Excellent, detailed review with a really exhaustive set of tests. Thanks for taking the time.
Shame about the use of the Asus P7H55D-M EVO in the 1156 testbed though, its used in lots of reviews and seems to skew the power consumption results against the 1156 cpu's. Were AMD gving these things away for free ?

From the SPCR review:

how about some S and T series reviews?

I saw a review somewhere and was quite disappointed. Energy efficiency gain is close to nothing at idle; on load energy drop is much smaller than performance drop

You'll be better off undervolting a regular processor.

Edit: its not that bad, but you still lose in peak performance
http://www.xbitlabs.com/articles/cpu/di ... 2400s.html

I'm shocked that these things are not > $700 USD. I'm waiting to see what Bulldozer brings to the table, but these definitely look attractive! Any word on when the 25nm (was it 25?) Sandy Bridge will be shipping?

Edit: 22nm, and says Wikipedia: "According to the keynote speech presented by Paul Otellini during the 2010 Intel Developer Forum (IDF), Ivy Bridge processors may be introduced as early as the second half of 2011. However, it now seems more likely that Ivy Bridge will be launched a year after Sandy Bridge, probably during CES 2012."

This has been brought up before... if we used the DH55TC, you can say it skews the results the other way. Neither is 'correct'.

Thank you for including the motherboard cost. Quality article, and definitely tailored to our needs more than other review/benchmark articles.

Thanks for that.... i'm thinking for a HTPC system which would be at idle most of the time i think, so the S and T series don't seem to have much benefit at all...

This motherboard has an abnormally high power consumption as I understand, it isnt representitive of the standard sample set as its one of the most inefficient available in its price range. "Correct" would be testing all CPU's on the mostpower effcient board available, which would give you the maximum efficiency attainable rather than a slightly random sample set.

Its easy enough to take a few watts off the 1156 results though, its still one of the most compreshensive reviews of the various platforms.

Another socket!? LGA775/1155/LGA1156/1366... why is Intel doing this?


PS:Every time I buy a heatsink from thermalright or prolimatech I get several sets of mounting hardware that i don't want or need.. Are AMD users subsidizing Intel customers?


//

They have never been good with backward CPU compatibility... but the heatsink mount holes are the same as 1156 -- which is actually just a bit smaller pattern than 1366. 775 - 1155/56 - 1366: They all have the same type of 4 holes in a square pattern. I can see the reasoning for the change from 775 to 1366 -- bigger heatsinks can be accommodated, better cooling, lower noise. Too bad they had to change the pattern for 1156 from 1366 -- the difference is pretty small.

I'd also be interested in reviews about the lower powered 'S' series.

SNB-E for socket 2011 comes out this fall. I personally don't see the point of such a platform when 22nm Ivy Bridge will release for socket 1155 in the Winter. Ivy bridge will have 16 PCI-e 3.0 lanes so SNB-E will only have the advantage of 4 channel RAM and 6 core CPUs. Ivy Bridge will likely get the much higher overclock, so once again enthusiasts will turn down the "enthusiast" platform for the better performing "mainstream" platform.

Ya gotta wonder what Intel are thinking.

They put out the new generation CPUs on the mainstream platform first, which perform better (for the most part) and are much cheaper than their enthusiast platforms which are always half a generation behind.