3DVelocity would like to thank Intel Corp and especially Mathias Raeck and Graham Palmer for their help and courtesy in providing this motherboard for review.

Setting the scene-

Before I get too involved in looking at Intel's assault on the mainstream buying public, let me first set the scene on both a technological and a personal front, starting with the latter. To attempt to remove any accusations of bias or unfair journalism, let me state up front that I own and use an AMD Athlon 1.4GHz. Does that mean I'm anti-Intel? not even slightly because long before Athlon came onto the scene I was Intel through and through, and I'm going back twenty years here. Intel was very much a part of my formative years on the PC scene, an association you don't get out of your system as a result of a single line of processors like the Duron/Athlon, however impressive they may be. What I'm saying is I like to think I can see both sides of the coin without particularly favouring either, and I hope this standpoint is evident as you read the rest of the review.

When Intel launched its new generation of P7 processors, the Pentium 4, it was far from certain amongst the hardware community that it was going to live up to its hype. Legions of websites queued up to show the world their benchmarks and proclaim that the Pentium 4 was a flop. Page after page of graphs, figures and scores showed the Pentium 4 trailing woefully behind slower clocked Athlons, Durons and its own ancestor the P3. It was heralded a "mistake" a "panic introduction" and was all too quickly dismissed by some as not worthy of further investigation. What was frustrating from my point of view was that although price stopped the P4 finding its way into my personal system, there was a certain level of misrepresentation going on regarding what the Pentium 4 was, and why it was not screaming through the benchmarks the way everybody had assumed it would. Let's take a closer look :

The Architecture -

Technology stands still for no company, and as we've seen on so many occasions you can't always make the next generation embrace the past. When Intel drafted the P4 architecture, they were immediately opening themselves up to the kind of accusations we've seen. In what way? let me explain. Intel are at a distinct disadvantage when it comes to introducing radically new technology in that its components are visible. By that I mean that the individual components need to be purchased separately, an act that immediately draws attention to them. When NVIDIA first introduced the GeForce card, the world got excited about the prospect of DDR. Nobody gave a damn that DDR was more expensive because it came on the card, they didn't have to physically go out and pay for it, yet Rambus is the antichrist. Likewise SSE2, which is the new and optimised instruction set designed specifically to cope with the demands of modern day multimedia by utilising 128-bit SIMD integer arithmetic and 128-bit SIMD double precision floating point instructions. Here again we see sites running benchmarks that are clearly not optimised for SSE/2, then stating that P4's floating point performance is lousy. Well, in fact from own benchmarking I think it's fair to say that the P4's x87 floating point performance when used sans SSE is pretty weak, but isn't this precisely what SSE is there for, to take these complex instructions and crunch through them? Can we really argue that the P4 is lacking because titles aren't yet optimised for SSE2 when we happily pay hundreds of dollars for graphics cards with no current games support or when we openly accepted MMX and 3DNow before titles were optimised to run on them.
No doubt about it, the P4 was a brave, maybe a risky move from Intel, but the faults lie much more in how the P4 has been marketed than in its new architecture.
Another thorn in Intel's side has been Rambus, though I doubt they'd publicly admit to that. I say Rambus have been a thorn in Intel's side, but again if we look at the raw facts, there's no existing memory technology that's better suited to P4's gargantuan 3.2 gigabyte of bandwidth running to the memory controller than RD-RAM. So why was Intel's decision to use RD-RAM so contentious? Well, to begin with it was mainly down to price, though as events unfolded Rambus' aggressive stance against memory manufacturers it claimed were infringing on its patents made it more of a personal dislike for the company than a matter of price or suitability.

Page 2 - Architecture Details and Chipsets