The company announced the departure just after the close of trading on the New York Stock Exchange. CEO Rory Read will handle Ghilardi’s sales responsibilities in the interim while the search for a replacement gets underway.

Ghilardi (pictured) joined AMD in 2008 as senior VP and general manager for Europe, Middle East and Africa. The following year, he was promoted to chief sales officer.

Before AMD, he worked in Hewlett-Packard’s consumer group, running sales operations of PCs and printers in the EMEA region. And before that, he did the same thing for HP’s commercial business, also in EMEA. He joined HP in 1982, and served on the board of directors of HP Italy (I didn’t know there was such a thing!) from 2001 to 2008. Ghilardi holds a master’s degree in electronic engineering from Italy’s Politecnico di Torino.

It’s hard to know what kind of signal this move is intended to send for AMD. Last week, in a meeting with financial analysts, Read confounded the audience with statements that AMD — whose chips are fundamentally similar to those of Intel and use the same x86 technology — might pursue an “ambidextrous” strategy and incorporate technology used in chips from other companies. This has generally been interpreted as a hint that AMD might incorporate designs from the British chip outfit ARM Holdings in some future hybrid design.

ARM-based chips, as you probably know, are in most of the world’s smartphones and tablets, and for AMD to embrace that technology would be a huge shift in strategic thinking. And by huge, I mean Republicans-embracing-Socialism huge. It wasn’t so long ago that AMD was actively sponsoring a campaign to get x86 chips everywhere, including phones, though it obviously hasn’t worked out that way.

Since then, Read hasn’t really elaborated on what he meant, other than to say that future AMD chips might have a modular design that would allow bits of other types of chips, including an ARM core, to be added. Clearly some big changes are afoot at AMD.

AMD’s statement on the departure is below:

AMD Announces Departure of Emilio Ghilardi as Senior Vice President and Chief Sales Officer

AMD President and CEO Rory Read to Serve as Interim Chief Sales Officer

SUNNYVALE, CA–(Marketwire -02/07/12)- AMD (NYSE: AMD – News) today announced the departure of Emilio Ghilardi as senior vice president and chief sales officer, effective immediately. Rory Read, AMD president and chief executive officer, will serve as interim chief sales officer while the company actively seeks a replacement.

“I’d like to thank Emilio for his contributions to the business and wish him well in his future endeavors,” said Read. “Developing relationships with our customers that are grounded in a foundation of trust through consistently delivering on our commitments is critical, and we are making progress toward that goal. AMD enters 2012 with significant momentum, and we are building upon that momentum by embracing the shifts occurring in the industry and marrying market needs with innovative technologies to become a consistent growth engine.”

Mr. Ghilardi joined AMD from Hewlett Packard in 2008.

A start-up called Tilera today unveiled a chip it calls the TILE-Gx. Essentially, it’s a super-chip with 36 cores which — so the company claims — beats a traditional Intel server chip on the key metric of performance per watt.

For those who don’t keep score in the arcane world of semiconductors, I’ll revisit some of the basics of the above paragraph. We all know that Intel and its one main rival, Advanced Micro Devices, sell chips for servers. Those chips, and those that go into PCs, are generally known as x86 chips, a name derived from the instruction set they share.

On the other hand, there are ARM chips, which are a different breed, and exist in a very different ecosystem. Scores of companies make ARM-based chips for all kinds of different uses, and they license the basics of the designs from ARM, the company, which last year did $636 million in revenue.

ARM chips show up in phones and tablets from the likes of Broadcom, Qualcomm, Texas Instruments and Nvidia, but not so much in PCs and servers. ARM is even the basis for Apple’s A4 and A5 chips. At CES last year, Microsoft said it would create a version of Windows 8 that will support ARM chips. And a company called Calxeda (which I initially got mixed up with Tilera) is aiming to bring ARM cores to chips running in servers.

Tilera, based in San Jose, Calif., is backed by investments from Bessemer Venture Partners, Walden International, Columbia Capital and VentureTech Alliance; plus a trio of strategic investors, Quanta Computer, NTT Finance and Broadcom. Its new chip is based around an entirely new architecture developed by Tilera’s CTO Anant Agarwal, a professor at the Massachusetts Institute of Technology. It forgoes both the traditional x86 and ARM architectures. Aimed squarely at servers, its intention is to get the same work done that a traditional Intel server chip does, while using less power to do it.

That’s not a trivial benefit, especially in data center environments where servers are bunched together and pushed to the performance limit. The biggest operational expense in running them is going to be power. So it’s on this point that server vendors and chip vendors obsess over saving a watt here and there — over the machine’s useful lifetime, the costs will add up considerably.

How it does this is what makes it interesting. Essentially, the cores on the chip do something that an Intel chip can’t do: They communicate among themselves. The way I understand it — and I admit I’m simplifying it greatly — the cores on an x86 chip rely on a single communications channel, called the Bus, to communicate. The Tilera architecture allows each core to communicate directly with the other cores, thus eliminating the need for the Bus and cutting back on the need for power.

The top-end chip — there are two versions — has 36 cores. A core is essentially the main computing engine on a chip. If you’re reading this on a PC, chances are the chip inside it has two cores, maybe four. It used to be that chips had only one core, until it became logical to put two or more on a single chip. I’ve always compared multicore chips to roommates folding laundry together. When there’s a big pile of laundry to be folded, one person can certainly do it, but two or four get it done faster and with less effort. Multicore chips basically prove the old adage that many hands make for fast work.

There’s an obvious appeal to a chip like this, but there are a lot of strikes against it. First, much of the server ecosystem is pretty well entrenched. Companies run what applications they already have, and are usually loath to mess with their computing environments much. Changing the architecture of the CPU chip inside the servers is about as major a decision as a CIO may ever make, and one they don’t make lightly. First they’ll have to test it and run it for awhile, and then see how it interacts with other systems. It’s not the sort of decision that happens just overnight. Also, a new architecture brings with it a lot of software compatibility questions that will give many IT departments pause.

Meanwhile, Intel, which sells chips that go into most of the world’s mainstream servers, will continue to push its power consumption down. At the same time, it’s been trying like crazy to use its Atom line of chips to mount an attack on ARM’s territory and win business from phone and tablet vendors. That effort is just now seeing its first early successes. If there’s a great long-term story in chips that bears watching, the grappling between Intel and the ever-expanding universe of ARM vendors is certainly it.

Correction: I initially thought the Tilera chip was based on the ARM architecture. I’ve revised the story to correct that.

AMD CEO Rory Read

First there’s Intel, which has parlayed its deep strategic command of the complex and expensive process of manufacturing chips into a commanding position at one of the crucial pivot points of tech hardware, despite the uncertain state of the global economy, and also despite the dire predictions of analysts.

Then there’s Advanced Micro Devices, which competes for the same sockets in PCs and servers that Intel does, but with a lot less success to show for it. While it reported a non-GAAP profit of 19 cents a share on $1.7 billion in sales, its outlook for the current quarter was weak enough to send AMD shares down 2 percent in after-hours trading. Its revenue forecast of $1.51 billion to $1.61 billion caught analysts by surprise, as they had been expecting average revenue of $1.6 billion.

Then there’s the matter of a $209 million charge from the writedown of AMD’s stake in Globalfoundries, the contract chip manufacturer that was cobbled together out of what used to be AMD’s factories in Germany, Texas, and one under construction in upstate New York; plus the former Chartered Semiconductor. That expense went a long way toward pushing AMD’s results into a 24-cent per-share loss on a GAAP basis.

