• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Sandy Bridge Preview

I want socket 1567 Westmere-EX (Eagleton) -- 10 cores, 20 thread, up to 8 sockets.
 
Read it earlier....
quad channel ram, I shudder at the thought. :eek:
Shame the overclocking will be pretty locked down on everything except the K line cpus, but as Anand hypothesized, hopefully the MSRP on the K chips stays around what people pay for i5-750s and such now anyway, so hopefully it won't be so noticeable.

Once thing's for sure....Sandy Bridge means no more Pentium Dual-Core type goodness. No more "buy a $50 cpu, oc it to $150 cpu levels". :( That makes me a sad panda.

Sounds like as a whole, SB won't really benefit folding at least not until the really ridiculous cpus come out and they start talking about octo cores with sixteen threads (which you know is coming). I say that because of this:
Anand said:
While Nehalem was an easy sell if you had highly threaded workloads, Sandy Bridge looks to improve performance across the board regardless of thread count.
 
Read it earlier....
quad channel ram, I shudder at the thought. :eek:
Shame the overclocking will be pretty locked down on everything except the K line cpus
I'm still hoping the socket 2011 performance lineup won't be locked down. I would wager those will be more overclocked friendly. Nobody knows for sure yet though.
Posted via [H] Mobile Device
 
I'm still hoping the socket 2011 performance lineup won't be locked down. I would wager those will be more overclocked friendly. Nobody knows for sure yet though.
Posted via [H] Mobile Device


should be.. if its not then intels fuxor'd.. but should only be the low end ones that have the overclocking locked.. most likely because of the onboard IGP..
 
Sounds like as a whole, SB won't really benefit folding at least not until the really ridiculous cpus come out and they start talking about octo cores with sixteen threads (which you know is coming). I say that because of this:
I don't know where you get that idea from, but the benchmarks it definitely looks like Sandy Bridge will bring improvements to folding performance since it does much better than Nehalem in other CPU-intensive tasks.
 
the chip looks to be a powerhouse and soo long as the K chips are in a good price range all is good. if i remember what i read correct the 4 core wasnt to fare behind the 980X and was at a much lower clock.
 
I don't know where you get that idea from, but the benchmarks it definitely looks like Sandy Bridge will bring improvements to folding performance since it does much better than Nehalem in other CPU-intensive tasks.
You can't overclock SB, though...so those 3Ghz SBs will have to compete with 4Ghz Nehalems.
 
You can't overclock SB, though...so those 3Ghz SBs will have to compete with 4Ghz Nehalems.


no ones really confirmed that.. the only thing that's been confirmed is that the lga-1155 Sandy Bridge non K series cpu's will not be able to overclock.. there has been absolutely nothing confirmed about the lga-2011 Sandy Bridges which we wont know for another year or so..

not to mention all the K series chips will be able to overclock since they have the unlocked multiplier..

why everyone's still stuck on this Sandy Bridge cant overclock craps beyond me..

the thing people should be more pissed about is that intel still only supports 2 sata 6Gbps ports..
 
You can't overclock SB, though...so those 3Ghz SBs will have to compete with 4Ghz Nehalems.
I posted about that a few weeks ago but we don't have all the details of the new architecture yet. It may be the case for some of the SB product line, not necessarily for all (I hope not).

Another thing that should be mentioned is the higher memory performance potential of S2011. From mine and others' observations, -bigadv is largely affected by memory subsystem performance. Much more than standard A3 SMP or any SMP WUs of the past. That alone can make a big difference with the new architecture whether it can OC or not. Something to ponder..
 
the thing people should be more pissed about is that intel still only supports 2 sata 6Gbps ports..
They support as many as the motherboard manufacturer decides to put on the board, though I don't see what the big deal is since no current HDs or SSDs have average read or write speeds that come near maxing out SATA2.
 
They support as many as the motherboard manufacturer decides to put on the board, though I don't see what the big deal is since no current HDs or SSDs have average read or write speeds that come near maxing out SATA2.


