Intel: we will ‘blow away’ any power advantage of potential G5 CPU

“Rumor has it that Apple is a potential customer for the chip because it wants to replace the Motorola PowerPC chips in its Power Mac desktop computers. If IBM were to market the chip not only to Apple, but also to Windows PC OEMs, it would be a revolution: a 64-bit chip that could scale from desktops to the high end of mid-range Wintel-class servers…

Thomas R. Halfhill, senior chip analyst with InStat/MDR, told NewsFactor that although IBM’s new chip seems ‘tailor made’ for professional publishing, he does not expect it to extend the company’s presence in the PC processor market. ‘I don’t see it being a breakthrough product,’ he said. ‘The market has standardized around the [Intel] x86 instruction set. There’s too much new design work to move to another instruction set, and I just can’t see Dell or HP doing that.’

In fact, IBM concedes that 64-bit computing power may be overkill for today’s desktops. Chekib Akrout, vice president of PowerPC and networking technology development at Big Blue, told NewsFactor that the company’s real target is servers running Xeon and, in the future, Itanium 2 chips. Intel’s chips run about 87 percent of today’s servers, according to Gartner Dataquest.

According to Akrout, developers struggling with the total power budget for racks of multiprocessor Web and database servers will be attracted to the 970’s power savings: It is expected to consume just tens of watts, compared with more than 100 watts for the next version of Itanium 2, code-named Madison.

‘If you’re going to compete on raw performance, you’re going to be fighting this installed base of applications,’ Akrout explained. ‘But when you find a technological barrier where designers really have to think differently, then I think you’ve got a chance to catch their attention.’

Performance is a moving target, however, and Intel contends it will blow away any power advantage the 970 currently holds with its next Itanium, dubbed ‘Deerfield,’ to be previewed in February at its developer forum. Moreover, Intel spokesperson Bill Kircos says the business relationships behind Itanium are unstoppable. ‘We’ve got 20 OEMs signed on to Itanium, and seven different operating systems running it,’ he told NewsFactor. ‘I think it’s going to be very difficult to overcome that lead,” reports Tiernan Ray for the NewsFactor Network here.

Supreme confidence on Intel’s part or is that a whiff of fear we smell?


  1. Sorry, but it’s likely supreme confidence, not fear.

    I may not like Windows, but I have nothing but respect for Intel. Years of “RISC is better” and “the x86 architecture is dead” statements have been proven wrong by Intel’s continued innovation and excellence in chip design and marketing.

    IBM’s new chip may be compelling, but not enough to harm Intel.

  2. The Ghz battles are pretty much over and done. It’s how they are being used that counts, and the overall speed and security of the server system as a whole entity. The new PowerPC (G5?) in Apple’s current Xserve system could make it much more attractive than it already is. A high-performing ( UNIX system for $4000 that INCLUDES the server software w/ UNLIMITED user license, high reliability, simple yet powerful admin software, and NOT prone to the viral attacks as recently seen, makes it a VERY attractive option. Intel can make the fastest engine in the world, but it’s useless if the vehicle that uses it constantly breaks down.

  3. Of those 20 OEMs and 7 operating systems, how many are actually in production? There are not a lot of companies selling Itanium machines right now. And as to the OSes, I remeber there being a lot of killed projects.

  4. It’s probably a bit of both. I have no doubt that Intel will continue to ship the majority of chips but with AMD no longer slavishly following in Intel’s lead and the PPC making the switch to a fast 64 bit chip from a company that is executing well, we’re likely to see an erosion down into the 60-70% range from the current 85%+ dominance. That’s enough systems shipping in the “not from Intel” category that more and more people will go multi-platform to chase the non-intel systems.

    If Apple merely goes from their current 0.2% of servers shipped to 5% they’ve gone a long way towards repositioning themselves as relevant for the enterprise and Intel ends up having one more competitor to worry about as realistic competition.

    That all being said, there is a lot of inertia in IT. It’s likely that the majority of systems will continue to be Intel based for at least the next few generations. Intel needs to continue to execute flawlessly to maintain dominance. Monocultures were in for quite a while but having variety in your computing ecosystem seems to be making a comeback with the lowered susceptibility to various computer attacks.

  5. I reply here more to Troy Curtis than to the original article above. From the beginning, the 8086/88 architecture was a kludge and not nearly as elegant a design as the 68000 chip from Motorola. And while it’s true that Motorola (once a shining example of a forward-thinking company) has been mismanaged in recent years, much of Intel’s “innovation” has come from the cannibalization of technologies from HP and the former DEC (and its Alpha chip). In fact, if anything, it’s an indictment of Intel’s LACK of innovation that it had to partner with HP in order to develop a 64-bit chip–one which HP has sinced backed away from to a degree. (Remember that the first Itanium was an embarrassment to Intel and was not even as fast by some measures as its Pentium 4 chip.)

