powerful midrange consolidation machine, containing special hardware and software that allows it to create many “virtual” servers on a single box.
IBM calculates that 30 SunFire v890s can be consolidated into a single rack of the new IBM machine, saving more than $100,000 per year on energy costs. According to IDC, IBM has gained 10.4 points of UNIX revenue share in the past five years — versus HP’s loss of 5.3 points and Sun’s loss of 1.4 points. IBM will use the new machine to target customers with less-efficient HP, Sun and Dell servers.
Demonstrating its remarkable versatility, the new IBM System p 570, running the POWER6 processor, claims the No.1 spots in the four most widely used performance benchmarks for Unix servers – SPECint2006 (measuring integer-calculating throughput common in business applications), SPECfp2006 (measuring floating point-calculating throughput required for scientific applications), SPECjbb2005 (measuring Java performance in business operations per second) and TPC-C (measuring transaction processing capability). This is the first time that a single system has owned all four categories. The new System p 570 now holds 25 benchmark records across a broad portfolio of business and technical applications.
The performance leadership is largely attributed the system’s balanced design. Unlike competing servers, IBM succeeded in scaling the new server’s processor performance and system design (cache sizes and bandwidth) in a balanced way. The POWER6 chip has a total cache size of 8MB per chip – four times the POWER5 chip – to keep pace with the awesome processor bandwidth. By contrast, many other servers concentrate mainly on processor performance, at the expense of the server’s ability to feed data to the chip at a rate that takes advantage of the processor’s speed.
“Like the victory of IBM’s Deep Blue chess-playing supercomputer 10 years ago this month, the debut of POWER6 processor-based systems proves that relentless innovation brings ‘impossible’ goals within reach,” said Bill Zeitler, senior vice president, IBM Systems and Technology Group, in the press release. “The POWER6 processor forges blazing performance and energy conservation technologies into a single piece of silicon, driving unprecedented business value for our customers.”
The POWER6 chip in the new IBM System p 570 server owns a number of industry “firsts.” It is the first UNIX microprocessor able to calculate decimal floating point arithmetic in hardware. Until now, calculations involving decimal numbers with floating decimal points were done using software. The built-in decimal floating point capability gives tremendous advantage to enterprises running complex tax, financial and ERP programs.
The POWER6 processor is built using IBM’s state-of-the-art 65 nanometer process technology. Coming at a time when some experts have predicted an end to Moore’s Law, which holds that processor speed doubles every 18 months, the IBM breakthrough is driven by a host of technical achievements scored during the five-year research and development effort to develop the POWER6 chip. These include:
• A dramatic improvement in the way instructions are executed inside the chip. IBM scientists increased chip performance by keeping static the number of pipeline stages – the chunks of operations that must be completed in a single cycle of clock time — but making each stage faster, removing unnecessary work and doing more in parallel. As a result, execution time is cut in half or energy consumption is reduced.
• Separating circuits that can’t support low voltage operation onto their own power supply “rails,” allowing IBM to dramatically reduce power for the rest of the chip.
• Voltage/frequency “slewing,” enabling the chip to lower electricity consumption by up to 50 percent, with minimal performance impact.
• A new method of chip design that enables POWER6 to operate at low voltages, allowing the same chip to be used in low power blade environments as well as large, high-performance symmetric multiprocessing machines. The chip has configurable bandwidth, enabling customers to choose maximum performance or minimal cost.
The POWER6 chip includes additional techniques to conserve power and reduce heat generated by POWER6 processor-based servers. Processor clocks can be dynamically turned off when there is no useful work to be done and turned back on when there are instructions to be executed.
Power saving is also realized when the memory is not fully utilized, as power to parts of the memory not being utilized is dynamically turned off and then turned back on when needed. In cases where an over-temperature condition is detected, the POWER6 chip can reduce the rate of instruction execution to remain within an acceptable, user-defined temperature envelope.