Spinning off its expensive factories was supposed to save AMD, and it may well have done so. AMD now owns less than 9 percent of Globalfoundries. But the strategic shift has yet to pay off in the kind of success that AMD envisioned when it first plotted the move.

This fact caused chipmaker Intel to slash its sales forecast for the quarter by $1 billion, to $13.7 billion plus or minus $300 million. Today we’ll see just how bad the damage was, and how bad it’s going to be going forward, as the company reports its results after the close of markets today.

In a note to clients, Deutsche Bank analyst Ross Seymore says not to expect many surprises from Intel. Despite the lower sales guidance, he sees little change to Intel’s overall profitability. He expects Intel’s gross margins, a key metric in measuring profitability, to come in only slightly below the 64.5 percent that Intel had previously forecast. “We see little risk to gross margins despite the lower revenue because product mix continues to be solid,” Seymore wrote.

And while the first half of the new year is always seasonally slower than the second half, Seymore expects it to be slower still for Intel. He’s expecting sales in the first quarter of 2012 to come in at $12.73 billion, with a per-share profit of 51 cents, which is below the consensus of analysts who expect Intel to book sales of $12.8 billion. He also expects the hard drive shortage to hit Intel harder in the first quarter, as the supply of hard drives dries up. Gross margins in the quarter, he expects, will drop below 61 percent. Expect conservative guidance on PC demand for the quarters ahead.

Even so, Seymore rates Intel a Buy, with a $27 price target. “We believe Intel is well-positioned to benefit from new product introductions, improved execution and stable PC demand,” he wrote. He said that competition in chip prices from Intel rival Advanced Micro Devices should “remain benign,” mainly because AMD is fabless and so is at a competitive disadvantage with Intel. Other sectors of Intel’s business, including its flash memory operations, are showing improvement.

That would be the unit that sells graphics chips to dedicated PC gamers and others that pay top dollar for blazing speed. Its latest offering is definitely styled as a silicon Ferrari, not a Honda Civic.

“A single one of these is the fastest thing on the planet,” says Devon Nekechuk, a product manager at the AMD operations in Canada, acquired as part of the 2006 acquisition of ATI Technologies.

Read the rest of this post on the original site »

And while it was Fujitsu, using SPARC chips, that made the top of the list, you couldn’t help noticing how many machines used chips from Intel. Of the 500 supercomputers on the list, 384 of them use chips from the semiconductor giant.

At the SC11 Supercomputing conference in Seattle today, Intel is making some important disclosures about what it is doing to maintain its role as the chip vendor of choice, and also offering its competitive response to a potential threat from the graphics chip specialist Nvidia.

As I’ve explained a few times before, the graphics chips, or GPUs, that Nvidia makes are starting to make some inroads into supercomputing and high-performance computing environments, thanks to their ability to handle floating point computations at a high rate of speed. Sometime next year, at the Oak Ridge National Laboratory in Tennessee, a machine called Titan, using a combination of chips from Advanced Micro Devices and Nvidia, is expected to break the 20 petaflop barrier when it begins operation.

The narrative that has emerged recently is that GPUs are generally better at the floating point operations that are increasingly used in supercomputing — better in many cases than traditional x86 chips from Intel and AMD. Even so, if you add up the number of systems on the Top 500 list using Intel and AMD chips, you’d hit a percentage that’s just shy of 90.

In a presentation today (on what just happens to be the 40th birthday of the Intel microprocessor — hence the two people I saw today outside the “Today” show at Rockefeller Center on my way to work), Rajeeb Hazra, Intel’s general manager of Technical Computing, detailed Intel’s response. First off, Intel is supporting a new technology, called PCI Express 3.0, that will speed up the ability of chips inside a supercomputer to share data. In systems this big, and working on such large amounts of data at once, the processors spend a lot of time tapping their feet and waiting for data to work on. Engineers call this latency, and the point of the new interconnect technology is to cut latency by doubling the bandwidth available. The result is an improvement in the raw FLOPS (floating point operations) available by 2.1 times in lab tests, and a 70 percent improvement in real-world workload tests. In supercomputing terms, that’s real progress, and it effectively means getting answers to big questions faster.

Another advance that Intel talked about today is a chip bearing the codename “Knight’s Corner.” It’s a coprocessor, meaning it’s an additional chip that would be added to a computer to boost its performance. Intel says it can do a full teraflop — a trillion floating point operations a second — and that’s just the result of demonstrations from the first silicon. When in full production, it will probably do even better.

And not only will it do a teraflop on a single chip, it will perform those calculations to what engineers call “double precision,” which is a fancy way of saying the result of each operation will be accurate to a higher level of granularity. As John Hengeveld, Intel’s director of technical computer marketing, told me last week, the rule of thumb in these matters says that moving from single to double precision boosts the amount of time you have to wait by four times.

Why is that important, when an off-the-shelf GPU from Nvidia can do 2 teraflops — though only at the single-point precision? Programming. If you’re a scientist who 10 years ago wrote a program to simulate weather patterns or nuclear explosions or some other classic supercomputing problem to run on systems running Intel chips, there’s nothing new to learn in terms of programming. While the GPUs are great, there are new programming rules to learn.

Finally, Intel is reiterating its plan to keep working on the exascale problem, which is the next great summit in supercomputing. Right now the world’s top supercomputer maxes out at 10.51 petaflops, and a candidate to top the list next year will go north of 20 petaflops, or quadrillions of floating point operations. Sometime this decade — say, about 2018 or so — the hope is that supercomputers will break the exaflop barrier, where machines will run quintillions of FLOPs.

The fundamental problem there isn’t the computing so much as it is power, as in electrical power. Already some of these machines consume as much power as a small city. Getting to exascale will require chips and other components that can run full out at speeds we can as yet only imagine, but doing it consuming a lot less power than they would otherwise be expected to. Think in terms of a Prius that could win the Indy 500 — and not just by a hair, but by a long mile — and do it day after day without really using much more gas than the other cars. It’s kind of like that.

Anyhow, Intel has said that it plans to enable exascale supercomputing that will require only a doubling of the power needed, rather than, say, 10 times as much. To that end, it said today it will open its fourth research lab in Europe. This one is in Barcelona and joins one in Paris; another in Juelich, Germany; and a third in Lueven, Belgium. They’ll all have a lot of work to do between now and 2018.

If you’re a regular AllThingsD reader, you’ve already been introduced to the world’s most power supercomputer: It is the Fujitsu K Computer, which the Japanese computing concern disclosed earlier this month, and it runs in Japan’s quasi-public research institution RIKEN. That’s it in the picture above.

It’s capable of performance as high as 10.51 petaflops, or 10.51 quadrillion floating point operations per second. The same machine had been rated in the top spot on the list before, but was less powerful then, because it was still being assembled, and then capable of only 8.16 petaflops.

The machine is based on SPARC chips — the chips for which Sun Microsystems, now part of Oracle, gained such renown. Fujitsu has been building SPARC chips under license and using them in its own servers and supercomputers for years. In this case, there are 705,024 SPARC64 processing cores in action. And if my memory is correct, the chips in question each have four cores on board, meaning there are 176,256 individual processing chips in the machine.

It’s the first machine on the Top 500 list to venture past the 10-petaflop milestone; however, work is underway in the U.S. on a machine known as Titan, which will supposedly break the 20-petaflop mark sometime next year.

In the meantime, the second most powerful machine in the world is in China. The Tianhe-1A system took the top spot on the list a year ago — and in the process, caused President Obama such consternation about the state of American leadership in innovation that he mentioned it in his State of the Union address to Congress. Its performance reaches 2.57 petaflops and it’s powered by a combination of Intel-made Xeon processors and Nvidia graphical processing units.