    Intel no more innovates than Microsoft does–and in fact they both have “borrowed” from the same company–DEC. Just as Intel could not make the leap from 32-bit to 64 without help from east-coast engineers, neither could Microsoft go from 16-bit Windows to the 32-bit Windows NT–except that they essentially appropriated the Mica project from DEC.

    But I realize this is a hot-button topic and somewhat a tangential to this thread. So let me just close by echoing the sentiments of others here: Intel’s dominance is a fact of life–just as is that of Microsoft. But dominance does not necessarily imply merit. If it did, then we would have to conclude that Chevys are very high quality cars. Meanwhile, time will tell whether IBM’s chips represent a worthy alternative (from a technology point of view). I’m inclined to think they will.

  6. If the WINTEL family wants to stay the course, fine. If Apple sticks with Motorola, they are going to die a slow death. Leapfrogging the market with the 970 series is just the jolt Apple needs to recapture market share and become viable again for many potential users.
    With a proper OS & software in place, all that is needed now is hardware to match. The economic recession & tech depression have left a lot of old equipment and software in the government, business & education markets. The orders have not been let yet, but the window of opportunity is getting ready to open. Mr Jobs, I hope you are ready, this is your chance to get Apple out of it’s “niche”.

  7. It�s not just war against Wintel. There�s no reason to believe in Intel innovation but integration around existent devices. I want to give a big question to Intel, nothing else.

  8. Itanium, a failure until now, will be pushed hard by companies like HP and Dell to the server market where superior and tested platforms such as UltraSparc and PPC are available. It would be far more interesting to see a Motorola G5 or IBM PPC970 going to the low end market than Itanium. Frankly, AMD’s Hammer has much more potential of impact at the low/mid level than Intel’s offering but x86 architechure is no much to PPC in terms of elegance of design, efficiency and scalability.
    Guess what ? x86 will again prevail (in the low end market) and the world will be full of incompetent products. That will not stop thinking people to select better solutions. There is no fear of “minority” products and commnunities eclipsing and many companies can be relatively successful and profitable without HAVING to own 85% of the market.

  9. Im more inclined to agree that it is arrogance than fear, the reality is that Intel will most likely overcome the issues (shear weight of R&D investment) it has with the Itanium to be very competitive with anything IBM can come up with. Hopefully the IBM/AMD deal means that AMD will be responsible for production and development from IBM (best in the world). Also, Apple MUST release the 970 in the PM’s this year or it can kiss its mid range desktop (design, publishing and video) market goodby.

  10. Intel this time is truly in trouble and they know it. When the G3 trounced the Pentium, they ignored it. They aren’t ignoring the 970. Nor will they when the Far East manufacturers are stuffing multiple 970s in inexpensive servers running Linux that sell like hotcakes. Dell and HP can push the Itanic all they want, but in order to compete, they will likely have to sell Itanium based hardware at losses. One thing is for sure, people are salivating over the 970. Others do the same for the Opteron. Virtually no one does for the Itanium, leaving companies like Intel and HP beating the drum. I am sorry, but there aren’t many folks who see the Itanium as a must have chip. The 970 was needed yesterday. The Opteron today. The Itanium, well, it would never have been missed. That is precisely Intel’s problem. It isn’t a must have chip for any environment. At least IBM recognized this regarding the Power4 and scaled it back into the 970. I am sorry, but just because Intel says we need the Itanium chip doesn’t make it so. It breaks backward compatibility with the x86 instruction set, consumes far too much energy, and is too costly to manufacture. Absolute speed is not the only factor motivating purchases. Even then, the Itanic will lose to the Opteron in running x86 instructions. 100 watts of power consumption. Yikes!!! I don’t want to be in a room with the fans blowing on one of these machines! Perhaps Intel might invest in a refrigeration company. To think of multiprocessor machines with the Itanium is downright ludicrous. Intel needs to lay this chip to rest. When Dell (they aren’t stupid, after all, they are selling iPods) decides to produce a 970 based machine, Intel’s days will be numbered. HP on the other hand, will ride the Itanium to their ultimate demise.

  11. Now that IBM is pushing Linux as the recommended OS for their server products, RISC chips such as the 970 are the competition to Intel’s Itanium. I guess if a customer specifically seeks a Microsoft based solution they would buy from Intel but if they seek Linux then IBM is a good option.

  12. ROFL, Intel just wants to create a monopoly and make money. I haven’t seen a drop of innovation or excellence from them. Do you work for them or something? ROFL.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.