Nvidia, in a partnership with AMD, has set out to oust Intel from the SMB channel.
Nvidia has quietly contacted many of the top channel vendors and tier-2 and tier-3 system builders, attempting to put in place an Nvidia Business Platform that will certify new machines and put in place a stable image process that will remain fixed for a year’s time.
Nvidia’s program is an outgrowth and complement to AMD’s own Commercial Stable Image Platform (CSIP),
Nvidia executives, however, said that they have gone out to system builders like Acma, Compusys, Equus, and Polywell, and certified them for a range of core components, excluding peripherals. The company has also worked with Microsoft and Altiris to tweak the company’s ActiveArmor firewall and ensure that remote-management data traffic can be passed through and the PCs managed. Only North American and European partners are being targeted in the current program.
“Stable images” got their start in the late 1990s, as chipsets from Intel, Via, and others were being designed at a fast and furious rate, and government procurements agents began calling for a more manageable release schedule. The argument they presented was that the accumulated support costs of maintaining and updating the drivers for the various hardware platforms was unnecessarily expensive, part of the underpinnings of the “total cost of ownership” buzzword that evolved as a benchmark for evaluating cost.
Intel later agreed to design a single chipset for its government and large OEM clients, and pledged to use a single software driver, or stable image, for the year-long life of the platform. In January 2003, Intel launched the Stable Image Platform Program, formerly code-named Granite Peak.
Nvidia historically has been a supplier of enthusiast graphics chips and chipsets for home PCs, while AMD has found success in selling its Athlon64 chips to home PCs, and its Opteron line of server microprocessors into corporate environments. But the tier-2 and channel market is a 20-million-unit opportunity largely served by Intel, according to David Ragones, a product manager at Nvidia.
“We’ve been in close alignment with AMD with their stable image platform, and we’re going out together,” Ragones said. “We’re synchronized with their deployment configuration.”
Nvidia has also participated in the workstation market and penetrated back offices with Nvidia-based visualization systems. For that reason, Gartner analyst Martin Reynolds said his firm is now recommending the systems over those from ATI. Leslie Fiering, vice president of mobile computing for Gartner said that when AMD first announced its CSIP program the analyst firm endorsed it strongly. At that time, however, the company did not endorse a specific CSIP partner.
“We are recommending them as a good platform for AMD processors in a business setting,” Reynolds wrote in an email. “We make this recommendation over ATI because of Nvidia’s experience in server systems. And we believe that they understand stability better than the chipset manufacturers.”
Nvidia’s program has certified four motherboard manufacturers: Asus, Foxconn, Gigabyte, and MSI. Each board contains either an AMD Athlon64 or Athlon X2 processor as well as an nForce 430 chipset, with an on-board GeForce 6150 graphics processor. Peripherals like hard drives and optical storage are left up to the discretion of the OEM, Ragones said.
Like the other stable image programs, Nvidia’s work on a fifteen-month cycle: three months of evaluation, twelve months of production, and a further 24 months of support, managed by the OEM or channel partner. The evaluation window begins on Sept. 1, with production commencing on Nov. 1 – closer to the actual government procurement timetable than what its competitors offer, Ragones said.