In fact, the supercomputers in the top 10 spots on the list are otherwise unchanged from the list released in June.

At No. 3 is Jaguar, the system at the Oak Ridge National Laboratory that is being rebuilt into the machine called Titan, which I mentioned before. It’s a system built by Cray primarily around Nvidia GPUs and Opteron processors from Advanced Micro Devices. Its current performance is just shy of 1.8 petaflops.

The No. 4 system is in China. It’s called Nebulae and is at the National Supercomputing Centre in Shenzen. Its performance is just short of the 1.3-petaflop mark. No. 5 is called Tsubame 2.0, and is at the Tokyo Institute of Technology in Japan.

Chip companies in particular like to crow about the use of their products in the systems that wind up on the list. That makes this a banner day for Intel. Of the 500 systems on the list, 384 of them — 77 percent — use Intel chips. Chips from AMD, Intel’s main rival, are in 63 systems.

It’s a banner day for Nvidia, too. Its GPU chips can be found in 35 systems, more than double the number from the previous list. GPUs were invented to make the graphics in computer games more stunning and realistic; as such, it meant they were, from the beginning, pretty good at performing a certain type of math problem known as a floating point operation. It turns out that the people who run supercomputers do a lot of floating point operations — or FLOPs — too. So as GPUs have gotten more powerful, they’re finding their way into an ever-larger number of the world’s top supercomputers. Two supercomputers on the list use GPU chips from AMD’s graphics chip unit, ATI. Two more use IBM’s PowerCell architecture, which is a sibling of the Cell processor chip found in the Sony PlayStation 3.

President Obama shouldn’t feel so bad about the U.S. not being in the top spot. For one thing, practically all of the systems on the list are built on American-made technology. And among the systems that can reach 1 petaflop in performance or more, the U.S. has five, more than any other country. China and Japan have two each, and France has one. And the U.S. has more supercomputers on the list than any other country: 263. European countries have a combined 127; China has 75 and Japan has 30.

Intel may furnish more chips to the Top 500 list than anyone, but the king of the systems vendors on the list is unquestionably IBM, followed by Hewlett-Packard. IBM built 223, or more than 44 percent, of the machines on the list; HP built 140 of them. IBM also led the performance pack: Its machines are responsible for more than 27 percent of the total. Fujitsu, which made the list-topping K Computer, was in second place, with 14.7 percent. Cray and HP were in a statistical dead heat, with about 14 percent each.

Here’s a link to the full list, and a bunch of other things related to supercomputing.

Patrick Moorhead, AMD’s corporate VP for strategy and an AMD Corporate Fellow, is leaving the company, and according to people familiar with his plans, will be launching a consumer-focused technology analyst and consulting firm around the time of the International Consumer Electronics Show in Las Vegas in January.

Moorhead is the last remaining VP been hired by AMD’s legendary founder and former CEO, Jerry Sanders. Over 11 years at AMD he led the company’s marketing efforts around its Athlon PC and Opteron server chips that led to a bit of a renaissance at AMD from about 2005 to 2007, when the chips won a lot of business away from Intel and thus gave the bigger company a major migraine headache.

To place Moorhead (pictured from his Twitter feed) appropriately in AMD’s history, it was during these years that AMD put into products a concept called x86-64, which essentially extended the x86 instruction set — the underlying code that chips from Intel and AMD share — into what was then the bright new world of 64-bit computing, thus paving the way for machines that could contain more than two four gigabytes of memory and could handle more complex computing tasks.

AMD first put forth its approach at a chip industry event in 1999 — one that I happened to cover for a now-defunct outlet called Electronic News — at a time when Intel was championing a different approach to 64-bit computing by starting from scratch with an entirely new design. Its technology was called EPIC, for “explicitly parallel instruction set computing.” The product that eventually resulted was the exotic Itanium chip, which is today the subject of a legal dispute between software giant Oracle and Hewlett-Packard, which is for all intents and purposes the only company selling hardware that runs on Itanium.

AMD ultimately won that argument and Intel embraced its own implementation of AMD’s x86-64, now common in their mainstream desktop, notebook and server chips, but only after giving Intel and its investors fits over lost market share during 2005 and 2006.

Moorhead joined AMD in 2000 from Compaq and had also worked at the not entirely forgotten search engine outfit AltaVista, which had been launched at Digital Equipment Company, then acquired by Compaq, and is now part of Yahoo.

In more recent years he had been known primarily for being an outspoken advocate for the opportunities in mobile computing. One suspects he’ll have more to say on that topic in the coming months.

The job cuts will hit the company globally and will be completed by the end of the first quarter of 2012, the company said in a statement. The company says the move will reduce operating expenses by $118 million in 2012 and by $10 million in the fourth quarter of 2011.

The cuts will amount to about 1,400 of AMD’s 12,000-strong workforce. What you’ll read elsewhere is that AMD is suffering from a worldwide slowdown in PCs, caused in large part by the growth of Apple’s iPad business and to a lesser extent other tablets. But it’s a little more complicated than that when you consider that AMD’s share of the market actually grew slightly in the second quarter, according to the latest numbers I have from Mercury Research, which tracks the market share between the two chipmakers. As of August, AMD was running a 19.4 percent share of the PC and server business, up from 18.2 percent in the first quarter, while Intel’s share dropped slightly from 81 percent in the first quarter to 79.9 percent in the second. Those fractions of a percentage point actually matter to both companies.

The bigger problem is that AMD is nowhere in tablets or smart phones and that was one of the reasons that Read was brought in to replace Dirk Meyer, who surprised everyone by resigning after a fight with his board of directors in January.

In the meantime, expect AMD to make a big fuss about server chips in the coming weeks.

AMD’s statement is below.

AMD Optimizes Cost Structure to Enhance Competitiveness and Accelerate Growth

Operational Savings of More Than $200 Million in 2012 Designed to Accelerate Future Growth in Lower Power, Emerging Markets and the Cloud

SUNNYVALE, CA–(Marketwire -11/03/11)- AMD (NYSE: AMD – News) today announced a restructuring plan and implementation of operational efficiency initiatives designed to strengthen the company’s competitive positioning. AMD expects that these combined actions will create a more competitive cost structure and rebalance the company’s global workforce skillsets, helping AMD to continue delivering industry-leading products while improving productivity, reducing time-to-market and better aligning with key industry trends that are expected to drive growth.

“Reducing our cost structure and focusing our global workforce on key growth opportunities will strengthen AMD’s competitiveness and allow us to aggressively pursue a balanced set of strategic activities designed to accelerate future growth,” said Rory Read, AMD president and CEO. “The actions we are taking are designed to improve our ability to consistently address the needs of our global customer base and stake leadership positions in lower power, emerging markets and the cloud.”

AMD expects that the restructuring plan will result in operational savings, primarily in operating expenses, of approximately $10 million in the fourth quarter of 2011 and $118 million in 2012, primarily through a reduction of its global workforce by approximately 10% and the termination of existing contractual commitments. The workforce reduction will occur across all functions globally and is expected to be substantially completed by the end of the first quarter of 2012. Based on anticipated savings from the restructuring plan, AMD expects fourth quarter 2011 operating expenses will be approximately $610 million.

As a result of implementing efficiencies across the company’s operations, AMD expects to save approximately $90 million in 2012 operating expenses in addition to the restructuring plan savings, resulting in more than $200 million of expected combined operational savings in 2012.

The company expects to reinvest a significant portion of the savings to fund initiatives designed to accelerate AMD’s strategies for lower power, emerging markets, and the cloud.