Source: http://www-03.ibm.com/press/us/en/presskit/21546.wss
One more time…
The PowerPC processors were based upon a merge of the single chip variant of the original POWER chip set and Motorola’s 88000 chip set (Motorola’s RISC chip of the day). The original versions of the POWER processor had either 5 or 7 separate chips (depending on the variation and capabilities you needed) to make up the full processor. After a couple iterations IBM got it down to a single chip. POWER is/was an acronym for Perfomance Optimized With Enhanced Risc. The original 88000 processor had one central chip and up to four supporting chips with the L-1 cache on them. Apple had a large hand in the tailoring of the chip to optimize it for the Mac System software. This showed up most strongly in the good performance numbers of the PowerPC G3. (The Apple, IBM and Motorola get together on the PowerPC was often referred to as the AIM alliance.)
The instruction set for the original PowerPC (the 601) was almost identical (but *not* identical) to the original POWER processor. What the PowerPC got from the 88000 was the chip’s internal backplane and the inherent design to be able to include additional processing units (e.g., a vector processor) from the very beginning rather than being a “bolted on kludge” like the original SSE implementation from Intel.
Power PC never really stood for anything except that it was derived from the POWER chip, and it was meant for the Personal Computer market. IBM and Motorola (who own the rights to the PowerPC designs) tried to market it to many other computer manufacturers than just themselves and Apple. However the only vendors that shipped significant quantities of anything built upon the PowerPC were based on Macs and Mac Clones, Motorola Unix boxes and IBM Unix boxes.
Motorola had a follow on to the 88000 (loosely named the 88010) but never went into full production on that chip so that side of the PowerPC’s lineage died with the original processor.
IBM went on to build subsequent variants of the POWER chips. The PowerPC G5 was a deviation from the “Book E” specifications agreed to with Motorola and could only loosely really be called a PowerPC. The PowerPC G5 was as much (or even more so) based upon the POWER G4 chip (often referred to as the POWER4) as it was based upon the original POWER and 88000. (Bye-the-bye, Motorola did its on PowerPC G5 variant. Early on it ran into problems with being able to manufacture it defect free and a redesign to cure that problem would have cost several million $$$. The few chips that were tested by various organizations outside Motorola and Apple were increadibly powerful. They were truly amazing. But with the cost to fix the chip for full production Motorola decided to kill the entire Motorola PowerPC line rathe than invest that much money for predominantly one customer: Apple.)
The POWER chips have always pulled more power (no pun intended) than most people would want in their personal computers. These new POWER G6 chips are no different. Most variants do *not* need liquid cooling, but they definitly will never be in a laptop unless you want to wear very thick insulating pants!
Since OS X is based upon Unix (more properly called UNIX) it is indeed *possible* to port it over to this new POWER chip, but it would be very costly to do so. The POWER chips don’t have the same implementation of AltiVec as the PowerPC did. They don’t have the same execution methodology. This would *not* be a trivial port. Therefore I doubt anyone will ever see any version of OS X (even Darwin) running on the POWER G6 chip.
@ Shadowself : Thanks for the summary and analysis. The only argument I would have with it is that Apple is not dependent on AltiVec anymore after offloading the processing to the GPU through Core Image / Core Video.
But is it any good for gaming?
I could respond to this, but it would cut into my masturbation time. So I’m going to go do that instead.
MW: “soon”, as in “I’ll be back soon”.
INTEL sucks!
It’s burning your Macbook, it’s slow, its fan sounds like shit, my photoshop runs like snails on Intel!
@ shadowself – Good recap, but a few possible corrections come by way of a CNET article (found here):
http://news.com.com/IBMs+Power6+Bigger+iron,+lower+power/2100-1006_3-6158739.html
You said: “Since OS X is based upon Unix … it is indeed *possible* to port it over to this new POWER chip, but it would be very costly … The POWER chips don’t have the same implementation of AltiVec as the PowerPC did. They don’t have the same execution methodology. This would *not* be a trivial port. Therefore I doubt anyone will ever see any version of OS X (even Darwin) running on the POWER G6 chip.”
From CNET: “For blade servers, where power consumption is a particularly acute issue, … IBM previously used a separate chip design …, the PowerPC 970 series, but that family was merged with the mainstream Power chips with the Power6 generation.”
