Does anyone else see GPU behavior like this?

https://i.imgur.com/PiM63xn.jpg

Clock speed, and voltage cycle up and down constantly if i'm running 1 WU or 4 WUs at once. In the image it says .25 gpu usage but it's only running one in this case, it was previously running 4 but went to a single to rule out if concurrency was the issue.

7970
Driver Packaging Version
17.30.1041-170720a-316467C-CrimsonReLive
Radeon Software Version
17.7.2
Radeon Settings Version
2017.0720.1902.32426

I haven't seen this, but what popped into my mind was you may be suspending computation because of settings.

Go to Options, Computing Preferences, under the "When to Suspend" area, may sure EVERYTHING is UNCHECKED.

If this isn't it, can't help.

The remaining runtime in the above photo is misleading, probably estimated on the 4x runs.

My single WU runtimes are about 1:05
Does anyone with a 7970 or R9 280x have a runtime I can compare with?

I haven't seen this, but what popped into my mind was you may be suspending computation because of settings.

Go to Options, Computing Preferences, under the "When to Suspend" area, may sure EVERYTHING is UNCHECKED.

If this isn't it, can't help.

It's set to use GPU always, those up-down cycles are seconds apart, it's a real-time graph.

I wonder if it downclocks while it's waiting for a slower double-precision function? The Tahiti's have amazing double-precision performance though, relative to other desktop cards.

For comparison this is the same machine that i switched to primegrid GFN.

I am not seeing it with my R9 280X. My runs have been averaging around 47 seconds for the set of units I have received to investigate this. So I'm not sure what is going on with your GPU, maybe a background program? My CPU usage on GPU-Z is smooth like your primegrid example.

One other possibility I can think of and its not likely. Make sure your power settings are set to high performance. Since your Primegrid seems to be working ok it probably isn't this but its worth a try.

What type of CPU / system do you have? I have seen patterns like this on CPUs that have very little L2 cache per thread. A lot of swapping may be going on. For example, I am showing results from an MSI Platinium with an "old" X5470. There are only 4 cores but each pair of core has access to 6mb of level 2 cache. Considerably more than, for example, a hex core xeon that is 6x256kb but with each thread only having 128kb.

The HD7950 on the left of the photo is running 4 WUs and the one one the right 5. Not sure why is there an imbalance.