The company’s actions pursuant to the restructuring plan will take place primarily during fourth quarter of 2011, with some restructuring plan activities extending into 2012. The company currently estimates that it will record restructuring expense in the fourth quarter of 2011 and in 2012 of approximately $101 million and $4 million, respectively. Of the total restructuring expense, approximately $56 million will be future cash expenditures in 2011, $33 million will be future cash expenditures in 2012 and $15 million will be future cash expenditures in 2013.

About AMD
AMD (NYSE: AMD – News) is a semiconductor design innovator leading the next era of vivid digital experiences with its groundbreaking AMD Fusion Accelerated Processing Units (APUs) that power a wide range of computing devices. AMD’s server computing products are focused on driving industry-leading cloud computing and virtualization environments. AMD’s superior graphics technologies are found in a variety of solutions ranging from game consoles, PCs to supercomputers. For more information, visit http://www.amd.com.

Today it was Fujitsu’s turn. The Japanese computing giant teamed up with RIKEN, the quasi-public Japanese research institution, to announce that they had built a machine they call the K Computer, which can perform 10.51 petaflops, or 10.51 quadrillion floating point operations per second.

And while all that may sound very impressive, it’s not quite as muscular as the Titan machine being assembled in the U.S. at the Oak Ridge National Labs, which can — or will — perform 20 petaflops.

The machine (pictured) is made up of 864 racks with 88,128 interconnected CPU chips, all of them based on the SPARC architecture for which Sun Microsystems, and therefore Oracle, are best known, though Fujitsu has long been a SPARC licensee. The new K Computer is basically an improvement and extension to the same K computer that took the top spot on the last Top 500 list in June, supplanting in the process a Chinese machine that had taken the crown last November.

Never mind that it contained all U.S.-made chips, the Chinese feat caused the leader of the free world to kvetch about the apparent sorry state of U.S. supercomputing, thus prompting, perhaps indirectly, the Titan machine at Oak Ridge.

It’s not as though China hasn’t been heard from on the supercomputing front recently. Last week its Sunway BlueLight MPP raised eyebrows not for its performance — a relatively pokey 795 teraflops — but rather for the fact that it’s built using all Chinese-made components.

So what will it be used for? Weather simulations, research into drugs and solar cells, and simulating earthquakes and tsunamis.

Here are the more formal descriptions from the announcement:

–Analyzing the behavior of nanomaterials through simulations and contributing to the early development of such next-generation semiconductor materials, particularly nanowires and carbon nanotubes, that are expected to lead to future fast-response, low-power devices.

–Predicting which compounds, from among a massive number of drug candidate molecules, will prevent illnesses by binding with active regions on the proteins that cause illnesses, as a way to reduce drug development times and costs (pharmaceutical applications).

–Simulating the actions of atoms and electrons in dye-sensitized solar cells to contribute to the development of solar cells with higher energy-conversion efficiency.

–Simulating seismic wave propagation, strong motion, and tsunamis to predict the effects they will have on human-made structures; predicting the extent of earthquake-impact zones for disaster prevention purposes; and contributing to the design of quake-resistant structures.

–Conducting high-resolution (400-m) simulations of atmospheric circulation models to provide detailed predictions of weather phenomena that elucidate localized effects, such as cloudbursts.

So what’s a petaflop anyway? A FLOP is a floating point operation. Its a type of mathematical function that involves decimal points. Adding 5.6 and 11.21 is a floating point operation and is therefore slightly more complicated from a computing standpoint than adding 11 and 5. But in computing, even day-to-day computing, it’s massively more complicated than all that.

A top-of-the-line NVidia GeForce GTX 590 graphics card, which specializes in floating point operations, can run about 2,400 gigaflops. Since a gigaflop is a billion flops, I guess that technically puts the GeForce GTX 590 into the teraflop, or trillion-flop range.

Petaflops are then in the quadrillion-flop territory, which as I noted before makes them fun because they’re among those rare numbers that are larger than the U.S. national debt. So 10.51 quadrillion flops gets written like so: 10,510,000,000,000,000. Didn’t I say this was fun?

All this is leading up to a big supercomputing conference starting in 10 days in Seattle. So expect lots more supercomputing news in the coming days!

AMD — which designs semiconductors that serve as calculating engines in computers, servers and game consoles — has benefitted from strong demand for its new chips that combine graphics and computing on the same piece of silicon.

Read the rest of this post on the original site »

The end of the search comes just as suddenly as it began, with Dirk Meyer’s surprise resignation in January. The search was a tough one, in no small part because AMD — whose business is already complicated by the fact that it has to compete with Intel — is widely seen as missing the boat on key new markets like mobile computing and tablets.

It’s equally surprising that there were no leaks about this hiring, though word had circulated that several A-list executives had been approached and turned the job down. Among them: Pat Gelsinger, COO of EMC and a former CTO of Intel, said no twice; both Michael Capellas, the former CEO of WorldCom and Compaq, and William Nuti, CEO of NCR, turned down AMD’s inquiries.

So, who is Rory Read? His bio at Lenovo says he led that company’s American unit to a $140 million surge in profitability in 2007, and revenue grew 14 percent on his watch. Before that, he spent 23 years at IBM — remember that Lenovo bought out IBM’s PC division for $1.75 billion in 2005.

AMD shares rose 10 cents, or more than 1.5 percent, as of 10:40 AM Eastern time.

Today it fell to Harry Wolin, AMD’s senior vice president and general counsel, to sound a bit like Orson Welles doing a 1970s wine commercial for Paul Masson. At the opening of AMD’s quarterly earnings conference call with analysts, Wollin made a statement that the search for a new CEO remains a “top priority,” but that meeting a timeline is “not the driving force for the search.” Finding the right person is. I hate to say it, but I knew it was going to be a complicated search.

Wolin doesn’t sound much like Orson Welles, but you can hear his statement below. And if you’re drawing a blank on the reference to Orson Welles in wine commercials, there’s an example of one from 1978 below that.

AMD-Harry-Wolin by ahess247

Doug Freedman, an analyst who covers the chip sector for Gleacher & Co. in San Francisco, trimmed his estimates on both Intel and on its smaller rival Advanced Micro Devices amid a weakened PC market that is running well behind the typical seasonal patterns. Last week, market researcher Gartner reported that worldwide PC shipments grew less than three percent over the year-ago period, as consumers remain focused on tablets and smartphones and hold off on upgrading their desktops and notebooks.

At a high level, that’s not good news for Intel and AMD, both of which have yet to penetrate the tablet market in any meaningful way. And both are grappling with the impending entrance of competing chips — based on designs from ARM, from vendors like Nvidia and Qualcomm — burrowing their way into new consumer notebooks.

In a July 15 note to clients, Freedman cut his estimates on both Intel and AMD for the quarter ending in June and the quarter ending in September. He expects Intel to report sales of $12.7 billion, which is about $100 million below the street consensus of $12.8 billion. He also expects Intel to report per-share earnings of 53 cents, which would amount to a two-cent improvement over the year-ago quarter.

The quarter being reported today isn’t the story, however: It’s September. Typically it’s a seasonally strong quarter, as college students head back to school with new notebooks under their arms. This year Freedman thinks PC sales will lag behind historical patterns. He trimmed his September quarter revenue forecast to $13.16 billion, down from $13.43 billion — or $300 million below the street view — and knocked it down by two cents to 59 cents, a penny above the street.