PowerPC 970 was/is the G5. IBM has actually incorporated a great deal of 970/G5 design tweaks into the POWER6 – particularly Altivec (this was not on any previous PWR chip), but most things that made G5 ‘Macified’ as well. Therefore, I do not think it’s reasonable to suspect a massive or costly project would be required to get even the next version of OSX, let alone the present one, to run on it.
You said: “The POWER chips have always pulled more power … than most people would want in their personal computers. These new POWER G6 chips are no different. Most variants do *not* need liquid cooling, but they definitly will never be in a laptop unless you want to wear very thick insulating pants!”
This isn’t exactly correct. The PWR chips have always been extremely power efficient when compared to their competitors, especially the PWR5 – that’s what made them such popular choices for supercomputing clusters particularly. Even the 970/G5 was an efficiency champ. The makers of both the VATech and Barcelona blade-based supercomputers chose 970/G5 precisely b/c of that, as well as their low unit prices. However, the original .130 nm 970/G5 did have a problem with heat – specifically, the die was smaller than typical, and the heat generated by the chip was very concentrated and thus difficult to dissipate. And the .90nm version didn’t clock up as well as predicted (an industry-wide problem), which required pushing thermal design limits to the wall to keep performance advancements (and marketing plans) on track. Apple dealt with this via watercooling and clocking up regardless, and get perfect pace with the similarly clocked Athlons & Opterons from AMD. Intel’s Netburst-based Pentiums were going balls to the wall, but they were inferior perfomers too. Nevertheless, the 3ghz thing stuck in Jobs’ craw & the Macintel switch was announced, Yet as soon as they announced it, IBM debuted the FX version of the 970/G5 a month later which largely solved the problem. The last G5 iMacs ran a single core version of that CPU and were not known for power consumption or heat problems – in fact, the MoBo used in the last G5 was the actual design for the now mythical (but in fact real) PowerBook G5 that became irrelevant once Intel became Apple’s entire platform supplier/designer.
Anyway, the new PWR6 is an efficiency and heat champ. It is in no way saddled with the original 970’s problems. From CNET:
“Overall, IBM promises performance double that of Power5+ without exceeding the earlier chip’s power budget (which was even better than 970/G5). Power consumption and waste heat are an increasing problem for chip and computer designers, so… Power6 employs several power-saving techniques. It has a low-power idle mode called a “nap” that can cut power consumption by 30 percent to 35 percent when a server’s operating system is also idle, … even when the operating system was busy with a stress test, napping cut power consumption 10 percent… Another power-saving technique dynamically adjusts processor frequency and voltage, lowering both when possible to cut power … And if there’s room within power limits, the chip can run faster [than spec’d]… IBM can accommodate higher temperatures and therefore higher clock frequencies [with PWR6, showing] … a chart in which processors ran as fast as about 5.8GHz without exceeding 85 degrees Celsius.”
The article doesn’t go into any of the materials advancement IBM has incorporated with this chip that also increases power & thermal efficiencies, but that last bit about the temp is pretty telling. Even if optimistic, that’s a number well within the range of the 1.33Ghz iBook G4 I’m running right now, with roughly 10X’s the performance. None of Intel’s Core2Duo chips, as good as they are, are close to matching that mix at this time.
MDN Magic Word = strong … I’d say that’s an apt description for this chip
” width=”19″ height=”19″ alt=”cool smile” style=”border:0;” />
Not my post at 6:12.
Why couldn’t this come out three years ago?
Because it’s typical IBM. Too little too late.
Now let’s see if IBM can:
1. Actually SHIP these processors and servers before they’re obsolete.
2. Avoid permanently alienating everyone on the backorder list.
3. Deliver enough of these processors and servers to matter.
meh… IBM can’t ship stuff in volume, and they can’t put out decent laptop CPUs. The best they ever do is to pull ahead by a tiny bit for a very short period of time, once every 7 years or so if they’re lucky.
