8 hours without error (but also decreased memory clock to 2500 Mhz)

[QUOTE=Redarm;332510]8 hours without error (but also decreased memory clock to 2500 Mhz)[/QUOTE]
The ether was wearing off. The acid was long gone. But the mescaline was running strong. Good mescaline comes on slow.

I realize CuLu and mfaktc could benefit the most from this new board, but
I hope someone gives mmff a shot before the boards get returned for refunds.

Good idea
 

[QUOTE=Redarm;332510]8 hours without error (but also decreased memory clock to 2500 Mhz)[/QUOTE]

I'm testing the same now and think this could really be it with regard to the symptoms.
Error free for 4 hours.

[QUOTE=Redarm;332510]8 hours without error (but also decreased memory clock to 2500 Mhz)[/QUOTE]

It would not be the first case of an nVidia chip with over-rated VRAM.

30% no error

[QUOTE=Redarm;332568]30% no error[/QUOTE]

Still no errors. :-) Redarm, you're the man.
God bless the Titan!

How is the speed with the lower clock rate?

[QUOTE=Brain;332682]Still no errors. :-) Redarm, you're the man.
God bless the Titan![/QUOTE]
Hmmm. So, buying a Dodge Viper and then driving it at 20mph (because otherwise it stalls or blows up) is now an achievement?
:confus:

[QUOTE=ATH;332699]How is the speed with the lower clock rate?[/QUOTE]

4.25ms for 57M @80% power target with 3.6 M FFT.
100% power target cannot be sustained cause of temps.

3.4 ms with 2750 mhz 100% PT 80°C Fanspeed 60%

[QUOTE]Hmmm. So, buying a Dodge Viper and then driving it at 20mph (because otherwise it stalls or blows up) is now an achievement?[/QUOTE]

more likely an injection throttlement so that you drive it with 180mph max