Half precision using GPU

10 vues (au cours des 30 derniers jours)
Fernando le 10 Avr 2023
Commenté : Fernando le 11 Avr 2023
Hello, I was trying to see if I can run some code using half precision rather than single.
before converting my code, I was trying a very simple example.
This gives me the error: No constructor 'half' with matching signature found.
Using the the half with the CPU works fawlessly.
Any idea if this is supported by all? Looking here, https://www.mathworks.com/help/gpucoder/ug/what-is-half-precision.html, it seems some GPU should support it?
I am using a 16 GB RTX3080 Mobile. R2022b.
  2 commentaires
Fernando le 11 Avr 2023
Unforunately, this won't work either, it gives: GPU arrays support only fundamental numeric or logical data types.

Connectez-vous pour commenter.

Réponse acceptée

Joss Knight
Joss Knight le 11 Avr 2023
As pointed out, gpuArray does not support half. The main reason is that half is an emulated type only meaningful for deployment to special hardware, it is not native to most processors. Feel free to investigate use of half for code generation.
Do you just want to store data in half to save space on the GPU? You can use the following code to get something like the behaviour you're after:
function u = toHalf(x)
realmaxHalf = single(65504);
x = min(max(x,-realmaxHalf),realmaxHalf);
[f,e] = frexp(abs(x));
sgn = uint16(x>=0);
sgnbit = bitshift(sgn,15);
expbits = bitshift(uint16(e+15),10);
fbits = uint16(f.*2.^10 - 1);
u = bitor(bitor(sgnbit, expbits), fbits);
function x = fromHalf(u)
if u == 0
x = single(0);
u = uint16(u);
sgn = single(bitshift(u,-15));
fbits = bitand(u,uint16(1023));
f = single(fbits+1)./(2.^10);
expbits = bitand(u,uint16(31744));
e = single(bitshift(expbits,-10))-15;
x = (sgn.*2-1).*f.*2.^e;
Note, this is a very crude implementation of fp16 that takes no account of nans, infs, correct overflow behaviour or denormals. The half version is just a uint16 with the data in it, you can't actually use it to compute anything in fp16.
  4 commentaires
Fernando le 11 Avr 2023
I see. The issue is that I gain more from having larger matrices as oppossed to have smaller ones with higher precision or digits in them.
I guess I could try to work with your solution while I figure out another way or buy a better gpu.

Connectez-vous pour commenter.

Plus de réponses (1)

Matt J
Matt J le 11 Avr 2023
Modifié(e) : Matt J le 11 Avr 2023
GPU Code Generation does support it, but not the Parallel Computing Toolbox, which is where gpuArray is defined.
  3 commentaires
Fernando le 11 Avr 2023
Ok, I will try to look into this.

Connectez-vous pour commenter.


En savoir plus sur Get Started with GPU Coder dans Help Center et File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by