But what about P67 chipset which doesn't support SB integrated graphics. Is there any access to QuickSync feature on P67?
Printable View
A very interesting quote from bit-tech have to me a deep sens...
Clock/clock SB is 10-15% more performant than Lynnfield
http://www.hardware.fr/medias/photos...IMG0030494.gif
http://www.hardware.fr/articles/815-...dy-bridge.htmlQuote:
Sans HyperThreading, les gains varient entre 5,6 et 20,4%, pour une moyenne de 13,2%. Avec HyperThreading, Sandy Bridge est 5,6 à 19,3% plus rapide, pour une moyenne de 11,3%
But if we judge about oc frecuency, from what i saw until now i5 2500K, and i7 2600K can do 4.8Ghz(may be even 5Ghz) 24/7, 600Mhz more than an average i5 750/870 - 4.2Ghz than we have also frecuency performance 19% more with SB.
So adding 10-15% + let say 20% we have 30-35% more performance with SB than with Nehalem.
And the quote from bit-tech
http://www.bit-tech.net/hardware/cpu...idge-review/14Quote:
However odd it sounds to be talking of an overclock of up to a frequency, it’s truly scary for AMD fans that a low-end Sandy Bridge CPU can hit speeds of up to 3.99GHz with no extra CPU voltage – it suggests that there’s masses of frequency (and therefore performance) headroom in the Sandy Bridge architecture.
You know me, I don't usually do single socket builds but went after this after hearing some rumors 2 months ago as to the efficiency and power in this little chip that I saw as a mid range replacement for the P55's.
What a pleasant surprise I had to find that all the rumors were true.
Beastly computational power and incredible electrical efficiency.
You guys hit it right with this one..
Now, how about a dualie version?:rofl:
For you those who couldn't tell movieman isn't interested in sandy bridge :rolleyes:
PS: click on his profile, find more posts...
dumb question time: socket mounting system the same as 1156? or 1366? (re: coolers) ;)
Let's not forget about the dual core 2100 being on par with older i7's and AMD hex/quad CPUs in games. In some games, it even beats them by a pretty fair amount. Sandy Bridge is not only a winner for overclockers, but gamers as well.
> The test of win.
MM is right, again. The 32nm SB is cool and power-efficient, and has a better OC-headroom (compared to good old 45nm i7). With this price, there is no doubt it's a good step forward.
Those who are expecting too many wonders, should notice that these are not high-end CPUs.
It would be nice to see some double-socket, or maybe even quad-socket SB tho, but maybe we have to wait for the high-end?
Dave, just show up a multi-CPU LGA2011 rig already :rofl: :up:
5Ghz is rare.. extremely rare if you talk about 24/7 which would imply <1.4V. only thing i saw reaching 5Ghz or higher was all above 1.5V so not even in the facinity of 24/7.
but SB does have alot headroom, they could go up 600Mhz (if the tdp wouldn't be in the way combined with the gpu).
Apparently you missed my review ;)
5.1GHz stable with 1.45V :up:
http://www.techreaction.net/2011/01/...n-p67-part-12/
Not trying to imply that it'll be common....but there are chips that are capable. ;)
I forget which review mentioned it but they said based on some 3rd party testing of about 100 retail SB skus, that something in the area of 10% were able to get a stable 5Ghz clock with about 50% being able to do 4.4-4.7. Forgot which review though :rolleyes:
I would like to see what was lowest voltage they could achieve on stock clock/turbo :up:
Then of course measure the wattage
I will when I can get my hands on one..
Maybe 3-5 months for that to happen..if I'm lucky.
Think on this:
SB at 3500 is equal to a i7-920 at 4Ghz in computational powerand SB should have no issues running at least 4500 for 24/7 usage.
Now will that be enough for you?:rofl:
I saw it too and now can't find it anywhere. I think it was Hardwarecanucks, but it isn't there now. Maybe it got pulled for "some" reasons. :rolleyes: To be fair, there wasn't a full description provided, how those test were concluded or what voltage they used as maximum, so those might not be showing a true picture of the situation.
It also doesn't converge with the reports of getting quite high clocks with stock voltage, which is to be expect when you power gate the IGP off and have some TDP headroom.