Integrating new frameworks¶
This section details some of the configuration options and some of the environment variables that can be used to tune for optimal performance when your system already has a version of nGraph installed with one of our supported backends.
|Backend||Current nGraph support||Future nGraph support|
|Intel® Architecture Processors (CPUs)||Yes||Yes|
|Intel® Nervana™ Neural Network Processor™ (NNPs)||Yes||Yes|
|NVIDIA* CUDA (GPUs)||Yes||Some|
|Field Programmable Gate Arrays (FPGAs)||Coming soon||Yes|
Regardless of the framework, after the nGraph Library for backends step, a good place to start usually involves making the libraries available to the framework. On Linux* systems built on Intel® Architecture, that command tends to looks something like:
export NGRAPH_CPP_BUILD_PATH=path/to/ngraph_dist/ export LD_LIBRARY_PATH=path/to/ngraph_dist/lib/
FMV stands for Function Multi-Versioning, and it can also provide a number of generic ways to patch or bring architecture-based optimizations to the Operating System that is handling your ML environment. See the GCC wiki for details.
If your nGraph build is a Neural Network configured on Clear Linux* OS for Intel® Architecture, and it includes at least one older CPU, the following article may be helpful.
Training Deep Neural Networks¶
Before tweaking various environment variables, be aware that how the computation
gets executed depends upon the ordering of the data format that the model is
NCHW are the two more common layouts in Deep Learning
models. Your ultimate runtime can vary greatly – even when all other factors
are exactly the same – when this detail is overlooked.
For CPU (and most cuDNN) backends, the preferred layout is currently
- N – Number of images per batch
- C – Channel of the image (expressed as a number like 3 for RGB and 1 for grayscale)
- H – Height of the image
- W – Width of the image
Intel® Math Kernel Library for Deep Neural Networks¶
-The following KMP options were originally optimized for models using the
Intel® MKL-DNN to train models with the
NCHW data layout; however, other
configurations can be explored. MKL-DNN is automatically enabled as part of an
nGraph compilation; you do not need to add MKL-DNN separately or as an
additional component to be able to use these configuration settings.
KMP_BLOCKTIMESets the time, in milliseconds, that a thread should wait after completing the execution of a parallel region, before sleeping.
KMP_AFFINITYEnables the runtime library to bind threads to physical processing units.
true) or disables (
false) the printing of OpenMP* runtime library environment variables during program execution.
OMP_NUM_THREADSSpecifies the number of threads to use.
nGraph-enabled Intel® Xeon®¶
The list below includes recommendations on data layout, parameters, and application configuration to achieve best performance running DNN workloads on Intel® Xeon® (CPU processor) systems.
The number of threads set by
OMP_NUM_THREADS ought not exceed the number of
physical cores. The threads should be pinned to their respective physical cores
and activated as follows:
Buffer pointers should be aligned on 64-byte boundaries. NUMA policy should be
configured for local memory allocation (
- When running inference, or training for forward-propagation and weight
updates, for best performance:
- the number of input channels should be 1, 3, or a multiple of SIMD-width (8 for AVX2 systems, 16 for AVX512 systems).
- the number of output channels should be a multiple of SIMD-width (8 for AVX2 systems, 16 for AVX512 systems).
- When training backward propagation, the number of input and output
channels should be a multiple of SIMD-width (8 for AVX2 systems, 16 for AVX512
- padding should not exceed \(0.5x\) where \(x\) is the kernel size.
- kernel width should be less than 14.
The best resource for this configuration option is the gnu.org site
OMP_NUM_THREADS defaults to the number of logical cores. To check the
number of cores on your system, you can run the following on the command-line to
see the details of your CPU:
Intra-op and inter-op parallelism¶
Some frameworks, like TensorFlow*, use these settings to improve performance; however, they are often not sufficient for optimal performance. Framework-based adjustments cannot access the underlying NUMA configuration in multi-socket Intel® Xeon® processor-based platforms, which is a key requirement for many kinds of inference-engine computations. See the next section on NUMA performance to learn more about this performance feature available to systems utilizing nGraph.
NUMA stands for Non-Uniform Memory Access. It indicates how each CPU can access memory attached to each socket.
Without the “knowledge” of CPU socket and NUMA configuration, a simple thread affinity (as in the case of thread pool) does not lead to optimal performance. In fact, it can sometimes prohibitively decrease throughput; a core from socket 0 might have to continually access cache lines from the memory bank of socket 1, increasing bandwidth demands on the Intel® Ultra-Path Interconnect (Intel® UPI). This situation is exacerbated with larger number of sockets found in 4, 8, and 16-socket systems. We believe that users need to be aware of system level optimizations in addition to framework specific configuration parameters to achieve the best performance for NN workloads on CPU platforms. The nGraph Compiler stack runs on transformers handled by Intel® Architecture (IA), and thus can make more efficient use of the underlying hardware.