fused_concat: Support fp8-fp32 - Githubissues

pulp-platform / snitch_cluster

An energy-efficient RISC-V floating-point compute cluster.

https://pulp-platform.github.io/snitch_cluster/

Apache License 2.0

51 stars 51 forks source link

fused_concat: Support fp8-fp32 #122

Open fischeti opened 7 months ago

fischeti commented 7 months ago

Almost all formats are now working:

FP64

[x] NAIVE
[x] OPT

FP32

[x] NAIVE
[x] BASELINE
[x] OPT

FP16

[x] BASELINE
[x] OPT_EX
[x] OPT

FP8

[x] BASELINE: kind of works, error is not far off. Probably saturation/cancellation
[x] OPT: Same