Citiraj:
Autor Manuel Calavera
|
Evo ti quote sa MS-ovog announcement-a za DXR:
Citiraj:
|
You may have noticed that DXR does not introduce a new GPU engine to go alongside DX12’s existing Graphics and Compute engines. This is intentional – DXR workloads can be run on either of DX12’s existing engines. The primary reason for this is that, fundamentally, DXR is a compute-like workload. It does not require complex state such as output merger blend modes or input assembler vertex layouts. A secondary reason, however, is that representing DXR as a compute-like workload is aligned to what we see as the future of graphics, namely that hardware will be increasingly general-purpose, and eventually most fixed-function units will be replaced by HLSL code.
|
A mislim da i AMD na to cilja, general purpose compute unit pa mu daj da radi sta hoces i koliko hoces.
Takodjer, odakle na onoj tvojoj slici 13 TFLOPS compute power za Pascal? Titan X je imao 11, 1080 oko 9 ako se dobro sjecam.
A pogledaj na Turingu koliko RT jezgre zauzimaju prostora. Da su na taj komad stavili compute jezgre, imao bi 23-24 TFLOPS-a compute power-a. Mislis da to ne bi vrtilo DXR demo kako spada?
Sve u svemu, veselim se vidit kako su MS i AMD to izveli, ja se kladim na cisti compute power.