Intel has, in recent quarters, taken a fairly aggressive stance on the state of the PC market, and has criticized analysts for fanning investor fears. “Management should offer more subdued PC unit growth expectations thereby alleviating investor fear that Intel is setting its bar too high,” Freedman wrote. One other thing Intel has in its favor is that the average selling price of chips is edging upward, which should give it a slight hedge against the weaker market. This should help keep gross margins — a key metric for Intel — in the higher end of the 59 to 63 percent range the company said to expect. He also says that Intel could deliver a surprise with better-than-expected results from other parts of its operations, namely its flash memory unit, which makes solid-state hard drives.

For AMD, which reports its results tomorrow, the picture is a mixed bag. The search for a new CEO is now in its sixth month, with no sign of being resolved anytime soon. Freedman doesn’t expect a CEO to be named today nor in the near term. Finding an external candidate is proving harder than expected. (Note to Freedman: You don’t say.) Pressure on its share price, thanks to short-sellers, has created a buying opportunity in the near term.

Even so, Freedman trimmed his estimates for AMD’s June and September quarters. He expects AMD to report sales of $1.55 billion, down from $1.6 billion previously, which would amount to a four percent decline in year-on-year sales. He also shaved a penny off his EPS estimate to nine cents from 10. For September, he expects AMD to report sales of $1.63 billion, down from $1.7 billion before, and cut his EPS estimate to 16 cents from 20.

I’ll be back later today to cover Intel earnings live. See you then.

Alex Gauna of JMP Securities cut his rating on AMD to “market underperform,” arguing that its Fusion line of chips isn’t winning sufficient business from PC makers. Having interviewed people at PC retailers and at PC manufacturers, he concludes Fusion is going nowhere. He says he was unable to purchase a Fusion-based PC from Dell via its Web site.

AMD’s Fusion processors aren’t winning many positive reviews, Gauna says, indicating a “lack of market interest.” Sales representatives at various online retailers steered his researchers “away from AMD platforms.”

As you might expect, it doesn’t take much to send AMD shares south these days. Intel is far and away the king of the hill in PC and server microprocessors, and chips from companies like Nvidia and Qualcomm based on the designs of ARM Holdings are coming into notebooks and tablets, minimizing AMD’s chances to win business there. He slashed his target price on AMD to $4.50 a share, which would constitute a drastic drop from the $6.80 at which it opened today, having closed Friday at $6.95. By today’s close of regular trading, investors knocked AMD’s shares down to $6.76, a drop of nearly 3 percent.

The day ended however, with AMD shares being defended by analyst JoAnne Feeney at Longbow Research. Dell is not an early adopter of the AMD technology, she says, but will be offering systems using AMD’s Fusion chips — the codename of the specific AMD product is Llano — soon. I heard the same thing from a source at AMD. On top of that, Toshiba and Lenovo are happy AMD customers. Tiernan Ray at Barron’s has a little more on Feeney’s note.

The downgrade came 10 days before AMD is due to report quarterly earnings and amid growing speculation on the company’s ongoing search to find a new CEO to replace the ousted Dirk Meyer. The search is about to drag into its sixth month, and I’m told it won’t be resolved before earnings are announced.

There were reports last month that several A-list executives, none of whom would have realistically considered the job in the first place — Oracle’s Mark Hurd, Apple’s Tim Cook — had turned down approaches by AMD’s recruiting agency, Heidrick & Struggles. Another who topped AMD’s list early on, Pat Gelsinger, the former Intel CTO who’s now in line to succeed Joe Tucci as the CEO of storage giant EMC, even said he told AMD no not once but twice. A lot of those people, Hurd and Cook especially, are really just names that any recruiter looking to fill a senior position would call in the course of building a list, just to make sure they’ve covered their bases.

The sense of urgency to find a new CEO seems to have dissipated at AMD, though not entirely. With the company running a profit again — its most recent quarter showed a $510 million profit on sales of $1.6 billion — it may be that directors feel they have the luxury of time in finding the right candidate, so they don’t have to rush the process, though obviously they’ll want to get it done soon. Expect the company to face a lot of questions from analysts when it reports earnings on July 21.

This is exactly the sort of problem I predicted in January. That Hurd, who is also a former CEO of Hewlett-Packard and of NCR before that, and Gelsinger, a former CTO of Intel once considered a possible successor to Paul Otellini, have been approached is not surprising, given their tech and managerial bona fides. Nor is the fact that they turned the job down.

The third name jumps out at me simply because I’m not familiar with Greg Summe. His bio on the Carlyle Group site says he spent 20 years as chairman and CEO of PerkinElmer, the $2 billion health sciences company, and before that, he ran the Avionics business at AlliedSignal, now part of Honeywell.

The search is being run by Heidrick and Struggles, Bloomberg says, and the fact that Summe was approached indicates how widely the company is casting its net. The clock, however, is ticking. When I last spoke to someone at the company, not directly involved with the search, I was told that the plan was to have a new CEO named before its next earnings report, scheduled for July 21. That’s 36 days away.

Historically, this is unlike AMD, which has always taken care to have a managerial bench, and like most big companies, has typically had a CEO successor waiting in the wings. Former CEO Dirk Meyer (pictured) was named COO in 2006, tapped by then CEO Hector Ruiz, who had himself been recruited from Motorola’s Semiconductor unit (now Freescale) to succeed AMD’s founding CEO, the colorful Jerry Sanders. Ruiz, however, had been recruited in 2000 because of the surprise resignation in 1999 of AMD’s heir apparent, Atiq Raza, who’s now a tech investor, backing, among others, Violin Memory.

That’s not to say there aren’t internal candidates who could step up. Rick Bergman, senior vice president and general manager of AMD’s products group, has been mentioned as on the list for consideration, though the board has favored an external candidate from the start. But the company has been bleeding talent. Two other internal contenders bolted in February — Bob Rivet, AMD’s onetime COO under Meyer, and Marty Seyer, the well-regarded senior vice president for corporate strategy, who in 2006 personally landed the deal to sell the first AMD server chips to Dell (until then an Intel-only shop) and had been known to occasionally jam with Ruiz on the electric guitar. Other senior managers are bailing out as well. Just last week Jeff VerHuel, corporate vice president of platform strategies, left AMD to join SMSC as its head of engineering.

Meyer’s sudden departure is said to have come after a row with the board of directors, impatient that AMD is not showing up in any meaningful way in the market for chips for mobile devices. The days when it was dealing perennial market leader Intel bruising punches in the punishing business of selling server chips are over. And its overall share of the market for PC and server chips has slipped to 13.2 percent versus Intel’s 86.5 percent as of March, according to Mercury Research. It still makes a compelling case as an alternative supplier for chips in notebooks and desktop PCs, as The Wall Street Journal reported yesterday, but Intel’s lead, given its powerful manufacturing infrastructure — AMD no longer owns its own factories, opting instead to farm those duties out to GlobalFoundries, its onetime manufacturing arm — will as the years progress prove ever more difficult to erode even incrementally.

And even trying will increase the operational costs of an already profit-challenged company. AMD delivered profits in 2009 and 2010, but only after undergoing a massive restructuring to rid itself of its manufacturing operations. Still, the profits are thin: In 2010, AMD reported income of $471 million on sales of $6.5 billion. Compare that to Intel’s $11.5 billion profit on nearly $44 billion in sales, and you see how hard a time even the new, leaner, fabless AMD has competing with Intel.

