A prototype MPI/CUDA stencil halo exchange library
Install MPI, CUDA, and CMake 3.17+, then
git clone [email protected]:cwpearson/stencil.git
cd stencil
mkdir build
cd build
cmake ..
make
make test
- Maximal communication overlap: each GPU communicates with neighbors in a point-to-point scheme with all communication overlapped
- Fast intra-node communication: communication can occur entirely through CUDA even for different ranks on the same node.
- CUDA-aware MPI support: Can be enabled if desired (
cmake -DUSE_CUDA_AWARE_MPI=ON) - Automatic partitioning: partition data to minimize communication volume
- Topology-aware data placement: place data to maximize communication bandwidth for large messages
- Communication/Computation Overlap: API to query grid regions that can be used concurrent with communication.
- Friendly: Access data in terms of grid coordinates, not memory offsets.
stencil optimized communication with CUDA, NUMA, and MPI, depending on whether those are available in the environment.
Depending on availability, the following compiler defines exist:
| Capability | Available | Not Available |
|---|---|---|
| MPI | STENCIL_USE_MPI=1 |
STENCIL_USE_MPI=0 |
| CUDA | STENCIL_USE_CUDA=1 |
STENCIL_USE_CUDA=0 |
Tested on
- CUDA 10.1 / 10.2 / 11.0
- OpenMPI 2.1
Create a toolchain file, like cmake/toolchains/deneb-openmpi-4.0.5.cmake
clear your build directory, and invoke CMake like this
cmake -DCMAKE_TOOLCHAIN_FILE=../cmake/toolchains/deneb-openmpi-4.0.5.cmake ..
Install MPI and CUDA, then
make && make test
To run specific tests
test/test_cpu "<case name>" -c "<section name>"
With USE_CUDA_GRAPH=ON, errors may be reported inconsistently, very late, or be nonsense (e.g., 715 invalid instruction when a stream was on a wrong device).
Using a newer CUDA version can improve aspects of error reporting.
If there is an error, it probably indicates an actual problem, just not necessarily the problem CUDA reports.
It may be useful to run tests under cuda-memcheck.
cuda-memcheck may report spurious errors from cudaDeviceEnablePeerAccess, so we disable reporting API errors.
All API calls are checked at runtime at all optimization levels, so errors will always surface.
With USE_CUDA_GRAPH=ON, cuda-memcheck may not report errors.
cuda-memcheck --report-api-errors no test/test_cuda
Some tests are tagged:
MPI tests only
test/test_all "[mpi]"
CUDA tests only
test/test_all "[cuda]"
With the default profiling settings, we sometimes see a crash on Nsight Systems 2019.3.7 on amd64. Restrict profiling to CUDA, NVTX, and OS calls.
nsys profile -t cuda,nvtx,osrt mpirun -n <int> blah
to enable IP sample, backtrace, and scheduling data collection
sudo sh -c 'echo 1 >/proc/sys/kernel/perf_event_paranoid'
Use the Nsight Systems application to view the resulting qdrep file.
Make sure your nsight-sys version at least matches the nsys version used to collect the profile.
Nsight Systems 2019.5.2 allows -t mpi, but on amd64 it causes the importer to hang.
nsys profile -t nvtx,cuda,mpi mpirun -n <int> blah
mpirun -n <int> nvprof -o timeline_%p.nvvp ...
To mount a remote directory (where there are nvprof files to load):
sshfs -o IdentityFile=/path/to/id_rsa user@host:/path /mount/location
cd build
rm -f CMakeCache.txt
cmake -DCMAKE_PREFIX_PATH=path/to/mpi ..
mpiexec from the MPI will be soft-linked into the build directory for convenience
This mpiexec should be used to launch any binaries.
./mpiexec -n 2 bin/<exe>
Setting an MCA param
mpirun --mca mpi_show_handle_leaks 1 -np 4 a.out
Checking for CUDA-Aware MPI support:
ompi_info --parsable --all | grep mpi_built_with_cuda_support:value
./configure --prefix="blah" --with-cuda=/path/to/cuda
apt-get install bison
./configure --prefix="blah: --enable-cuda --with-cuda=/path/to/cuda
may need a quick patch for CUDA 11:
src/mpid/ch3/channels/mrail/src/gen2/ibv_cuda_util.c:844:37: error: ‘struct cudaPointerAttributes’ has no member named ‘memoryType’
844 | is_dev = (attributes.memoryType == cudaMemoryTypeDevice) ? 1 : 0;
the struct member should be attributes.type
may also need
--disbale-mcast --enable-fortran=no
Run with MV2_USE_CUDA=1.
./configure --prefix="blah" ---with-device=ch4 -with-ch4-shmmods=gpudirect
uncertain of the difference between ch4 and ofi.
may also need
--enable-fortran=no
Run scripts are in scripts/hal
srun --partition=gpu --time=4:00:00 --ntasks=2 --nodes=2 --ntasks-per-node=1 --sockets-per-node=1 --cores-per-socket=8 --threads-per-core=4 --mem-per-cpu=1200 --wait=0 --export=ALL --gres=gpu:v100:2 --pty /bin/bash
SpectrumMPI with CUDA-Aware MPI
module load spectrum-mpi
cmake .. -DUSE_CUDA_AWARE_MPI=ON
make
Show reservations: scontrol show res
Show queue: swqueue
Suggest using the latest version of CUDA:module load cuda/11.0.3.
Earlier versions work but have poor error reporting from CUDA graphs and cannot overlap CUDA graph execution and data transfer.
Run scripts are in scripts/summit.
nsight-systems 2020.3.1.71 can crash with the osrt or mpi profiler turned on.
