Posts

DirectML platform architecture showing Direct3D 12 integration for GPU-accelerated machine learning

Deep Learning for Graphics Programmers: Performing Tensor Operations with DirectML and Direct3D 12

In the rapidly evolving landscape of machine learning and artificial intelligence, harnessing the …

Side-by-side code comparison of matrix multiplication in SYCL, OpenCL, and CUDA

Comparing SYCL, OpenCL, and CUDA: Matrix Multiplication Example

Matrix multiplication is a core operation in scientific and engineering applications, often …

DirectX 12 graphics pipeline diagram showing command lists, PSOs, root signatures, and descriptor heaps

Intro to DirectX 12 Pipeline

DirectX 12 organizes graphics rendering into pipelines.

Components of DirectX 12 Pipeline:

  1. Command …

PyTorch 2.0 compiler architecture showing Dynamo graph capture and AOT Autograd flow

The Simple Path to PyTorch Graphs: Dynamo and AOT Autograd Explained

Graph acquisition in PyTorch refers to the process of creating and managing the computational graph …

PyTorch Profiler chrome trace view showing operator execution timeline for ResNet model

Profiling ResNet Models with PyTorch Profiler for Performance Optimization

In the realm of deep learning, model performance is paramount. Whether you’re working on image …

Benchmark comparison chart of PyTorch, TorchScript, and ONNX inference throughput on Intel Arc 770

Accelerating Deep Learning Inference on Intel Arc 770: ONNX and PyTorch Go Head-to-Head

When deploying deep learning models, the choice of framework can significantly impact performance. …