bsprings
Re:CUTLASS, a CUDA C++ template library for matrix multiply on GPUs
on Wednesday December 06, 2017 @04:33PM (#15865221)
For most formats it achieves >90% of the performance of the hand-tuned assembly implementations in cuBLAS.
RESTful Inference with the TensorRT Container and Nvidia GPU Cloud
How Jet.com Built a GPU-Accelerated Fulfillment Engine with F# and CUDA
Malware Detection in Executables Using Neural Networks
DeepStream: Next-Generation Video Analytics for Smart Cities
Programming Tensor Cores in CUDA 9
Mixed-Precision Deep Learning Training: Perf of FP16 with Accuracy of FP32
Training AI for Self-Driving Cars: Challenges of Scale
Cooperative Groups: Flexible CUDA Thread Programming
Re:Numba: High-Performance Python with CUDA Acceleration
on Thursday September 21, 2017 @06:40PM (#15307952)
When I originally wrote the post in 2013, the GPU compilation part of Numba was a product (from Anaconda Inc., nee Continuum Analytics) called NumbaPro. It was part of a commercial package called Anaconda Accelerate that also included wrappers for CUDA libraries like cuBLAS, as well as MKL acceleration on the CPU.
Continuum gradually open sourced all of it (and changed their name to Anaconda). The compiler functionality is all open source within Numba. Most recently they released the CUDA library wrappers in a new open source package called pyculib.
Some other minor things changed, such as what you need to import. Also, the autojit and cudajit functionality is a bit better at type inference, so you don't have to annotate all the types to get it to compile.
We thought it was a good idea to update the post in light of all the changes.
Accelerated XGBoost Gradient Boosting on GPUs
SigOpt: Deep Learning Hyperparameter Optimization with Competing Objectives
2X Faster Embedded Small Batch Inference with Jetpack 3.1
Deep Learning for Automated Driving with Matlab
AI Co-Pilot: Driver Assistance via RNNs for Dynamic Facial Analysis
GOAI: Open, GPU-Accelerated Data Analytics (demo Walkthrough)
Explaining how end-to-end deep learning steers a self-driving car
Re:Caffe2 adds 16 bit floating point training support on the NVIDIA Vo...
on Friday May 12, 2017 @02:08AM (#14321771)
Tensor Cores: 120 TFLOP/s mixed-precision (peak). Typo in your table: V100 FP64 is 7.5 TFLOP/s.