Nvidia may develop an integrated x86-based chip for use in low-cost computers, an Nvidia executive said this week, a move that would step up its rivalry with Intel.
Nvidia is considering developing an integrated chip based on the x86 architecture for use in devices such as netbooks and mobile Internet devices (MIDs), said Michael Hara, vice president of investor relations at Nvidia, during a speech that was webcast from the Morgan Stanley Technology Conference on Tuesday.
Why would NVIDIA do this? Because, as a recent and widely linked Jon Peddie report correctly points out, the integrated graphics processor (IGP) will eventually go away, as integrated GPUs move off of the chipset and onto the processor die. IGPs are a huge and very successful part of NVIDIA's volume business, and the company can't afford to see that segment left solely to Intel at a future date when all mobile x86 processors come with an on-board GPU.
If a smartphone maker like Nokia goes with ARM, then it can choose from a variety of suppliers like TI or Qualcomm. This competition among sources is healthy and gets the handset maker a lower price; it also guarantees a supply of ARM chips, because if one supplier goes under you can just move on to the next. Intel can't offer this with x86, but NVIDIA as a credible second source of x86 mobile parts might be enough to get the world's most successful ISA into the phone market.
Before everybody gets their knickers in twists, NVIDIA is not going to show up in your general purpose desktop any time soon. Hara explained that it would most likely be a system-on-chip design for MIDs and netbooks. He spelled out a time frame of two to three years.
There isn't a firm timescale on such a development, though Hara suggested that "two or three years down the road I think it's going to make sense" but adding: "we won't talk much more about what we think about that timeframe, but there's no question it's on our minds."
Such speculation is all well and good; Nvidia already has SoC solutions using ARM CPUs, and it seems likely that tweaking those designs to accommodate an x86 processor wouldn't be particularly complicated. However it does gloss over the slight issue that Nvidia doesn't own an x86 license, which it needs if it wants to sell a CPU.