Hi there!

I have a Win 7 x64 box with a Radeon HD6450 using the WHQL 15.7.1 driver. For some reason WCG it says that it's unusable, even though it's Open CL 1.2 compatible. I realize it's old, but it seems to me that it should work fine. Did I just forget to enable something?

4/12/2021 9:01:01 AM |  | CAL: ATI GPU 0: AMD Radeon HD 6350/6450/7450/7470/R5 230 series (Caicos) (CAL version 1.4.1848, 1024MB, 991MB available, 400 GFLOPS peak)
4/12/2021 9:01:01 AM |  | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6350/6450/7450/7470/R5 230 series (Caicos) (driver version 1800.8 (VM), device version OpenCL 1.2 AMD-APP (1800.8), 1024MB, 991MB available, 400 GFLOPS peak)

4/12/2021 9:19:10 AM | World Community Grid | Not requesting tasks: don't need (CPU: job cache full; AMD/ATI GPU: GPUs not usable)

Need more units. My machine is ready to process. It goes is spurts because no more 'units' are available.

But which are the apples, which are the oranges?

Actually, the 'request work' query is much more server intensive (when there's work available) than reporting completed work. It involves checking your preferences for what you want to run, finding the best app version consistent with those, looking through the tasks in the feeder cache, checking that it isn't prohibited because you can't be your own wingmate, and so on. Reports just update the existing records.
But you also need to tweak it and run it to suit the project's needs - it's their science, after all, and their server. Yes, I'm running the scripts, for now - partly because I've worked out how to run the tasks which cause the iGPUs to fail, and this project has configured those 'normal' server settings to require a matching re-issue.

So, were now in WCG, and back with the old SETI discussions with the same credit hounds and WU hoarders from SETI, with several computers containing 4-5 or more powerful GPU's, in each computer? As per usual SETI behaviour, complaining that they can't get as many WU's as they want.....

Incredible......

Seems like I'm getting decent amount of gpu tasks by running the script. Doesn't feel like it is 1700 tasks generated every 30 mins for a total of about 81,600 tasks per day or worst case 81,600 x 2 if every task require validation. I don't see high number of pending validation on my hosts, at least for now.

OP daily project stats shows more than 81,600 tasks on average processed per day since Apr 6th. On the CPU side, historically prior to Apr 6, we have been averaging around 1.014M tasks per day with standard deviation of 0.021M. Since Apr 6, we are averaging 1.156M so a delta of 142k tasks been processed.

I'm running only OPNG right now. Perhaps most people who just want to run gpu task also run cpu task which could explain the above observation. Maybe in the future good to split the cpu versus gpu after the gpu tasks run stable for some time.

In anyway, let's hope the admin can add more gpu tasks in the very near future.