Competing with Intel for share of its traditional markets is hard enough. If AMD’s board is determined to push the company into the business of selling chips for mobile devices, the path to success looks nearly impossible. Just look at the troubles Intel is having in that space competing with ARM Holdings and its numerous licensees, which include Nvidia, Qualcomm, Broadcom, and Texas Instruments to name but a few. When it comes to mobile devices — tablets and smart phones — ARM-based chips are as ubiquitous as x86 chips are in PCs and servers: They are the standard. Intel’s low-power Atom-based chip has so far been largely unsuccessful in penetrating that business. And if Intel is not scoring any significant wins there, why would anyone want to take on the job of leading AMD into a likely failure? No wonder potential candidates are finding it easy to say no. Bloomberg quotes Gelsinger: “I said no, and I said no again.”

So where does that leave AMD now? There are two paths. First, consider an internal candidate to lead the company. As more external candidates spurn AMD’s approaches, the list of objections AMD’s board may have to hiring internally could shorten. Bergman may get a second more serious look.

The other is to sell the company to someone bigger. That’s another complicated question, mainly because with the terms of its settlement with Intel (or what I like to call the Treaty of Maui) and the terms of its complicated patent-cross licensing agreements that date back the the 1980s, any buyer would have to first pass muster with Intel or find themselves in a very expensive lawsuit. Then there’s the fact that AMD is 20 percent owned by Mubadala Development Company, the investment arm of the Arab Emirate of Abu Dhabi. Buying AMD — at current valuations it would take about $7 billion — would be, to paraphrase Steve Jobs, a “big bag of hurt.”

Even private equity players who specialize in buying troubled companies, fixing them up and spinning them off at a profit, are wary of AMD, having learned well the lessons of the disastrous 2006 buyout of Freescale by the Blackstone Group, Carlyle, TPG Capital and Permira Advisers.

Ultimately there will be no easy options at AMD. No surprise, its shares are trading down by 11 cents or more than 1 percent as of 9:45 am New York Time this morning.

The firm has cut its growth estimate on PC sales in 2011 almost in half. In February, it said PC shipments would grow more than 7 percent. Now it says shipments will grow by only a little more than 4 percent. The reasons are a combination of things you can probably guess: A tightening economic outlook, a saturation in developed markets, and, well, the iPad. IDC uses the phrase “competing products,” but we all know that the competing product of the moment is tablets, and when you talk tablets, you’re talking about Apple’s iPad. Even so, IDC says 2011 will be a down year compared to 2012 through 2015, when growth is expected to head north of 10 percent.

The 2009 consumer-driven boom cycle in notebooks and mini-notebooks has faded. And shipments in the first quarter of 2011 declined by more than 4 percent versus the same period a year ago. That drop was only offset a little by a 3 percent growth in shipments to companies. The consumer slowdown was worst in the U.S., Canada and Europe, IDC says. Add to that the Japanese earthquake, the Arab Spring, and a dour economics outlook, and you end up with a market that’s going to grow a lot slower than previously expected.

As we’ve seen, these forecasts initially tend to impact the share prices of companies in the PC ecosystem like Intel, Advanced Micro Devices, hard drive makers like Seagate, not to mention the PC makers themselves like Hewlett-Packard and Dell. But remember what happened when Intel last reported its quarterly earnings. CEO Paul Otellini slapped down the market research firms in comments made on a conference call with analysts.

At the time, he said, “Our views differ from the views of the analysts,” and that, “Our projection for 2011 remains in the low double-digit range.” If you see Intel cutting its estimates ahead of its next earnings call–and there’s no evidence of that happening–you know it’s getting serious.

Update:An Intel spokeswoman says the company is standing by the forecast it gave in the second quarter. For the record the company said last month that it expects PC unit shipments to grow 11 percent this year notwithstanding any impact from tablet sales. It also said it expects revenue to be flat in the second quarter versus the first quarter, which would be an improvement over the usual decline the company sees at this time of year. I talked briefly with Intel about this today, but Don Clark at The Wall Street Journal has more on it.

Clearly it still sees a different market than IDC does, and wants the market to know it. We’ll see how it all turns out soon, as the second quarter comes to a close in about three weeks.

But what if you’re a company who doesn’t always sell directly to your customer, but rather relies on a reseller who stands between you and the end customer? That’s a different dynamic entirely. And it may seem like an insignificant detail until you consider that there are more than $2 trillion worth of goods sold annually through indirect business-to-business sales relationships.

Companies tend to call these “channel sales,” or use some variation of that phrase to describe this aspect of their business. And it brings with it a level of complexity that’s different from conventional direct sales.

And as Mark Geene, the CEO of Channelinsight tells me, it’s a sector of sales that has yet to benefit from the kind of productivity gains that Salesforce.com and SAP have brought to direct sales. The company has created a cloud-based service that does two things: First it aggregates live data, including inventory and point-of-sale data provided by a network of some 5,000 business-to-business resellers, distributors and retailers. Then it combines that with some screening and analytics tools that can run either as a standalone application or as an add-on to Salesforce.com. You can see who bought your stuff, who the end customer was, whether or not incentives or other programs are working, and whether or not they have inventory in the right places given demand trends.

“Indirect sales have been sort of the stepchild of salesforce automation,” Geene told me. “It hasn’t benefited at all from the kinds of things that Salesforce and SAP have been doing.” Managing indirect sales is often a rather labor-intensive process involving a lot of time looking at spreadsheets. Channelinsight’s play is to automate that process.

It must be working. Customers include printer manufacturer Lexmark and the chipmakers Advanced Micro Devices and Nvidia, both of which sell a great deal into the channel, as well as the German industrial giant Siemens. That sounds like momentum.

Channelinsight just closed a $10 million Series C round led by Rho Ventures, with participation from Sevin Rosen Funds, Sequel Venture Partners and Vendanta Capital. Its total venture funding so far is $21 million. Paul Bartlett, a Rho Ventures Partner, is on Channelinsight’s board.

“Our expertise is in identifying and investing in companies that redefine the status quo,” Bartlett said in a statment. “Channelinsight fits this profile by reinventing the way channel sales are managed, saving companies millions of dollars in the process.”

Separately, Channelinsight said it added Ram Gupta, the former CEO of Cast Iron Systems, which IBM acquired last year, to its board of directors.

Geene is a former Oracle vice president who ran its mid-west sales, and has held senior management jobs at Tenfold and Dorado Software.

He told me one big problem companies often run into with indirect sales is screening for regulatory compliance–in particular the Foreign Corrupt Practices Act, which covers, among other things, bribery.

You may remember that last month IBM settled allegations by the U.S. Securities and Exchange Commission that it paid bribes to officials in China and South Korea during a period beginning in 1998 and ending in 2009. IBM never admitted to any wrongdoing, though as the SEC noted, the purpose of the payments was to “secure the sale of IBM products through IBM-Korea and LG-IBM’s business partners.” Hewlett-Packard had its own headache with bribery allegations in Russia last year.

This is the sort of thing that Channelinsight can help a company watch for, Geene said. “This is a big issue for any company that works with partners,” he said. “We screen every transaction that gets processed on our system, looking for red flags.” It also looks for instances where products might be sold indirectly via partners to countries that are subject to trade embargoes. This sort of screening is something that companies have till now generally had to manage manually with some sort of custom solution.

For openers, the first quarter of the year is always the one that’s seasonally slower, as consumer PC sales slow down. And they are, as we saw in the latest estimates from Gartner and IDC last week. Of the $43.6 billion in sales Intel reported in 2010, nearly $25 billion or 57 percent of sales was derived from the sale of microprocessors into PCs. Add in another $6 billion and change for sales of chipsets and motherboards and Intel’s exposure to PCs jumps to 72 percent. By comparison, its Data Center segment, which sells chips used in servers, accounts for $8.7 billion or less than 20 percent of sales. If the PC market is slowing, then it’s hard for Intel not to slow down. However, its technology, its market strength versus rival Advanced Micro Devices, and its best-in-the world manufacturing efficiency allows it to roll with whatever punches the marketplace throws. Plus, Intel does a lot of business with Apple, which continues to boom.

