[QUOTE=xilman;258420]Ah, it worked when all four results were submitted.

Very strange. Something took a dislike to me the first time around. No matter. It appears that I've 4.2387 GHz days to my credit. For an expenditure of 0.0622 days, the GPU seems to have an effective clock rate of 68GHz.

Paul[/QUOTE]

For a GTX460 - that looks like you need to throw another core at it, and run another instance. Well to maximize it anyhow.

-- Craig

[QUOTE=nucleon;258472]For a GTX460 - that looks like you need to throw another core at it, and run another instance. Well to maximize it anyhow.

-- Craig[/QUOTE]

Yes, but if it runs correctly, he's still running circles around any CPU you want to talk about!

I've got some stuff I need to do (including new computer and answer SM88, since he's at least trying to contribute compute cycles) before I get a GPU.

How's the next release coming?

[QUOTE=nucleon;258472]For a GTX460 - that looks like you need to throw another core at it, and run another instance. Well to maximize it anyhow.[/QUOTE]Ok, you talked me into it. Just grabbed another two exponents to see how well it goes with them. A few minutes in and it looks as if the pair of them will take around 25-30 minutes. Two exponents running serially took 45 minutes yesterday.

Doesn't seem to have fried the display yet, though the gpu temperature has reached a maximum of 62C with a fan speed of 52% --- well within acceptable limits IMO.

Paul

[QUOTE=xilman;258502]Ok, you talked me into it. Just grabbed another two exponents to see how well it goes with them. A few minutes in and it looks as if the pair of them will take around 25-30 minutes. Two exponents running serially took 45 minutes yesterday.

Doesn't seem to have fried the display yet, though the gpu temperature has reached a maximum of 62C with a fan speed of 52% --- well within acceptable limits IMO.

Paul[/QUOTE]

Are you also blowing enough air into the box to cool the mainboards interconnects?

This as mfackt will fry the interconnects of course as it ships loads of data to the GPU.

There is no temperature sensors at the interconnects, but touching careful with your hand you will be able to confirm me that you can cook an egg on 'em.

In case of the AMD 4 socket board here, i've put 4 additional fans on the 4 HT links. They did have passive heatsinks, but that wasn't enough.

In case of emergency you can still use a standing fan and blowing into the open case with it. Creating overpressure inside the box is what you'd want i guess.

[QUOTE=xilman;258502]Doesn't seem to have fried the display yet, though the gpu temperature has reached a maximum of 62C with a fan speed of 52% --- well within acceptable limits IMO.

Paul[/QUOTE]

Maybe you should increase fan speed manually? When I ran mfaktc I set fan speed to 80% and made a rule that if temperature exceeded 60C fan speed increased to 100%.

Isn't it best to use the fan to its max to keep GPU temperature low?

[QUOTE=ATH;258506]Maybe you should increase fan speed manually? When I ran mfaktc I set fan speed to 80% and made a rule that if temperature exceeded 60C fan speed increased to 100%.

Isn't it best to use the fan to its max to keep GPU temperature low?[/QUOTE]

If you keep it real low it will easily eat up to 10% less power.

[QUOTE=ATH;258506]Maybe you should increase fan speed manually? When I ran mfaktc I set fan speed to 80% and made a rule that if temperature exceeded 60C fan speed increased to 100%.

Isn't it best to use the fan to its max to keep GPU temperature low?[/QUOTE]How to do that under Fedora? It's easy to read the fan settings but I haven't yet discovered how to set them.


Paul

[QUOTE=diep;258503]Are you also blowing enough air into the box to cool the mainboards interconnects?

This as mfackt will fry the interconnects of course as it ships loads of data to the GPU.

There is no temperature sensors at the interconnects, but touching careful with your hand you will be able to confirm me that you can cook an egg on 'em.[/QUOTE]Good suggestion. I'll check it out when I next have a reason to open the box. At the moment, everything appears to be working smoothly again.


Paul

[QUOTE=xilman;258512]Good suggestion. I'll check it out when I next have a reason to open the box. At the moment, everything appears to be working smoothly again.


Paul[/QUOTE]

Until then you can run it as mfockt :)

mfaktc-0.16p1.performance - OS: Linux 64 bit - Drivers: 270.40 GPU: GTX 470 @ 607/1215/1674 MHz core/shaders/memory - CPU: Q9550 @ 3.6 GHz - Trial factoring from 68-69 bit / Exponents in the 56.5M area.

[CODE]inst./avg. rate / total:

1 156M/s 156M/s
2 146M/s 292M/s
3 100M/s 300M/s[/CODE]

[QUOTE=xilman;258420]Ah, it worked when all four results were submitted.[/QUOTE]After less than a day's work mfaktc has propelled me well into the top 1000 in the TF list. That's rather impressive, IMO, though I can't really take any credit for it.

One thing that puzzles me, though, is the report page:[code]
[FONT=Courier New][SIZE=2] Rank Member Name GHz-Days Attempts Successes |90day 30day 7 day 1 day | 90 days 30 days 7 days[/SIZE][/FONT][FONT=Courier New][SIZE=2]
685 Paul Leyland 101.127 95 0 | [IMG]http://www.mersenne.org/images/up_arrow.gif[/IMG]*** [IMG]http://www.mersenne.org/images/up_arrow.gif[/IMG]*** [IMG]http://www.mersenne.org/images/up_arrow.gif[/IMG]*** [IMG]http://www.mersenne.org/images/up_arrow.gif[/IMG]*** | [/SIZE][/FONT]
[/code]What does the "Successes" column mean? I thought it may be the number of factors found. A few factors have been discovered and reported through the manual reporting page but I have no "Successes", whatever they are.

George: if you're reading this could you explain please? The results.txt files will live on in the incremental backups for a few more days and can be recovered if needed.

Paul