GPU Computing
← Back to Data Parallelism
Using graphics processing units for general-purpose computation. GPUs have thousands of cores optimized for data-parallel workloads, providing massive throughput for matrix operations, neural network training, simulation, and other embarrassingly parallel tasks.