14:17fdobridge: <![NVK Whacker] Echo (she) 🇱🇹> NFS Underground still works after the rebase
16:52republic83: And about slapping you all know if you came face to face without fraud you had no chance to slap me, not your jack, indrek nor Alex, with broken hand I slap those to death too, but the way you handle life is that you can I can't I'd be fabricated to jail cause you are born retards ok, but I can not interfere when one slut wants to fuck retards either, now she is without teeth after abusing me over this, but only cause the retard crashed her bike, I
16:52republic83: did absolutely nothing, but she did it to her own or some of the screwed crankgangsters did beam again.
16:56republic83: But the work there's not a lot left to do , because instead of nothing and clown playing I did a few and that proved to be almost enough, I have only little work left.
16:56tertl8: karolherbst: do you run asahi or macOS?
16:57tertl8: i saw your rusticl talk recently, have you tried any gpu compute on the mac
16:58karolherbst: I did, but compute support on agx is still very experimental
19:30fdobridge: <![NVK Whacker] Echo (she) 🇱🇹> https://youtu.be/ujuCYZLeSsw
19:44tertl8: fdobridge: wow thats impressive
19:44tertl8: I finally figured out that my old 1060 doesnt support nvk
19:44tertl8: i think its only rtx cards that work
19:47tertl8: although they use confusing language like Pascal or something
19:50karolherbst: Pascal is the generation without raytracing
19:50karolherbst: but it should still work
19:50karolherbst: might crash randomly though
19:58fdobridge: <Mohamexiety> 1060 should work. My 1030 works except for some things which crash.. I don't really use it though, just to run CTS
20:10tertl8: i would like to try it but I dont have a test machine right now unfortunately
20:12tertl8: i use the proprietary driver right now because ive become curious to see how far I can get trying to do my python stuff with the cuda wrappers
20:18tertl8: I have to stick with python due to not being experienced with low level languages
20:20tertl8: Single-source C++ shaders for graphics and compute
22:47tertl8: pycuda vs c-cuda