13434 unusually high PPD

[gunnarre]

I'm folding on an GTX 1660 Super which usually produces about 1.2M to 1.4M PPD, but on 13434 (Run 353, Clone 0, Gen 3) I was getting 2.6M PPD. Is this just high atom count, or is the score for this project too high? Anyone else see this?

[PantherX]

It might be a Christmas bonus I have asked around so let's see what happens

[toTOW]

I heard that each run in these projects (134xx) doesn't have the same number of atoms ... so some might be faster than others ... since the benchmarks are trying to average it, some WUs might be faster than usual, and some will be slower ...

[Joe_H]

Your 1660Super may also be hitting the spot for best return, more shaders would be partially unused.

[rewron]

I just deleted a 13434 WU with high points.
It would have completed about 24 hours after the expiration date.

Edit: corrected WU#--Must have clumsy digits today. Thanks Gunnarre for the headsup.

[gunnarre]

I just deleted a 13143 WU with high points.
It would have completed about 24 hours after the expiration date.

This is a different project. Project 13434 is a COVID Moonshot sprint. Project 13143 studies the NMDA receptor on nerve cells.

[Knish]

what a christmas bonus
V100 usually gives around 230k per WU... 13434 just returned 606k \o/

[Kjetil]

11029085 ppd 3090 p 13436.

[HaloJones]

6195096 ppd 1080ti P13434 (349, 4, 13)

[BobWilliams757]

No big bonus WU's for me lately. Then again, I was wondering why it seemed all my WU's took a performance hit. Then I noticed that at some point (after some power losses at home) my memory was reset to 1066 vs 1600. I'm just happy to be getting back to average normal returns again.

Being I've been folding on this little rig 24/7, I just got too busy to notice and assumed that there was a change in PPD averages due to larger work units or something. DOH!