Macintosh would be dead by now without Intel. The Core 2 Duo easily beats the G5, it absolutely crushes the G4, it gets frequent updates and it’s cheap. If AMD somehow pulls ahead of Intel, Apple can switch to AMD chips. Competition is a beautiful thing.
I have to agree with wabewalker and AP–this POWER6, in just reading the short blurb here implied that this is only for places crunching the entire human genome or the entire US GDP and has to have it done before lunch. This is just a little “show and blow” time–notice the timing too, on the eve of WWDC from apple–a typical “FU” to Jobs et al. And AP is right, unless you invest in your own cooling refrigerant building (at least a block long) this thing is nowhere near–and never will be for home use.
Some in the infinite loop there is a Mac computer running the Mac OS X on Power6! <EOL>
If this processor turn out to be a good one, is there a reason why Apple couldn’t make a PPC-mac using that Power6? After all almost every Mac software are now UNIVERSAL …
Probably not impossible to port, but it would only be worth it if: a) Apple chooses to make an exponentially larger push into the big iron server market, and b) IBM has manufacturing capacity for the Power6 sufficient to make them interested in selling to other manufacturers rather than just keeping that business for itself.
I don’t see either scenario being realistic. Apple will probably buy a machine from IBM and do a “skunk works” version of the OS for it, but I don’t see it going further than that.
Wow man – so much ignorance in such a little space, with only a few signs of any actual knowledge.
” width=”19″ height=”19″ alt=”grin” style=”border:0;” />
The fact that Mac OS X 10.5 is UNIX certified does not magically mean that it can be readily ported to any processor architecture Apple feels like. All it means is that the OS provides an API that is compatible with the UNIX standard – it means that *applications* written to that standard can be readily ported to run on computers running Mac OS X. These applications are command line tools with no user interface.
This capability is of no assistance whatsoever in bringing Mac OS X to different processor architectures.
Apple has custom code within their XNU kernel (which is at the heart of Mac OS X) that are specifically written for PowerPC, Intel, and now ARM too (for the iPhone). It is not UNIX that helped them support ARM, but the modular way in which their kernel is written (as well as the rest of the OS), where the parts that need to be processor architecture specfic are kept to a minimum.
As for speculation about supporting POWER6, well, the architecture is related to PowerPC, and the chips support a similar instruction set, so OS-level support could be derived from the existing PowerPC code.
The differences between POWER and PowerPC would most likely dictate that you could not run a PowerPC kernel on a POWER chip. Similarly applications compiled for PowerPC would most likely not be compatible with POWER and would need to be recompiled to work. A Rosetta emulator could of course be written to allow POWER to run PowerPC applications.
@ AP You are correct circa 2003-2005. IBM at the time did not or appeared not to have a direction in which to focus their developement, hence Apple leaving as they did.
The wake up call came when Apple actually left. They proceeded to sell of all the divisions that were not profitable enough and spent the years to date on projects that would be of interest to the big bad boys with the muscle to use their products.
The interesting thing that I believe is being missed out is the fact this processor or chip is capable of multiple processing at a level hitherto unachieved. The maximum usage of the chip would result in the scenario you describe, on the other hand if it were not in full capacity mode it has the capability of switching off bits not in use thereby reducing energy consumption & heat output.
As with all new technologies there is a gap between the product first being announced, usage defined and applications discovered or derived to put the product into use.
The iphone announcement being a classic case of Apple sourcing a market, advertising & user demand for features from a product only just revealed to the World long before it has been refined and mass manufactured.
IBM Thinking Different?
You’re the last post, Crabman.
TMF: “You’re the last post, Crabman.”
oh, the delicious irony of that post.
This thread is classic example of Mac sheep being no different than the Windows sheep they mock. Quite humorous.
As some have pointed out before, this is NOT a desktop chip and never will be in the near future. Even if Apple kept IBM as a chip provider, POWER 6 would not find it’s way to a Mac.
Any references to a “G6” or potential Mac desktop equivalents shows just how ignorant some of you are.
Exposed: We are here only for your amusement. Nothing else.