Thanks. again, my 680 will report 13 invalid tasks. The only difference I (emphasis on I) could see was a different driver. Mine was 301.10 and his was 301.25, so I just switched to this new driver.

Results don't look too bad overall though (as far as running to completion is concerned) Keep up the great work!!!

In regards to dp, that's the thing I could think of, albeit very little apparently. FP32 cuda projects are showing a 25% inc. Over 580 and 44% over 570.

On my I5@4.2 they were running at speed as posted on average.

EDIT: the statement that it LIMITS it's FP64 precision is also SOMEWHAT incorrect. It doesn't limit it like it does on other series. It's that the Kepler series only has 8 FP64 cores on the board. They're actually not even added in the core count. They're completely seperate.

The other change coming from GF114 is the mysterious block #15, the CUDA FP64 block. In order to conserve die space while still offering FP64 capabilities on GF114, NVIDIA only made one of the three CUDA core blocks FP64 capable. In turn that block of CUDA cores could execute FP64 instructions at a rate of ¼ FP32 performance, which gave the SM a total FP64 throughput rate of 1/12th FP32. In GK104 none of the regular CUDA core blocks are FP64 capable; in its place we have what we’re calling the CUDA FP64 block.

"The CUDA FP64 block contains 8 special CUDA cores that are not part of the general CUDA core count and are not in any of NVIDIA’s diagrams. These CUDA cores can only do and are only used for FP64 math. What's more, the CUDA FP64 block has a very special execution rate: 1/1 FP32. With only 8 CUDA cores in this block it takes NVIDIA 4 cycles to execute a whole warp, but each quarter of the warp is done at full speed as opposed to ½, ¼, or any other fractional speed that previous architectures have operated at. Altogether GK104’s FP64 performance is very low at only 1/24 FP32 (1/6 * ¼), but the mere existence of the CUDA FP64 block is quite interesting because it’s the very first time we’ve seen 1/1 FP32 execution speed. Big Kepler may not end up resembling GK104, but if it does then it may be an extremely potent FP64 processor if it’s built out of CUDA FP64 blocks."

!

@knreed- FWIW 15 of the 31 errors that were reported for the GTX460 hosts came from 1 of my machines. I took that machine off line long enough to troubleshoot a few things. I put it back online after reinstalling the drivers and rebooting and it completed 20 tasks w/o any errors. Maybe this can help some others that had error problems.

knreed,
These Beta's are working Great for me ! Thanks for the effort.
I am quite ready to start running them fulltime when ever ya'll are

Note:
I noticed there is only one version of the 560 Ti listed.
You may want to note the 448 version is a very different card.
They really should have called it 565 Ti

While the GTX 560Ti is based on the GF114 graphics core , the 448 Core 560Ti is based on the GF110 architecture shared by the GTX 580 and GTX 570
-----------------------------------------------------
The 560 Ti is identified in the result log as
CL_DEVICE_GLOBAL_MEM_SIZE: 993 MByte
CL_DEVICE_COMPUTE_CAPABILITY_NV: 2.1
-----------------------------------------------------
The 560 Ti 448 is identified in the result log as
CL_DEVICE_GLOBAL_MEM_SIZE: 1280 MByte
CL_DEVICE_COMPUTE_CAPABILITY_NV: 2.0
-----------------------------------------------------
and at least in the case of my EVGA Classified version power is handled very differenty...
"6-phase digital PWM (compared to four from the reference design)"