--set
|
b521b2eadb
|
Fix build breaks, breakeventHtoDMemcpy.cu
|
5 months ago |
--set
|
35fbd1c8aa
|
Merge branch 'master' of github.com:ArchaeaSoftware/cudahandbook
|
7 months ago |
--set
|
41c9e13c7a
|
Tune up nullKernelSync.cu and cudaGetLastErrorIsAsynchronous.cu
|
7 months ago |
Nicholas Wilt
|
8d09575359
|
Make nullKernelAsync.cu minimal again
|
8 months ago |
Nicholas Wilt
|
e467e68672
|
Make nullKernelAsync.cu minimal again
|
8 months ago |
Nicholas Wilt
|
4f8dad585d
|
Add cudaGetLastErrorIsAsynchronous.cu
|
8 months ago |
Nicholas Wilt
|
ea91af905f
|
Furher hipification
|
2 years ago |
Nicholas Wilt
|
70c310f639
|
Async NULL kernel launch
|
2 years ago |
Nicholas Wilt
|
e7304fc146
|
Make nullKernelAsync.cu portable between CUDA and HIP
|
2 years ago |
Nicholas Wilt
|
1cda62ac16
|
HIPify histogram sample
|
3 years ago |
Nicholas Wilt
|
96a8b8240b
|
Get Reduction sample to build under HIP
|
3 years ago |
Nicholas Wilt
|
a0148b0f79
|
Get multiple GPUs working
|
3 years ago |
Nicholas Wilt
|
a5fa9d069d
|
Multi-GPU support
|
3 years ago |
Nicholas Wilt
|
a7cc291a2a
|
For multiple GPUs, report max time
|
3 years ago |
Nicholas Wilt
|
46a7c2283c
|
Checkpoint multi-GPU implementation, vectorize events
|
3 years ago |
Nicholas Wilt
|
99a7c348d4
|
Create vector of GPU device arrays
|
3 years ago |
Nicholas Wilt
|
4a9666b6ed
|
Separate CUDA and HIP specific wrappers in chError.h
|
3 years ago |
Nicholas Wilt
|
bdae5e7544
|
Continue porting stream samples to HIP
|
3 years ago |
Nicholas Wilt
|
eeed4dfa1d
|
Port stream2Async.cu to HIP
|
3 years ago |
Nicholas Wilt
|
ca65c1eb48
|
Port globalCopy.cu to HIP
|
3 years ago |
Nicholas Wilt
|
976d78b94d
|
More HIP compat
|
3 years ago |
Nicholas Wilt
|
eafe3b9439
|
Update a few apps to use HIP
|
3 years ago |
Nicholas Wilt
|
5027f096cb
|
FMA implementation
|
3 years ago |
Nicholas Wilt
|
f816c78b6d
|
AVX version
|
3 years ago |
Nicholas Wilt
|
074873cb9d
|
Got SSE working
|
3 years ago |
Nicholas Wilt
|
8cc116ecf7
|
Add SOA implementation
|
3 years ago |
Nicholas Wilt
|
871c6a954d
|
Use absolute error instead of relative error
|
3 years ago |
Nicholas Wilt
|
ac4c587721
|
Checkpoint cross checking code
|
3 years ago |
Nicholas Wilt
|
d6dfe50cbb
|
Fixed some bugs; checkpoint GPU version
|
3 years ago |
Nicholas Wilt
|
1ae650a4f0
|
Checkpoint reworked, more modern C++ nbody
|
3 years ago |
Nicholas Wilt
|
a0cd917a52
|
Initialize now initializes posMass/velInvMass
|
3 years ago |
Nicholas Wilt
|
e18a57fc26
|
C++-ify: class hierarchy for algos, use std::vector for AOS
|
3 years ago |
Nicholas Wilt
|
647f8b2161
|
Checkpoint working version
|
3 years ago |
Nicholas Wilt
|
f9a5490496
|
Fix bit rot, update to use modern idioms
|
4 years ago |
Nicholas Wilt
|
d71cd98c31
|
Checkpoint divergence performance test
|
4 years ago |
Nicholas Wilt
|
8757372fc4
|
Fix warnings re format string disagreement w size_t
|
9 years ago |
Nicholas Wilt
|
c759e29bfa
|
Port microbenchmark directory to new error handling.
|
9 years ago |
Nicholas Wilt
|
096e5506f0
|
Port reduction sample code to new error handling
|
9 years ago |
Nicholas Wilt
|
4bfea533a8
|
Port SMs sample code to new error handling
|
9 years ago |
Nicholas Wilt
|
593c9a8ded
|
Port texturing sample code to new error handling
|
9 years ago |
Nicholas Wilt
|
2c503e590c
|
Port Scan to new error handling
|
9 years ago |
Nicholas Wilt
|
1508f66c6a
|
Port streaming samples to new error handling
|
9 years ago |
Nicholas Wilt
|
fd03ad3694
|
Ported nbody to new error handling
|
9 years ago |
Nicholas Wilt
|
83bca6802c
|
Rest of memory directory now uses new error handling
|
9 years ago |
Nicholas Wilt
|
2637973750
|
Port histogram sample to new error macro
|
9 years ago |
Nicholas Wilt
|
4abe052e99
|
New error handling
|
9 years ago |
Nicholas Wilt
|
a679378499
|
Initial checkin, new error handling macros
|
9 years ago |
Nicholas Wilt
|
b7708e90b2
|
DOS2UNIX chError.h
|
9 years ago |
Nicholas Wilt
|
7f85b3ccbb
|
Add chNUMA.h
|
9 years ago |
EC2 Default User
|
4139b98c2d
|
Add coins.pgm to histogram directory
|
10 years ago |