Comments (3)
attached code from main branch and main branch without promote hoisting of vectorize predicates.
high_perf.txt
low_perf.txt
from fuser.
Generated with:
NVFUSER_PROF=print NVFUSER_DUMP=cuda_to_file,ptxas_verbose,scheduler_params,launch_param pytest /opt/pytorch/nvfuser/benchmarks/python/test_layernorm_bwd.py -k "float16 and 16384_16384" -v -s --disable-benchmarking 2>&1 |tee 1.log
from fuser.
The possible reason of regression may because of compiler optimization from cuda to ptx code.
The slow version has 2500 more lines of PTX code.
In the fast version , #2742 ,b47
is used in both if
and else
branches. Compiler can optimize out the else
branch, a very short PTX code generated for else branch.
In the slow version (current main), b47
is only used in the else
branch. The compiler generates PTX code for the else branch. Thatβs why we see 2500 more lines of PTX code.
slow.ptx.txt
fast_revert.ptx.txt
from fuser.
Related Issues (20)
- Hopper-only crash: Unsupported loop structure. Two loops are mapped together. HOT 26
- Sharding propagation
- ExpressionEvaluator errors out when running `logical_and` HOT 1
- Add runtime check for circular buffer correctness
- Codegen error: `have dynamic allocations but are placed in local memory.` coming from qkv_split_rope backwards HOT 9
- benchmarking thunder.jit
- MarkAliasPrepare does not preserve shardings
- Generalize pattern matching of matmul patterns
- SDPAOp and ComputeAtMap HOT 2
- makeInitialSchedulerEntry requires more than the current segment's inputs. HOT 6
- nvFuser confuses sharded dimension size and unsharded. HOT 3
- crash in transformOutputFromAllocationToLogical: frontier.size() != logical.size() HOT 2
- Indexing failure due to self-mapped broadcast domains
- Indexing error in matmul with certain parameters HOT 1
- Allocs must be based on constant integers HOT 7
- Promote reuse over specialization for small reductions
- INTERNAL ASSERT FAILED: Allocations must be based on constant integers for local memory HOT 2
- performance issue on dynamic shaped tensor HOT 8
- Evaluation for rFactored reductions is not supported. HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π 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. πππ
-
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 β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from fuser.