Both H67 and P67 support 6Gbps SATA, however only on two ports. The remaining 4 SATA ports are 3Gbps.

the chipset its self will only support 2 sata 6Gbps so the only way it would support more is using a third party chip.. which quite frankly is dumb since AMD already supports 6 of them.. while yes i agree a single hard drive or SSD doesnt max out sata 2.. sata 6Gbps even on those have showed a slight increase in performance and an even bigger increase when running raid.. the added problem to this is that if intels going to half way support sata 6Gbps then why would drive manufactures waste their time on developing sata 6Gbps hardware to take advantage of the speed increase..
 
I don't know where you get that idea from, but the benchmarks it definitely looks like Sandy Bridge will bring improvements to folding performance since it does much better than Nehalem in other CPU-intensive tasks.

I gave my source. ;) Folding is highly threaded, is it not? :)
 
I gave my source. ;) Folding is highly threaded, is it not? :)
And? The quote you posted says that Sandy Bridge brings performance benefits to both multithreaded and non-multithreaded workloads. That means folding performance should improve as well.
 
I'm just sad that LGA 1366 is finally approaching its EOL. At least we know that AM3 has possibly significant life left as they will get some variant of Bulldozer according to the previews we saw last week.
 
I'm just sad that LGA 1366 is finally approaching its EOL. At least we know that AM3 has possibly significant life left as they will get some variant of Bulldozer according to the previews we saw last week.
1366 still has a year before its EOL
Posted via [H] Mobile Device
 
it works on AM3+, older cpu's with work on the newer boards. http://en.wikipedia.org/wiki/Bulldozer_(processor)
so in other words, no, bulldozer will NOT work with our current AM3 boards. I fail to see how "AM3+" is terribly beneficial other than the rare case of somebody buying a new mobo to use with an old cpu.

Also, still only dual channel DDR3? When LGA2011 will bring quad channel ram to Intel?
 
Ahh, I must have read it wrong. I was all happy thinking that my Gigabyte GA-790XTA-UDR4 in my gaming rig would be Bulldozer compatible one day.
that would have been pretty amazing.

I wonder though if this backwards-compatibility emphasis is really good for AMD. Maybe it's time to ditch AM2/AM3 and move to a triple channel or quad channel RAM interface.

I would be extremely curious to know if the performance of an AMD rig would change depending on the amount of memory bandwidth available to the CPU. If anybody has a quad channel board that they would like to run some bigadv tests on, in quad- and dual-channel modes, that would be great :)
 
I wonder though if this backwards-compatibility emphasis is really good for AMD. Maybe it's time to ditch AM2/AM3 and move to a triple channel or quad channel RAM interface.
The way I see it, AMD needs to improve memory performance if they wish to keep up with Intel. Not stating that their best technology currently available is poor in that regard, but a new balanced architecture release demonstrates improvements in every area. In the past, substandard memory performance has proven to be a bottleneck in an otherwise good CPU architecture.

I would be extremely curious to know if the performance of an AMD rig would change depending on the amount of memory bandwidth available to the CPU. If anybody has a quad channel board that they would like to run some bigadv tests on, in quad- and dual-channel modes, that would be great :)
The only person who could demonstrate AMD quad-channel memory performance on this forum TMK is 10e.
 
I still say if there was a mobo that could overclock g34 by even a modest amount, I would be tempted to try it myself....
 
I'm just sad that LGA 1366 is finally approaching its EOL.

I wouldn't say that.

AMD has left AM2 around for a long time probably because they've been in 2nd place so they've had to try harder to keep fans from switching. If it wasn't for that, I don't see how AMD wouldn't be doing anything differently from Intel, which is the usual "new architecture means new sockets".

I don't expect one without the other, most other people don't. If you get lucky, you get lucky - I don't count on it though. 1366 still has more legroom than 1156; Intel's said that themselves. Anything high end will be 1366 only. I'm still hoping to see cheaper hexas in the next 6-9 months, but who knows whether that will happen or not. It may happen once the higher end SB is a lot closer to fruition, but we won't know until then.

Hell I'm still happy with both of my s775 systems because they serve their purposes and do what I need them to do. Sure I wish I could get more PPD out of them, and I could (if I wanted to go to MC and pick up a quad and swap out my e8400)...but money's too tight right now and we've got less than a month before baby girl #2 so $ is really tight. ;)
 
Back
Top