Quote from: D.A. Bentley on June 08, 2020, 02:44:14 PM...
For example, if you could on my 24-core CPU target Thread 0,2,4,6,8,10,12,14,16,18,20,22 as the group of threads to use but leave the odd threads unused. I would expect if my theory is correct that the rendering results would be similar to if SMT was OFF which just a small performance decrees (not 50% less performance).
...
From what i remember in the past i doubt that using SMT OFF is the way to go. Even in the past it wasn't much about getting 50% faster render speeds. It depends on the software you use but mostly you will see faster rendering like in the 20%-30% range. It can be more and less but it will be mostly always faster with SMT on. I would leave it ON and wouldn't think much about it except if needed and since so many years i have always used it ON. This is about rendering. Different programs that don't use that much cores are of course another thing where you might (maybe) consider it.
The last buckets problem is annoying and the best way would be if Matt could make rendering adaptive.
In animation i get around it by changing the bucket size ( i tried one time even to use a fully image wide size with small high size just out of curiosity).
Or sometimes depending on the scene just using 2 or more instances of Terragen. This is not ideal but was faster then not using it.