For today’s report the focus will be less on Intel’s results–unless there’s a surprise–than on what Intel says about its guidance for the second quarter. The consensus view of analysts calls for Intel to report Q1 sales of $11.6 billion, which would be a year-on-year increase of 12.6 percent, and 46 cents in per-share profits versus 38 cents a year ago. The consensus view on Q2 calls for $11.87 billion in sales and a profit of 45 cents.

Here is where the analysts start parting company with each other. Hans Mosesmann of Raymond James passes for an Intel bull, seeing Q2 sales at $12 billion and profits at 49 cents. He argues that after accounting for Intel’s acquisitions of McAfee and the wireless chip unit of Infineon, if Intel issues guidance that is in line with seasonal exceptions (Q2 is usually a slow quarter for consumer PC sales as well), that would be a sign that most everything is on track.

There are other optimists out there, some more cautious than others. Take last week’s note from Doug Freedman at Gleacher and Co. He reminded clients that Intel had to fix a manufacturing glitch with its Sandy Bridge chip, one that caused some PC shipment schedules to slip. Though Intel made short work of the problem, Q1 ended “weaker than expected,” he says; he forecasts microprocessor unit sales to fall by 8 percent from Q4, which is in line with seasonal patterns. He expects Intel to report sales of $11.7 billion and a per share profit of 48 cents, but has a more positive outlook on Q2, with sales at $12.2 billion and profits at 49 cents.

Then there are the pessimists. Michael McConnell at Pacific Crest Securities says the consensus numbers are too high, and trimmed his own expectations. He’s calling for Intel to miss, with sales at $11.6 billion. Between the Sandy Bridge problem, the earthquake in Japan, and the acquisitions, there are “too many moving parts,” he says, for Intel to hit its numbers, let alone beat them. Despite relatively strong sales of chips into notebooks and a slight rise in the average price that PC makers pay for those chips, McConnell writes that Intel’s implied guidance for a decline of about 7 percent in chip sales is “likely to prove optimistic.”

Moreover, he sees Q2 estimates at too high as well. First of all, the quarter is at 13 weeks, one week shorter than Q1. Second, the ramp-up of PC makers turning out machines with the Sandy Bridge chip isn’t enough to offset other factors. He expects Intel sales in Q2 to hit $11.7 billion. Overall, he expects Intel to grow its core PC business by only 6 percent this year, well below Intel’s projection. We’ll see how it all shakes out later today. Check back this afternoon. I’ll be covering Intel’s results and the conference call.

Intel used its usual hyperbole to describe the big step forward it has taken with with Xeon E3 and E7 families of chips. In one slide of its presentation deck to reporters on a conference call today, it said that 18 racks of 2006-vintage servers using its dual-core chips from back then could be replaced by a single rack full of servers using its newest Xeon E7-4800 chip. Swapping out the old servers for the new ones would also cut the annual energy cost by 93 percent.

The refresh is part of Intel’s bid to go after the $15 billion ultra-high end of the server market, where powerful, expensive and often very customized machines running IBM Power7 chips or Oracle/Sun SPARC chips hold sway. But here’s where the shifting sands of semiconductor politics get a little complicated. For years, Intel has coveted this market and had planned for a very long time and spent billions on designing and building its Itanium chip to go after the big iron.

Remember last month when Oracle said it would stop building software that supports Intel’s ultra-high end Itanium chip? The kerfuffle that followed included rebuttals by Intel and Hewlett-Packard. Finally, Oracle had the last word, saying in effect that Intel wasn’t being entirely honest about its plans and that in fact it eventually planned to kill Itanium in favor of its higher-volume Xeon chip. (While Oracle happily buys lots of Intel’s Xeon chips for its Sun servers, the real target in all this was HP, which is for all intents and purposes the only company selling Itanium-based servers in meaningful numbers and one competitor Oracle has been taking every opportunity to disparage.)

With memories of that three-way scrum still fresh, it was no surprise at all to hear Kirk Skaugen, VP of Intel’s Data Center Group, defend Itanium, reminding reporters and analysts assembled for the launch event today that the anniversary of the Itanium chip’s release in 2002 is coming up on May 29. He then went on to defend Itanium and score points againt Intel rival AMD in one breath, saying that the Itanium ecoystem, constantly criticized for being much smaller than Intel initially envisioned, amounted to about $4 billion in server sales last year, bigger than the $2.8 billion in sales of servers running AMD chips. He then reiterated Intel’s promise to keep development of Itanium chips on track: Poulson and Kittson, code names for future Itanium chips, are in the pipeline for release.

Yet even after all that, it was hard not to side with Oracle’s interpretation of the situation. The key moment came when Skaugen said, “There’s no workload in the world that Xeon can’t handle.” Hmm. Does that mean a Xeon chip could in some scenarios replace an Itanium?

Confused, I turned to Dean McCarron of Mercury Research, an Arizona-based outfit that tracks the market for x86 chips (chips used in PCs, servers and workstations manufactured by Intel, Advanced Micro Devices and VIA Technologies).

Years ago, he said, there wasn’t much that differentiated a Xeon server chip from a Pentium used in a PC. But over the years that’s changed. “Xeon has grown up,” he said. Many of the features that Itanium included as suitable for specific tasks have migrated down the line and into Xeon chips. “While there’s still a lot of cases where an Itanium is the right choice for the customer, the fact is that over time Xeon has gotten better and is able to fit into more segments of the server market.” There may be some cases, McCarron said, where a customer who would have insisted on an Itanium a few years ago might be more willing to consider a less expensive Xeon.

Here’s what may be going on. McCarron told me that Intel’s share of the market for server chips stood at 93.4 percent at the close of 2010, while AMD’s stood at a paltry 6.6 percent. In a market where one percentage point amounts to about 50,000 servers–which on average contain roughly two server chips each–Intel’s Xeon is in need of new territory in which it can grow. By definition, that means attacking the market for big iron where systems running IBM’s Power7 and to a lesser extent Oracle’s Sun SPARC chips dominate. This will also mean Intel will bump into itself: Xeon will probably eat into Itanium’s business. That’s going to take some complicated, nuanced positioning.

As Skaugen put it today, just before his more memorable line about Xeon being able to handle any workload in the world, he said the coexistence of Xeon and Itanium is one of choice. Xeon, once positioned at the lower-cost-but-higher-volume end of the server market, is now, as he put it, “side by side” with Itanium. What’s unclear is how much longer that will continue. Expect Intel to shed more light on this at its developer forum in Beijing later this month.

But as I learned, the effects wizards at ILM have a secret weapon that shares a lot in common with your PC at home. If you play any graphics-heavy games, your PC probably has a graphics processing unit in it, and chances are pretty good that GPU card came from Nvidia. As you might expect, displaying ever-more realistic scenes in a PC game is similar in many respects to what you need to make a wicked cool effect in a movie. And in certain cases they’re better than even the most powerful traditional CPU chips from the likes of Intel or Advanced Micro Devices.

The story goes that when he was working on a scene for “Harry Potter and the Half-Blood Prince,” Chris Horvath was asked to create a “tornado of fire” (the picture above is borrowed from that scene). At the time, the conventional way of doing it just didn’t produce a satisfying result. “We needed to do this very complicated fire simulation and we just didn’t have a solution to do it,” said Craig Hammack, an ILM visual effects supervisor who was sitting near Horvath at the time.

