url: https://github.com/tensorflow/tensorflow/issues/30227

https://github.com/tensorflow/tensorflow/issues/30227#issuecomment-506783788

Searching deeply, I found that the first timestep is also used to determine the cell output shape and its dtype.

(py3-tf2-gpu) sephiroce@bike:/usr/local/cuda/samples/5_Simulations/nbody$ ./nbody -benchmark -numbodies=2560000 -device=0

Run "nbody -benchmark [-numbodies=]" to measure performance.

-fullscreen (run n-body simulation in fullscreen mode)

-fp64 (use double precision floating point values for simulation)

-hostmem (stores simulation data in host memory)

-benchmark (run benchmark to measure performance)

-numbodies= (number of bodies (>= 1) to run in simulation)

-device= (where d=0,1,2.... for the CUDA device to use)

-numdevices= (where i=(number of CUDA devices > 0) to use for simulation)

-compare (compares simulation results running once on the default GPU and once on the CPU)

-cpu (run n-body simulation on the CPU)

-tipsy= (load a tipsy model file for simulation)

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

> Windowed mode

> Simulation data stored in video memory

> Single precision floating point simulation

> 1 Devices used for simulation

gpuDeviceInit() CUDA Device [0]: "Ampere

> Compute 8.6 CUDA device: [GeForce RTX 3090]

number of bodies = 2560000

2560000 bodies, total time for 10 iterations: 69005.547 ms

= 949.721 billion interactions per second

= 18994.415 single-precision GFLOP/s at 20 flops per interaction

1. install boost using python3

ref: https://github.com/pupil-labs/pupil/issues/874, huangjiancong1

tar -xzvf boost_1_65_1.tar.gz

cd boost_1_65_1

echo "using mpi ;

using gcc : : g++ ;

using python : 3.6 : /usr/bin/python3 : /usr/include/python3.6m : /usr/local/lib ;" > ~/user-config.jam

./bootstrap.sh --with-python=/usr/bin/python3 --with-python-version=3.6 --with-python-root=/usr/local/lib/python3.6 --prefix=/usr/local

sudo ./b2 install -a --with=all

2. install rdkit

modifying CMakeList boost version 1.5.1 to the installed version of boost.

change all the path below!

cmake version needs to be ~= 3.1

cmake -DPYTHON_LIBRARY=/usr/lib/python3.6/config/libpython3.6.a \

-DPYTHON_INCLUDE_DIR=/usr/include/python3.6/ \

-DPYTHON_EXECUTABLE=/usr/bin/python3 \

-DBOOST_LIBRARIES=libboost_python3.so.1.65.1 \

-DBoost_INCLUDE_DIR=include_foldr ..

3. Add rdkitpath to PYTHONPATH, libpath to LD_LIBRARY_PATH

https://github.com/tensorflow/tensorflow/issues/42146#issuecomment-671484239

Message: "Consider either turning off auto-sharding or switching the auto_shard_policy to DATA to shard this dataset."

If your Tensorflow scripts leave this log message, then it falls back to use DATA type sharding. Thus, to turn off the log message you can set auto_shard_policy to DATA using tf.data.Options() as follows:

options = tf.data.Options()

options.experimental_distribute.auto_shard_policy = tf.data.experimental.AutoShardPolicy.DATA

dataset = dataset.with_options(options)