Octane renderer - what type of precision / FLOPS ?
Posted: Wed May 17, 2017 1:11 pm
Hi all,
What with all the new GPUs coming soon - nVidia Volta and AMD Vega, it got me wondering as they throw around specs of FLOPS procession power.
For example, the new AMD Vega cards -
" this graphics card packs in 64 next-gen compute units providing 4096 shader processors, 16GB HBM2, supports 8K displays, and offers 12.5TFLOPS FP32 performance or 25TFLOPS of FP16 performance."
What precision does a GPU renderer use?
FP16? FP32?
Which set of specifications on a card should be what starts to get us excited by the new technology?
Thanks!
Dom
What with all the new GPUs coming soon - nVidia Volta and AMD Vega, it got me wondering as they throw around specs of FLOPS procession power.
For example, the new AMD Vega cards -
" this graphics card packs in 64 next-gen compute units providing 4096 shader processors, 16GB HBM2, supports 8K displays, and offers 12.5TFLOPS FP32 performance or 25TFLOPS of FP16 performance."
What precision does a GPU renderer use?
FP16? FP32?
Which set of specifications on a card should be what starts to get us excited by the new technology?
Thanks!
Dom