#asahi-gpu · 2021-07-06 02:41
Yuzu486, not pentium#asahi-gpu · 2021-05-10 19:59
Yuzuthat's a significant failure mode#asahi-gpu · 2021-05-02 23:03
Yuzuthat's a milestone. congratulations! :)#asahi-gpu · 2021-05-02 16:54
Yuzuin no time at all, you'll get your apple mark, and then you can start working the apple farm until the day you die... just like everypony in our family has since the dawn of time!#asahi-gpu · 2021-05-02 16:51
Yuzubloom: because i don't expect to see FiW in most places!#asahi-gpu · 2021-05-02 16:50
Yuzuwas not expecting to see FiW in this channel, but it's a welcome surprise...#asahi-gpu · 2021-05-02 16:42
Yuzushould have gotten that earlier. :)#asahi-gpu · 2021-05-02 16:41
Yuzubloom: because of Apple Bloom?#asahi-gpu · 2021-04-30 23:08
Yuzuyou really like Winx?#asahi-gpu · 2021-04-19 13:27
Yuzuand if anything, the stuff in the linked presentation sounds in-order (the optimizations in slide 57, for instance, would be of fairly limited use in a true OoO machine, since independent ops would happily execute while waiting for blocked ops)#asahi-gpu · 2021-04-19 13:19
Yuzuit seems like the structures needed to support OoO would be expensive when architectural state is as large as it tends to be in GPUs#asahi-gpu · 2021-04-19 13:18
Yuzui was wondering about the same thing#asahi-gpu · 2021-04-02 22:08
Yuzuiirc no licensable core is FPless or even FP-optional, and i don't think any of the architectural licensees have gone FPU-free either#asahi-gpu · 2021-04-02 22:08
Yuzui'm surprised the spec doesn't mandate FP to be present#asahi-gpu · 2021-04-02 22:08
Yuzuopticron: bloom hadn't actually said ARM64, so i figured M or R were possible (albeit unlikely - apple has a long and proud history of using lots of auxiliary ARMv8-A cores)#asahi-gpu · 2021-04-02 20:38
Yuzuif it's an M (or an R, i guess), that's interesting on its own#asahi-gpu · 2021-04-02 20:37
Yuzuisn't FP non-optional in ARMv8-A anyway? maybe i'm remembering wrong#asahi-gpu · 2021-03-27 00:18
Yuzuvariable length ops = higher density (typically, if done right)#asahi-gpu · 2021-01-07 20:01
Yuzubloom: Fair - was just wondering if you had seen any patterns in generated code that would be obviously inefficient unless there's delays#asahi-gpu · 2021-01-07 19:03
Yuzuso AGX is a fully dynamically scheduled machine? have you seen any sign of mandatory delays before dependent ops or anything like that?