- last edited on by
ROGBot
Hey all,
I'm seeing Aida64 report some strange temp readings for my three HD 7950s.
I've attached two screenshots that detail the report:
A note:
Opt 1: Air temperature intake
Opt 2: Case internal air temperature over CPU
Opt 3: Water temperature in loop (CPU and motherboard)
As you can see, my middle GPU is a full 15 - 17 degrees C cooler than the top and bottom cards.
Clock speeds are all identical at time of reading, and as you can see, voltage and power usage are comparable across all three.
I ALWAYS re-paste new GPUs with Noctua's TIM, so these are not factory pasted.
Is it possible I did a great job on one and awful job on the other two? I haven't repasted yet, but will soon and update.
Any ideas?
Z
Edit: I forgot to mention that this screenshot was taken after twenty minutes of RealBench v2 stress testings.
I'm seeing Aida64 report some strange temp readings for my three HD 7950s.
I've attached two screenshots that detail the report:
A note:
Opt 1: Air temperature intake
Opt 2: Case internal air temperature over CPU
Opt 3: Water temperature in loop (CPU and motherboard)
As you can see, my middle GPU is a full 15 - 17 degrees C cooler than the top and bottom cards.
Clock speeds are all identical at time of reading, and as you can see, voltage and power usage are comparable across all three.
I ALWAYS re-paste new GPUs with Noctua's TIM, so these are not factory pasted.
Is it possible I did a great job on one and awful job on the other two? I haven't repasted yet, but will soon and update.
Any ideas?
Z
Edit: I forgot to mention that this screenshot was taken after twenty minutes of RealBench v2 stress testings.
The Realbench stress test has always caused my GPUs up into the higher temp ranges, so I didn't think about it further.
I'll try it out with Furmark and update.
:::UPDATE:::
5 minutes using Geek3D Furmark - same thing is happening. Temps are just a little higher.
And after a half-hour on Skyrim running at max graphics settings.
:::ADDITIONAL UPDATE:::
I have just finished testing the system by running the Heaven Benchmark under several configurations.
Say the original card setup was 1:2:3, top:middle:bottom, and that's what I used for the aforementioned reports.
I moved cards to the following configurations, ran the benchmark, and had the same results as before:
1:3:2, 2:1:3, 3:1:2, 3:2:1 - all temperatures high:low:high, relatively speaking.
So, it must just be the way the work is distributed to cards in TriFire, and I suppose this thread can be retired now.
On a side note, my best score in Heaven was achieved with the 2:1:3 config, with all settings set at required values for the affiliated forum:
Dr. Z
I'll try it out with Furmark and update.
:::UPDATE:::
5 minutes using Geek3D Furmark - same thing is happening. Temps are just a little higher.
And after a half-hour on Skyrim running at max graphics settings.
:::ADDITIONAL UPDATE:::
I have just finished testing the system by running the Heaven Benchmark under several configurations.
Say the original card setup was 1:2:3, top:middle:bottom, and that's what I used for the aforementioned reports.
I moved cards to the following configurations, ran the benchmark, and had the same results as before:
1:3:2, 2:1:3, 3:1:2, 3:2:1 - all temperatures high:low:high, relatively speaking.
So, it must just be the way the work is distributed to cards in TriFire, and I suppose this thread can be retired now.
On a side note, my best score in Heaven was achieved with the 2:1:3 config, with all settings set at required values for the affiliated forum:
Dr. Z
I have 4 XFX HD 7950's. They always ran hot. 1 or two cards running wasn't that bad. If I added a third temps were a bit higher. Adding the fourth card then much higher. The cards are so close they cant breath. on the BENCH I lay two fans on top of the cards one over the power cable area and one laying flat. Drops temps by 10 -15%
