before

Today, I've broken the 3DMark2001 SE' 40k barrier with a X1900XTX @648/774MHz stock cooled and the Opty 146 @3.4GHz under Cascade...

@PL4YD34D: yeah row cycle time is @0 clock @loc.o: Very difficult to have -7°C while average core temp is -60°C with my cascade... BTW this chip won't do anything good at -60°C; it seems to be cold-bugged at 355HTT @-25°C so with SMP mutli CPU enable, extreme vcore and no thermal compound between the evap and the IHS, I can get some nice benchies BTW I'll turn my chiller into direct on die unit for benching A64 chips. :nod:

Thank you! 1.68V is very low... 3632.84MHz @1.84V - Cascade: CPU core @-7°C http://valid.x86-secret.com/show_oc?id=89382 Thx dude! The liquid used here was Glycoshell... This chip isn't so much cold-bugged, but actually its very difficult to run it under cascade. 23.875s - 3505MHz @1.84V - BH5 2*256Mb @250MHz 1.5-2-2-0 @3.7V

Opteron 146 CABYE 0540FPBW - Liquid Chiller CPU@+3°C http://valid.x86-secret.com/show_oc?id=89153 More to come :nod:

[email protected] under water

improved! [email protected] chiller CPU@+11°C http://valid.x86-secret.com/show_oc?id=79603

I've received the FX back today... Bob told me that he didn't succeed in validating above 3.6GHz; that's not bad hehe, but not impressive considering the stepping of the chip; Tanker will send me my cascade on next Wednesday, so I'll soon run it myself under cascade...

In personally prefer StressCPU to assess for overheating issues. This is a small program to torture-test your CPU in order to make sure that you don't have overheating problems. It will only run on SSE-equipped x86 CPUs, and it is executing a special version of the Gromacs innerloops that mixes SSE and normal assembly instructions to heat your CPU as much as possible. On the other hand, it can be used as a benchie; and seconding me, Fouge @OverClocking-Masters.com has modded the program that gives now computation time as well as CPU clock speed. There are 3 choices: 10K iterations which is like Super Pi 1M, 100K iterations which is like Super PI 8M, 1M iterations which is a real stability test during about 1 hour. http://olivier.fouge.free.fr/Bench/StressCPUBench.zip I've started a database for this bench, some of you may wish to participate here. http://www.beforeoverclock.com/b/showthread.php?t=33 :cat:

Very nice CACJE 0603FPMW :nod: and a good Spi32M time :angel:

Those CACJE 0603FPMW are pretty good in watercooling conditions... I've pretested 3 of the 10 in stock here; and all of these do 3GHz @ 1.55V or 1.52V... but under phase change cooling, they are probably the worst ever because of cold-bug at approximately 10°C :drool:

I've some CPUs of this stepping to pretest; didn't test with less or more volts, but [email protected] is pretty good actually

Some fresh results CPU Max Screen @1.81V = 3444MHz chiller CPU@6°C | http://valid.x86-secret.com/show_oc.php?id=76935 StressCPU 10 000 Iterations 27.03s - [email protected] CPUMark99 421 - [email protected] Super PI 1M 27.734s - [email protected] - BH5 @267MHz 1.5-2-2-6 3.7V

15MHz more with .01V and 4°C less [email protected] | http://valid.x86-secret.com/show_oc.php?id=75997

Thx mate! Unfortunately, the chiller doesn't help me much for stable speeds because load temps do not stay descent (for a chiller of course). Anyway, the cpu passed OCCT at [email protected] with a full load temp of about 33°C.

a little higher clock for this 146 CAB2E 0546GPAW... :nod: [email protected] | http://valid.x86-secret.com/show_oc?id=75504