Not anything about the product, but the Intel briefing is full of stuff with nebulous numbers and non-specific terms.
Here, ny main beef about the things Intel keeps droning on about is that it still steadfastly refuses to give any sales numbers other than painfully vague system revenue sales. No one will take it seriously until it puts out core sales numbers.
I will spare you the rest of the Zerg rush "we are going to take over the world" analogies, as I said, hard CPU numbers or you won't be taken seriously.
One interesting note is that it is pushing RAS over CPU power. This is because it has RAS, a real benefit, but are going to get blown out of the water by the eight core Beckton on CPU power. When you make a 500mm x86 die, well, IPF pales, so retreat to the high points.
There was a lot of flowery talk about OSes and software, but since Intel insists on using proprietary formats to relay this information, all I got was a jumbled mess. Until it puts out hard numbers however, I don't believe any of it anyway. Did I say that already?
Technical tidbits, Poulson will be all new and built on a 32 nanometre process. Pulling IPF onto the current process will help quite a bit, but will anyone still care when it comes out? This represents a fundamental change in the IPF philosophy, but so?
Itanic is such a compelling product that SGI is dropping it from most of its line in the CSI time frame, relegating it to the lowest end clustered boxes. If that isn't a show of confidence, I don't know what is. Did I say I'd like hard numbers, already? µ
* INTEL HAS NAMED the successor to Intel's Itanic Poulson part, it is called Kittson. Yay. µ
Sign up for INQbot – a weekly roundup of the best from the INQ