Someone suggested to Horvath that he try working with GPUs, and not only that but writing an effects program to take advantage of their unique computing capabilities. (Horvath tells the story in the video below.) The result was a piece of internal ILM software called Verté that reduced the desired fire effects to a series of flat two-dimensional images linked together to look like they were 3-D.

Next came a new tool called Plume, used to simulate the movements of fluids. It’s written to take advantage of a newer Nvidia parallel computing technology called CUDA. Suddenly, the work required to create deeply complex visual effects involving images of fire or water sped up considerably because the software written for Plume could talk directly to the graphics chip itself.

“We used to have a smoke simulation engine that we had been using for I don’t know how long, and it rendered on CPUs, and the turnaround time was about a day,” Kirk Haller, ILM’s director of research and development told me. “You’d have to set things up and feel pretty confident that you were doing the right things,” because you wouldn’t see the result until the day after.

Plume changed the game in a big way. Developed first for use on “The Last Airbender,”, it allowed artists working on simulations to mess around with them on the fly, literally changing settings on software dials. “It’s a very friendly system. People who aren’t experts and don’t know exactly what numbers and settings to put in on the old system could tweak the settings and learn how it behaves, and get the artistic refinement and the look that they want,” Haller told me.

Plume sped things up so much on “Airbender” that effects artists were able to work with director M. Night Shyamalan in near real-time, allowing him to have input on how the simulated fire and water and air would look in each shot and how it would affect the characters on the screen, Olivier Maury, a research and development engineer at ILM, told me. Instead of waiting a day to see the results of each day’s work, an artist could work up as many as several versions of a complex simulation every day.

That’s why Plume–which is a proprietary tool–is now being used on every film currently in the works at ILM. And that’s a lengthy list, including “Cowboys and Aliens,” “Pirates of the Caribbean 4” and “Transformers 3,” to name but a few.

Plume simulation renders are run on a rack of 12 Nvidia Quadroplex 2200′s. Each machine in the rack contains two GPU chips, but each chip has 240 cores, the central computing brain of the chip. That means this rack has 5,760 computing brains jamming on simulations and effects shots at any time. That’s some serious horsepower. The rack is situated only steps away from the “Death Star” rack I showed you last week, though, silly me, I didn’t have the presence of mind to shoot any footage of it.

I wasn’t able to shoot any video to demonstrate any of this. But the folks from ILM like the results they’re getting from the Nvidia GPUs so much they appeared in a video about it last year that coincided with the release of “Airbender,” and it shows a good bit of the evolution of the process from “Potter 6″ to “Airbender.” Chances are you’ll be seeing a lot of shots first created in Plume in movies coming this summer.

PREVIOUSLY:

• A Virtual Tour of the Town of Dirt, from the Animated Film “Rango”
• Lucasfilm’s Data Center, and an Encounter With the Real Death Star
• Meet Kevin Clark, Master Not of the Force, but of Data

The move is no doubt a blow to Intel’s meager Itanium business, but it’s surprising that Oracle waited this long. Both Microsoft and RedHat announced they were jumping ship last year. Microsoft said last year that Windows Server 2008 R2 would be the last operating system to support Itanium. RedHat dropped its support with the release of Enterprise Linux 6.

From an operating system standpoint that leaves HP-UX, Hewlett-Packard’s variant of Unix as the only real option, which makes some sense since HP sells nearly all the Itanium-based servers on the market.

It’s also a bit of a turn of events. As CNet’s Stephen Shankland points out, it was only five years ago Oracle CEO Larry Ellison said “There is no more important platform for Oracle than HP and Itanium.”

That of course was before Oracle owned Sun Microsystems. Since then, and after former HP CEO Mark Hurd left to become an Oracle co-president, Ellison has turned bashing HP into something of a hobby. It wasn’t hard to detect a bit of a sneer as Oracle’s press release pointed out that in his strategy remarks last week, HP CEO Léo Apotheker didn’t mention Itanium once.

Intel had gone to a great deal of effort to develop the 64-bit server chip, and at the beginning of the last decade portrayed it as a significant leap forward in server computing. The problem was that it wasn’t backward compatible with existing software written for Intel’s standard x86 chips, and so software meant to run on Itanium systems required extra work or had to developed separately.

This caused the kind of controversy that only computer scientists could love and prompted one of the more interesting chapters in the decades-old rivalry between Intel and its much smaller rival Advanced Micro Devices. AMD suggested a very different approach to 64-bit computing by simply extending the standard x86 platform. Software initially developed for standard 32-bit chips could run without any of the extra work on a 64-bit system. The industry liked it, and within a few years AMD had implemented the idea on its Opteron server chips. During 2005 and 2006 AMD built up enough momentum to take away some of Intel’s share of the server market.

Under pressure, Intel changed its mind and, beginning in 2004, started putting out its own x86-compatible 64-bit chips. Its Intel64 technology is now a standard across its server, desktop and notebook chips. And it got most of its share of the server market back with its Xeon line.

Despite the fact that it never really caught on, Intel is still putting out Itanium chips, though it’s really a niche product. It updated the line last year with the Itanium 9300. It also has two future chips in the family still in the pipeline, one codenamed Poulson, the other codenamed Kittson. Time will tell if either ever see the light of day. Few companies buy servers built around Itanium chips.

HP was an early partner in the chip’s development, just last week announced a new Itanium-based blade system, and continues to sell its Itanium-based Superdone line. However the HP unit responsible for selling those systems, Business Critical Systems, is showing no sign of progress. In its most recent quarter, revenues fell slightly to $555 million from the year-ago period.

And while it will continue to run existing Oracle installations and still enjoy Oracle support, these machines won’t be able to run future versions of Oracle’s database. That’s not going to help HP sell any more of these boxes.

His name is Michael Wolfe. He’s 52 and has worked for HP for five years, most recently as VP for Information Technology. This will be his second go as a CIO. Before his stint at HP, he spent 24 years at Motorola’s Semiconductor Unit and was CIO during the period it was spun out to become Freescale Semiconductor.

His new boss, AMD’s interim CEO Thomas Seifert, had high praise. “Mike has effectively led IT transformations constantly focusing on reducing operating costs and significantly improving business innovation,” he said in a statement. “His considerable talent and experience will help AMD to continue strengthening our IT infrastructure and streamline our business based on our own products and platforms.”

This hiring is taking place against the backdrop of the complicated, difficult search for a new CEO at AMD following the surprise resignation of Dirk Meyer in January. COO Robert Rivet soon followed.

AMD shares haven’t moved much since then, and it has been the subject of recurring problematic buyout rumors. Today the shares closed at $8.55, unchanged from the prior session, and that’s up only a nickel from where it was at the start of the year. Shares fell five cents in after-hours trading. Investors seem to consider AMD a company in a holding pattern until there’s some resolution in the corner office.

The potential acquisition of ATI Technologies Inc., a Canadian graphics company, by chip maker Advanced Micro Devices Inc. was so secret that the project was at first code-named “Supernova,” and then “Go Big,” said Anil Kumar, a former McKinsey & Co. consultant who advised AMD on the deal. In discussions, AMD was referred to as “Los Angeles” and ATI was called “San Antonio,” he said.

But for months before the deal was announced in July 2006, Mr. Kumar testified, he had been discussing AMD’s plans with Mr. Rajaratnam. Mr. Kumar has testified that Mr. Rajaratnam had been paying him $500,000 a year for inside information, which was reinvested in Galleon through an account in India under his housekeeper’s name.

Read the rest of this post on the original site »