New Prescott & Northwood/ A64 Benchmarks

[Ray Austin]

The Benchmarks

Flicker, flicker ~Intel was here.

[schmosef]

Fascinating. Thanks for the link.

[Sciascia]

Very interesting. So at the same clock speed it is actually slower, but we also have to take into account that there are more instruction sets that aren't utilized right now. Personally the power dissipation would be the killer for me, PC's are already too loud and with that much more heat they will get louder unless the BTX form factor really works.

I suffer from terminal upgraditis, but with these numbers I won't be changing anything for a while. I mean we are talking about maybe 4GHz and a lot of programs that can use SSE3. PCI Xpress will have to be involved as well. I also wonder if there will be a new chipset out anytime soon just for the Prescott? That could make a huge difference.

[Kris1]

For what it's worth, a bit of old news : (articles at these guys always should be handled with care)

http://www.theregister.co.uk/content/51/35535.html

(new Prescott stepping, lower temp and a couple of other enhancements)

It stands to reason that Intel will have to adress the hot-headed nature of the Prescott if they want to keep up with their roadmap. Also I've been told unofficially by an Intel representative that the reason for the delay between the sample shipments and the retail shipment involves the length of the validation-procedure for these new parts.

Also, other tests on the net showed the Prescott "breaking even" with Northwood around 3.6Gig (checkout Anandtech for example). After that, the relative cpu-power gain ramps faster than with Northwood, in other words, the Prescott scaled better and showed a faster increase in speed in function of Frequency raise when it was being overclocked then a comparable Northwood CPU did.