2.33G Yorkfield spi test 19.750 sec

[zaarath]

Remember the last penryn spi score of 14.xxx without a cpu-z shot?

This time twgsm from xfastest.com brought us a lot more, from Taiwan, obviously.

http://www.xfastest.com/viewthread.php?tid=3055&extra=page%3D1
You need to reply to view the original post so I'll just link the pics as follows, they look REAL, at least to me.

http://pic.xfastest.com/twgsm/12M/2.jpg
http://pic.xfastest.com/twgsm/12M/1.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-02.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-03.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-04.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-05.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-06.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-08.jpg
http://pic.xfastest.com//CPU/Intel/45n/45n-10.jpg

[Bigchrome]

Just over a second and a half faster than conroe clock for clock

[safan80]

any chance we will get to see 7x500 on these chips?

[zaarath]

It's about 10% faster than 4MB conroe clock for clock in spi. I'd say a 5% gain from the additional L2 and 5% from architecture enhancement.

[Cooper]

I would say SSE4 doesn't give any improovements to SPi. It's all in cache/bus speed

[blind_ripper]

that looks bloody fast to me :D

[nn_step]

It's about 10% faster than 4MB conroe clock for clock in spi. I'd say a 5% gain from the additional L2 and 5% from architecture enhancement.

10% faster with 50% more Cache and 15% larger SSE set.

[Jacky]

kinda slow cpumark, only 5% faster clock for clock if I am right...

[Noob-ftw]

Is that 50ºC idle?

Pretty impressive chips.

[revogamer]

wonder how far it'll overclock tho

[FatHippo]

Can't wait to see yorkfield in all of its oc'ed glory!! That looks quite fast for a 2.33 ghz

[massman]

I want to see some oc's :)

[Movieman]

I want to see some oc's :)

Very soon..;)

[wittekakker]

wonder how far it'll overclock tho

The billion dollar question we all want to see answered :cool:

[ineedaname]

hmmm its nice to see improvements but with g0 steppings comin out the door now it pains me to have 2 choose between these 2.

[ludeboy12]

lookin good so far....def holding out for one of these.

[FabricioGS]

Sorry for the dumb question but I'll do it anyway: is it compatible with socket 775?

What about memory support: only DDR3 or DDR2 also?

I'm trying to understand if this is already the new chip from Intel that has the integrated memory controller or it still uses the Conroe architecture only with some improvements.

[largon]

ineedaname,
Yorkfield won't be out until Q4 so there's plenty of time for anyone to play around with G0. Don't you think?


Viper.Br,
Penryn a Conroe shrink/derivate, it has no IMC on it.
It's compatible with DDR2 and DDR3 P35 boards and atleast some i965P boards.

[tvdang7]

WTF? im running those times with 2.8ghz e6300 7 x 400..........damn and thats only running 2.3 something wrong with my set up or something? or is the 2m cache really hurting it.

[nn_step]

WTF? im running those times with 2.8ghz e6300 7 x 400..........damn and thats only running 2.3 something wrong with my set up or something? or is the 2m cache really hurting it.

well have I had to guess, your memory is running slow compared to it

[dinos22]

WTF? im running those times with 2.8ghz e6300 7 x 400..........damn and thats only running 2.3 something wrong with my set up or something? or is the 2m cache really hurting it.

it's just the extra cache man helping in 1M
try comparing 32M and you will see there will be very little difference

[taurus01]

very nice!

[hipro5]

it's just the extra cache man helping in 1M
try comparing 32M and you will see there will be very little difference

Yea......about 45sec in 32M..... VERY little difference..... :D :p:

[AliG]

Just over a second and a half faster than conroe clock for clock
I suppose that's good, but this is why I'm telling people with g0 quadcores to hold tight. Penryn is just a die shrink with sse4 and less transistor leakage, so while performance may slightly increase, the only real benefits are the lower temps and power consumption. And honestly, unless you must have the high end or are a professional ocer like shamino, I think waiting for native octo cores (nehalem) will be the smart move. but this is coming from a person with a small budget and an am2 system

[Fixxxer]

Yea......about 45sec in 32M..... VERY little difference..... :D :p:

Good one! :D

[CyberDruid]

Having already owned and clocked the piss out of that exact chip I can say that 3.220 is the top speed you will likely get from that lil Yorkie. RE temps that CPU will not register temps correctly.

Unfortunately the larger portion of my screenshots dissolved when I toasted my board and lost the RAID array. But I do have a few I posted up before that point.

In terms of parity the 2.33 Yorkfield at 3.2 ghz is producing nearly the same Super Pi and Folding Results as the Q6600 at 3.5 ghz.

I was sort of underwhelmed so I did not even bother to post my results here...just not Xtreme enough IMO.

[hipro5]

You can't say this......The "lower" speed Yorks out there where/are the VERY FIRST batches.....Stepping 1 ..... :D

C1 will be out and we will see what it can do.... ;)

[Xtasy]

i have such some of those cpus too but i have booting problems. the boards needs 10 starts before the cpu is initialized. but i dont know wheres the problem. i used x38 board from gigabyte, dfi p35, gigabyte p35... same problems..

[RPGWiZaRD]

Hmm it doesn't say memory timings and speed anywhere does it? That's also a huge factor to take into account in SPi. Tightening all the subtimings alone gives me a huge boost, especially lowering the default performance lvl of 11 to 7. So therefore I think it's useless to try and make any Conroe clock for clock comparisions with this result.

[AliG]

You can't say this......The "lower" speed Yorks out there where/are the VERY FIRST batches.....Stepping 1 ..... :D

C1 will be out and we will see what it can do.... ;)
If that's the case, then you're correct because the original yorkies had a wall at around 3.2ghz anyways

[CoW]8(0)]

A few things need to be said here.
Why would anyone think SuperPI would gain at all with SSE4? Has anyone been recompiling SuperPI lately?
No it will not benefit AT ALL from SSE4. It doesn't even use multiple cores much less instructions that were never available when it was created.
SuperPI is a rather dated program.

Additionally, the wall isn't with Yorkfield. It's with your boards. Your motherboards can't handle the bus speeds Yorkfield needs to OC.

[dinos22]

^^^ short SuperPi is all about L2 cache
but longer calculations don't really need L2 cache all that much