Disable with nsys profile -t cuda,nvtx.
To control the compute mode, use bsub -alloc_flags gpudefault (see https://www.olcf.ornl.gov/for-users/system-user-guides/summitdev-quickstart-guide/#gpu-specific-jobs)
To enable GPUDirect, do jsrun --smpiargs="-gpu" ... (see https://docs.olcf.ornl.gov/systems/summit_user_guide.html, "CUDA-Aware MPI")
To run tests, do bsub 1node_test.sh or get an interactive node (bsub -W 2:00 -q debug -nnodes 1 -P csc362 -alloc_flags gpudefault -Is /bin/zsh) and run that script.
During configuration, you may see an error like this which causes cmake to fail:
CMake Error: Remove failed on file: /ccs/home/cpearson/repos/stencil/build/CMakeFiles/CMakeTmp/CMakeFiles/cmTC_50eb9.dir/.nfs000000001473310900000eaf: System Error: Device or resource busy
Re-run cmake again.
First, get some paraview files: for example, mpirun -n 2 bin/jacobi3d 60 60 60 --paraview --iters 1000.
File>Open. Open individually, not as a group- Ensure the delimiters are correct (
,), then hitApplyon thePropertiestab of the Pipeline Browser. This may hang for a while while the files are loaded. - Select the two files, then
Filters>Common>Group Datasets. Apply. This will create a new combined GroupDatasets1 in the pipeline browser.
- Ensure the delimiters are correct (
- Select the dataset and
Filters>Alphabetical>Table to Points- Select the x,y,z columns in the
Propertiestab (should beX,Y, andZrespectively). - Click on the vizualization window
Apply. A bunch of points should appear.
- Select the x,y,z columns in the
- In the
Coloringsection of thePropertiestab, choose the name of the data field. - Select the TableToPoints and add the
Glyphfilter.- Glyph type
Box - Scale Factor
1.0 - Glyph Mode
All points - Use the eye in the Pipeline browser to turn off the TableToPoints and leave the glyph turned on
- Glyph type
Experimental implementation of the sataroth code using this library for halo exchange.
mpirun -n <N> astaroth/astaroth
The astaroth configuration is defined in astaroth/astaroth.conf
-
v1 (iWAPT)
- joint stencils over multiple data types (Astaroth)
- uneven partitioning
- edge communication (Astaroth)
- corner communication (Astaroth)
- face communication (Astaroth)
- overlap MPI and CUDA
- Same-GPU halo exchange with kernels
- Same-rank halo exchange with kernels
- Same-rank halo exchange with
cudaMemcpyPeer - co-located MPI rank halo exchange with with
cudaIpc...andcudaMemcpyPeer - automatic data placement in heterogeneous environments
- Control which GPUs a distributed domain should use
DistributedDomain::use_gpus(const std::vector<int> &gpus)
- Control which exchange method should be used
-
v2 (Thesis)
- ParaView output files
DistributedDomain::write_paraview(const std::string &prefix) - support uneven radius
- "Accessor Object" for data
- Index according to point in compute domain
- Support overlapped computation and communication
- interface for extracting interior/exterior of compute region for kernel invocations
-
cudaMemcpy3D - CUDA runtime timer
- pitched allocation
- sort communication for weak scaling?
- ParaView output files
-
v3
- Message bundling
- Improved performance by sending all data for a remote node in a single message?
- allow a manual partition before placement
- constrain to single subdomain per GPU
- Message bundling
-
future work
- Autodetect CUDA-Aware MPI support
- testing at build time with
ompi_info MPI_T_cvar_read/MPI_T_cvar_get_info?
- testing at build time with
- N-Dimensional data
- with cuTensor?
- selectable halo multiplier
- fewer, larger messages and less frequent barriers
- larger halo allocations
- factor out placement solver code
- Autodetect CUDA-Aware MPI support
-
needs info
- exchange subset of quantities
- this would potentially split some of the operation bundling opportunities (pack, CUDA graph, ...)
- exchange is not async now, e.g. CPU polls communicators. May want a worker thread or something.
- mesh refinement
- we would rely on the user to specify the conditions
- inferface for asking for refinement?
- how to rebalance load and communication
- would require test machine with non-homogeneous intra-node communication
- mapping multiple subdomains to each GPU
- exchange subset of quantities
-
wontfix
- Non-rectangular regions
- probably getting to close to a general task-based runtime at that point (like Legion)
- Remove requirement of CUDA (HPCG)
- research vehicle for CUDA + MPI
- Non-rectangular regions
include/stencil/rcstream.hpp
A C++-style class representing a shared CUDA stream. The underlying stream is released when the reference count drops to zero. Each stream can be explicitly associated with a device.
include/stencil/gpu_topology.hpp
The Distance Between GPUs is computed by using Nvidia Management Library to determine what the common ancestor of two GPUs is. This is combined with other NVML APIs to determine if two GPUs are directly connected by NVLink, which is considered the closest distance.
Various repeated communication patterns are accelerated through the CUDA graph API.
A construction like this can inadvertently time evaluation of argument expressions:
auto start = get_wall_time();
cudaDoSomething(a.lots_of_work())
auto elapsed = get_wall_time() - start();
use stencil/include/rt.hpp to time only the cuda operation after arguments have been evaluated:
rt::time(cudaDoSomething, a.lots_of_work())
Similarly, for MPI operations
mpi::time(MPI_Isend, srcBuf, ...);
The timers (include/stencil/timer.hpp) use compiler barriers to prevent reordering of instructions around the system calls.
- Don't put state in abstract classes
- use signed integer types for subscripts