3 ms·
> If everything constantly depends on the prior action you will never get any uplift. Not always. For differential equations with large enough matrices, the in
by Archit3ch 1y ago
> If everything constantly depends on the prior action you will never get any uplift.
Not always. For differential equations with large enough matrices, the independent work each core can do outperforms the communication overhead of core-to-core latency.
- jayd16 1y agoIf its independent work then it's work that doesn't rely on the prior action... At least not in the way the parent means.
- Archit3ch 1y agoState can depend on the previous time point, or even the same time point. I see this misconception often in audio programming "you cannot parallelise work because it depends on the previous sample". As long as you can find parallelism somewhere and it's less than the overhead, you can benefit. Obviously if there's zero parallelism in the problem, no amount of cores will help.