Comments (2)
airmler
commented on September 3, 2024
1
Thanks for fast clarification.
I conclude that the correct approach is to neglect the slowest number.
from cosma.
rasolca
commented on September 3, 2024
Am I right that there is a 300 ms overhead finding the optimal "parallelization strategy"?
No, the overhead is very likely due to library initializations during the first run in the miniapp.
Multithreaded MKL is usually the the library that introduces more overhead as it has to initialize the OpenMP environment and allocate some memory during the first library calls. MPI on certain systems introduces as well some overhead during the first communications.
from cosma.
Related Issues (20)
- COSMA cublas crash after job finished HOT 5
- Fixing CI/CD issues HOT 1
- GPU-Aware MPI Version HOT 2
- Error crash at the end of the job execution HOT 7
- Unable to use an internal RCCL build HOT 4
- COSMA crash on Perlmutter when dealing with complex values HOT 9
- Crashes with the latest COSMA release HOT 25
- (Still) Excessive memory usage HOT 3
- cmake project version v2.6.1 does not match git tag v2.6.2
- Add ability to disable NCCL at runtime HOT 1
- build failure with nccl HOT 1
- enable overlap comm with computation for cosma_miniapp HOT 2
- Configure fails to find costa submodule and build fails to build with costa-2.2 HOT 1
- Link errors HOT 6
- cmake project version v2.6.4 does not match git tag v2.6.5
- COSMA build fails on CRAY HOT 1
- How to run pdgemm on multiple GPUs? HOT 26
- undefined reference to `void costa::transform<float>(std::__1::vector<std::__1:: ... HOT 3
- Switching to a proper memory-pool implementation HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cosma.