From the very beginning of the nForce (aka Crush) times back in late 2000 different mass media agencies have been talking about NVIDIA nForce chipsets for Intel Pentium III or Intel Pentium 4 processors. However, the company itself has never confirmed or denied the existence of such project claiming in unofficial talks that its profitability would have been too low. These days the rumours about an nForce chipset for Intel CPUs started to appear again on different web-sites, including NFI, GZeazy.com and so on. Let us try to summarize the unofficial facts and figure out possibilities and consequences of such product.

First of all, everybody understands pretty clearly that there are hardly any problems with technological implementation of any processor’s bus in core-logic products. Even though the difference between Quad Pumped Bus utilized by Intel NetBurst processors, EV-6 used by AMD K7 CPUs and HyperTransport interconnection of AMD K8 (Hammer) chips have tangible differences, experience of VIA Technologies, SiS and ATI Technologies show that it is possible to develop various types of chipsets.

Secondly, a concern that appears from the second paragraph is whether a company is able to successfully market a number of chipset families or not. Well, we do know that VIA used to be very successful with its chipsets for Intel Pentium III and AMD Athlon processors when their products were powerful enough. ATI Technologies who announced its RADEON IGP desktop and mobile chipsets for AMD Athlon and Intel Pentium 4 processors last year could not become big in all market segments, but managed to occupy 1/3 of mobile integrated graphics market. This year ATI is expected to put some pressure on the desktop integrated chipset market as well. All in all, it is not possible to become successful rapidly, but a company that left a footprint in one market segment can also start attacking another hoping to broaden its market presence.

Thirdly, NVIDIA is now a provider of one of the most sophisticated core-logic families for AMD platform and it would be logical for the company to try expanding its chipset market share by addressing Intel CPUs as well.

One thing that holds NVIDIA nForce2 from becoming extremely popular is its pricing. It is still more expensive compared to solutions from VIA and nothing helps to move the nForce2-series to the mass market quickly. NVIDIA has very good position to occupy less price-conscious K8 market, however. It seems that the company will become big in this market; nevertheless, since its revenues from graphics processors have been falling for months, NVIDIA has to do something with it in order to maintain its level of earnings. This “something” may be a chipset family for Intel Pentium 4 processors.

Intel is a company that dislikes all rivals and all possible rivals. It does not have any rights not to license its QPB to NVIDIA, but it may charge the Santa Clara, California-based graphics firm so heavily for its Quad Pumped Bus technology that there will be almost no sense for NVIDIA to make chipsets for NetBurst CPUs. However, given that the ramp of the GeForce FX products rapidly shrinks NVIDIA’s margins, the company may reconsider its strategy and start making core-logic sets for Intel CPUs at some point. At least, some money will be earned and there will be even fiercer competition in the integrated Socket 478 market what seems to be negative for the arch-rival ATI.

Keeping in mind that NVIDIA has won a number of designs in the AMD market, it has all chances to win another bunch in the Intel market and it is very likely to happen next year, roughly three years after the initial announcement of the nForce core-logic for AMD Athlon processors. At that time the market will consider NVIDIA as a provider of stable, fast and reliable products, just what is needed. Nevertheless, I believe that NVIDIA should roll-out a number of solutions to fit into different segments from the very beginning of its Socket T market penetration.