848P Mobos Introduced: This is basically a castrated single memory channel Springdale. If you care about performance, you buy the real dual-channel deal. Perfectly fine for the Joe Sixpack in your life who wants to save a few bucks, but that’s about it.
ATI wins XBox Competition You get the feeling nVidia is losing it? Like those orphan Voodoo engineers infected them with the disease their former employers had? 🙂
Their market share dropped a bit.
nVidia needs to understand that they are not Intel. They are not Microsoft. In the video card market, it’s easy come, easy go. ATI is, relatively speaking, a much stronger competitor to nVidia than anyone Intel or Microsoft ever had to face.
They’re in a dogfight, and you don’t win dogfights with attitude.
Intel Modifies Its Boards XBitLabs has an article pointing to this Intel document which outlines a number of modifications
to its 865 mobos. A much less descriptive description of changes to 875 boards can be found here.
The article also states that “Intel seems to had (sic) known the possible problems with Prescott support long ago, on the stage of its mainboard product design late last year.” Presumably, they heard that through the grapevine.
Unfortunately, most people reading this would jump to the conclusion that this lack of Prescott support was a deliberate act by Intel.
What is far more likely to have happened is that early versions of Prescott were power-hungry, but Intel felt at the time that they could get the wattage under control by the time Prescott actually came out.
I’m not saying that because I love Intel. I’m saying that because the documentation on the original 865 and 875 chipsets very clearly stated that these chipsets were supposed to support Prescott. Intel didn’t hint or suggest that, they flat out said it.
When you say in writing that A will support B, and it doesn’t, you have problems, and it’s not like you have to be extra evil, too.
This is a little bit different than 820 MCH problem. In the 820 fiasco, the issue was self-contained; it was a mobo division issue. They put out boards people in that division knew had problems. Here, it at least appears at the moment to be a matter of the CPU people saying they could fix it, and failing to do so. That doesn’t make Intel any less potentially liable; you don’t have to be evil to be liable.
Does that mean Intel may have some legal problems as a result of this? That could well be, and if that happens, it could be an ugly fight between Intel and the mobo makers as to who is responsible for what and who promised what.
You might want to download those motherboard datasheets if you own one of these boards.
Outside of that, provided the tale is true, the real moral to the story is that Intel has been trying and failing to fix this for a long time. And that could have some big reprecussions in the months ahead.