65nm Pentium D 950 with 4MB L2

[Salvador]

.... There were special on-die thermal diodes, and Over head management components that needed to be added, they even orginaly wanted to use indpendent busses for each core, however that now won't appear until paxville...

You do mean the Dempsey, right? With dual 1066FSB.

Now bench this baby.. Have been gone too long now!

[Kadaj]

You do mean the Dempsey, right? With dual 1066FSB.

Now bench this baby.. Have been gone too long now!

technically conroe is supposed to impliment dual 1.33ghz FSB's, but we'll have to see.

[Salvador]

technically conroe is supposed to impliment dual 1.33ghz FSB's, but we'll have to see.

Heh, but the Dempsey is to arrive long before Conroe.! So it's (Dempsey) the first one to have dual FSB

Dempsey-> Q1 2006 (jan)
Conroe -> Later Q2-Q3(?)

[Kadaj]

Heh, but the Dempsey is to arrive long before Conroe.! So it's (Dempsey) the first one to have dual FSB

Dempsey-> Q1 2006 (jan)
Conroe -> Later Q2-Q3(?)

yea, but the difference is that conroe is going to cost several hundred dollars less, and work with rev 2.0 i975X chipsets

Looks like dempsey has been leaked already
http://www.tecchannel.de/server/hard...57/index2.html
its in german, but they're claming they have performance results for a dual 1.06ghz FSB sequence 5000 dempsey.

[Salvador]

yea, but the difference is that conroe is going to cost several hundred dollars less, and work with rev 2.0 i975X chipsets

Looks like dempsey has been leaked already
http://www.tecchannel.de/server/hard...57/index2.html
its in german, but they're claming they have performance results for a dual 1.06ghz FSB sequence 5000 dempsey.

There are to other sites having testresults aswell.!

http://www.realworldtech.com/page.cf...WT110805135916
(Only Est, but part 2 is real benches)

http://www.2cpu.com/review.php?id=110

[Cooper]

Well Yonah working prototypes showed off already in the summer. So I`m not surpised that there`re already few working Conroes.
And NOT every presler/cedar mill will have vanderpool tech. Actually only few companies/ppl really need it. I`ve read that 65nm NetBurst family won`t have vanderpool, but i doubt it.
P.S. there are already two 6xx models available with vanderpool tech.

[i found nemo]

those chips mop the floor with the loe-end opty

[Kadaj]

Well Yonah working prototypes showed off already in the summer. So I`m not surpised that there`re already few working Conroes.
And NOT every presler/cedar mill will have vanderpool tech. Actually only few companies/ppl really need it. I`ve read that 65nm NetBurst family won`t have vanderpool, but i doubt it.
P.S. there are already two 6xx models available with vanderpool tech.

defentaly good news, from all the sources I read it gave me the impression that it was going to be as standard as HT.

[AkXb70]

i dont see the big deal with enabling/disabling VT...if its there it should be on. its not like it hurts the performance of the chip any or makes it more expensive to produce (apparently its already in the core since presler has it), and if you dont need it, you dont have to use it

[hicookie]

update pi 1m with stock cooler


CPU: INTEL 950
RAM:trandscend ddr2-667 1g*4
mobo:P5WD2-P
cooler:stock cooler
power:echance 600W



4.5G runs sp2004 only 44 C

[16floz470ml]

That is a really low temp for a stock cooler.

[HawainPanda]

holy....almost crapped in my pants :P that's some awsome temps at nice ocs

[HawainPanda]

can you try running pi on both cores? I heard that the first intel D chips automatically reduced the clock of the second core when temps got too hot, is it the same with the new chips? also any idea how much these suckers will be?

[Coustang]

This may seem like a lame question, but I've been out of the loop for a while...

What socket are these new CPUs going to be running on? 775/478/somthing else?

[Dumo]

Welcome to XS Coustang

65nm ..This will be interesting..

[Salvador]

This may seem like a lame question, but I've been out of the loop for a while...

What socket are these new CPUs going to be running on? 775/478/somthing else?

LGA775.
i975X chipset for the 955XE processor.
i955X (i945) and i975X for the other 9xx processors?
Correct me if I'm wrong


Also heard something about i865P "supporting" 955XE
http://www.hkepc.com/bbs/viewthread.php?tid=506271

Anyway, thats a amazing clock on stock cooler. And nice low temps!

[Salvador]

Came over som benches comparing the 955Xe and the 840Xe.
http://www.hkepc.com/hwdb/65nm975x-6.htm

[dpa]

looks like SLI works on the 975xe?

[Pressure]

looks like SLI works on the 975xe?

It does, along with CrossFire. Although, not officially yet.

[Sesto Sento]

I think that processor is screaming for some LN2

[Kadaj]

looks like SLI works on the 975xe?

955X supports SLI and crossfire, the only problem with SLI is driver support, which isn't there.

i dont see the big deal with enabling/disabling VT...if its there it should be on. its not like it hurts the performance of the chip any or makes it more expensive to produce (apparently its already in the core since presler has it), and if you dont need it, you dont have to use it

no, but VT chips in i955X have caused some stability bugs that may or may not be fixable(if they create a whole series w/o it though, my guess is it's not), which could lead to some chips having VT and others not (so they may work in 955X and earlier chipsets)

[AkXb70]

no, but VT chips in i955X have caused some stability bugs that may or may not be fixable(if they create a whole series w/o it though, my guess is it's not), which could lead to some chips having VT and others not (so they may work in 955X and earlier chipsets)

backwards compatibility would be a good reason for leaving it out. i didnt think intel was going to allow that though

[amd3001]

Do those 950's have hyperthreading also? Will these 950's run on a 955x chipset?

[AkXb70]

Do those 950's have hyperthreading also? Will these 950's run on a 955x chipset?

the 950's do not have HT...only the EEs will have HT (the 955EE will be a 3.46GHz 1066fsb part)

[Sentential]

Hmmm..... seems like my interest for i975X may have actually increased after reading HKEPC's article.

(EDIT I will have an i975X to test very soon)