Archived from groups: comp.sys.ibm.pc.hardware.chips (
More info?)
Tony Hill wrote:
> On Fri, 27 May 2005 13:05:11 -0400, Robert Myers
> <rmyers1400@comcast.net> wrote:
>
>
>>On 27 May 2005 07:41:57 -0700, "YKhan" <yjkhan@gmail.com> wrote:
>>
>>
>>>http://www.eetimes.com/news/semi/showArticle.jhtml?articleID=163701592
>>
>>The question is: what will the performance be per watt?
>>
>>Historically, the Via processors have significantly underperformed
>>Intel Celeron ULV clock-for-clock and watt-for-watt. The 64K extra L2
>>cache will help, but it's not a big jump from the 128K L1+64K L2 of
>>previous editions, and it's still in-order, as far as I can tell.
>>
>>Nice that Via is working with IBM.
>
>
> I might just be missing things, but I don't see any mention of core
> improvements beyond the addition of SSE2.
SSE3
NX
faster FSB (up to 800 MHz - probably 4 x 200)
90 nm
much lower idle power than C3 (100 mW)
hardware encryption engine
> The extra cache will help
> and bumping up the bus speed should help as well. I also expect that
> there will be a few tweaks here and there, but it doesn't sound to me
> like we're looking at any significant core improvements.
>
> What this means is that we're probably looking at a 2.0GHz VIA C7
> being roughly comparable to a 1.0GHz Celeron-M, though even that might
> be a bit optimistic.
The FPU now runs full speed instead of the half-speed in the C3,
so that should help VIA close the gap a bit.