Shanie Winitz
|
96debb96a8
|
cpu_ntt pre-parallel
|
2024-07-28 16:43:48 +03:00 |
|
Yuval Shekel
|
2a3dcd776a
|
fix format
|
2024-07-28 15:36:30 +03:00 |
|
Yuval Shekel
|
8d03fb6bc8
|
trigger examples build in ci
|
2024-07-28 15:33:00 +03:00 |
|
Yuval Shekel
|
445eec88af
|
refactor C++ example risc0
|
2024-07-28 15:30:48 +03:00 |
|
Yuval Shekel
|
a5a8bf8e3d
|
refactor C++ example poly-api
|
2024-07-28 15:20:23 +03:00 |
|
Yuval Shekel
|
68802bc14a
|
refactor C++ example polynomial-multiplication for V3
|
2024-07-28 14:49:02 +03:00 |
|
Yuval Shekel
|
63dcd4ef5d
|
remove multiply c++ example and fixed some readme
|
2024-07-28 13:01:01 +03:00 |
|
Yuval Shekel
|
c302aebeca
|
refactor pedersen-commintment example to V3
|
2024-07-28 12:55:12 +03:00 |
|
Yuval Shekel
|
b07d02096f
|
update ntt example for V3
|
2024-07-28 12:38:50 +03:00 |
|
Yuval Shekel
|
454fe5b922
|
update ntt-best-practice and msm examples to V3
|
2024-07-25 18:32:31 +03:00 |
|
Yuval Shekel
|
99e58532ad
|
update readme of C++ msm example
|
2024-07-25 15:41:58 +03:00 |
|
Yuval Shekel
|
9618e6db74
|
split backend registration part to backend headers to avoid dep on concrete fields/curves
|
2024-07-24 20:34:05 +03:00 |
|
Yuval Shekel
|
fc2ac41710
|
refactor msm example
|
2024-07-24 20:27:26 +03:00 |
|
Yuval Shekel
|
49beb324a8
|
add script for generating C apis
|
2024-07-24 20:12:42 +03:00 |
|
Yuval Shekel
|
b52191f9dd
|
field generation update from V2
|
2024-07-24 17:08:37 +03:00 |
|
Yuval Shekel
|
05d918fcda
|
let PR choose the cuda-backend branch to use
|
2024-07-21 17:37:44 +03:00 |
|
Yuval Shekel
|
bdf3cd5672
|
add info logs to NTT test to find the issue
|
2024-07-21 11:36:44 +03:00 |
|
Yuval Shekel
|
58780f0115
|
fix compilation issues for mac
|
2024-07-18 18:40:01 +03:00 |
|
Yuval Shekel
|
4f47c9d4f3
|
remove redundant link to stdc++ in build.rs scripts
|
2024-07-18 18:11:11 +03:00 |
|
Yuval Shekel
|
1baabd93d6
|
add feature to rust crates to choose local/remote cuda backend
|
2024-07-18 18:03:42 +03:00 |
|
Yuval Shekel
|
a8e794fdfd
|
use ssh-key in ci workflows to pull cuda backend
|
2024-07-18 17:27:49 +03:00 |
|
Yuval Shekel
|
934f934976
|
pull cuda backend when building in ci
|
2024-07-18 14:25:51 +03:00 |
|
Yuval Shekel
|
cd3353362a
|
move cuda backend to private repo and clone it in ci
|
2024-07-18 12:41:27 +03:00 |
|
Yuval Shekel
|
afe48317a6
|
skip fast twiddles for domains smaller than 16
|
2024-07-18 12:02:22 +03:00 |
|
Yuval Shekel
|
7de94fff84
|
fix bug in cuda MSM when scalars are in montgomery form
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
d3b27a7b68
|
randomize device for C++ polynomial tests
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
0409c6f5d7
|
replace msm config bases_size with boolean flag indicating bases are shared for batch elements
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
9c35aaa5c3
|
add doxygen comments to frontend APIs
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
db19c3846d
|
split default poly backend from cuda and support CPU too
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
06243e5493
|
add polynomial evaluation api and cuda backend. Use it in poly backend
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
8ade7ef393
|
use vector_add in polynomial backend to add monomial inplace
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
749d3f9b6e
|
add vec_op api for find highest non zero idx of vector, and CUDA backend
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
0885195619
|
add vec_ops apis for scalar+-vector and use in polynomial backend
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
6f8c480fd1
|
use vector_div in polynomial backend
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
f13f24cff2
|
add vec_div api
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
028f59cef0
|
use vec_mul in polynomial api insted of another kernel
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
61af965b41
|
add api for mul scalar and use in poly backend
|
2024-07-17 19:11:47 +03:00 |
|
Yuval Shekel
|
091a19354e
|
add slice api to vec ops
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
0c8a849d7c
|
move default polynomial context from cuda backend to icicle FE
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
87a25afee7
|
use icicle runtime apis in polynomial implementation instead of cuda apis
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
4d0b1a03a3
|
removed device_id from polynomial view
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
de88ad144e
|
minor: CPU_REF reuse device_api class like CPU
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
9b46b8489f
|
api for memset and tests for copy() and memset()
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
caf3b4fab2
|
add generic copy and copy_async with auto device inference
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
e49540de20
|
polynomial refactor to be multi-device
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
74f7a3cbcd
|
move tracking logic to runtime.cpp rather than device API
|
2024-07-17 19:11:46 +03:00 |
|
Shanie Winitz
|
36c879912a
|
NTT simple cpu backend (#533)
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
ca7fb84cf0
|
switch device when releasing memory of inactive device
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
76cf23d02a
|
ecntt bug fix from V2
|
2024-07-17 19:11:46 +03:00 |
|
Yuval Shekel
|
5a7c0ccd22
|
print domain size when ntt fails on that. same domain size for ntt and ecntt to avoid issues
|
2024-07-17 19:11:46 +03:00 |
|