Get Full Performance from Your Code with Tech.Decoded
Parallel Programming
December 2021
Dive into the details of using Intel® Software Development Tools to deliver compute-intense applications that take advantage of the hardware they’re deployed on. Key topics are time-stamped in the description for easy finding.
[1:59:42]
December 2021
Get a first look at the 2022.1 release of tools designed for cross-architecture parallel programming, portability, and performance.
[59:50]
September 2021
Learn about the Intel® oneAPI DPC++ Library (oneDPL), an alternative software development library for creating heterogeneous applications and solutions.
[1:03:41]
Compilers
January 2022
Learn about the recent evolution of SIMD (single instruction, multiple data) vectorization technology, including how to expose hardware capabilities.
November 2021
FPGA development can be slow due to compile time, among other things. Find out how the Intel® FPGA Add-on for oneAPI Base Toolkit can solve the issue.
[1:27:49]
October 2021
This session of Ask the Compiler Experts takes a closer look at the C++ standard support with ICX (the driver for the Intel® oneAPI DPC++/C++ Compiler).
[56:49]
August 2021
Parallel programming expert James Reinders details Intel’s adoption of LLVM—the benefits, what it means for developers, and key benchmarks.
Analysis & Debugging
January 2022
Performance optimization is a bit like solving a mystery: It can get tricky unless you know where to look. This session shines a light on a key profiling tool that helps you do exactly that, whether you’re targeting one or multiple architectures.
[55:08]
December 2021
Measure and visualize the actual performance of GPU kernels using benchmarks and hardware metric profiling against hardware-imposed performance ceilings and determine the main limiting factor.
November 2021
Find out how Cinesite used Intel® Open Image Denoise to analyze, debug, and render vivid, noise-free, photorealistic visuals for The Addams Family 2*.
Tuning & Optimization
October 2021
Facebook*, Amazon*, and Intel argue that measuring a hyperparameter optimization routine by time, rather than iterations, is a more effective cost metric.
September 2021
Get clarity about the unique and distinct programming needs of these architectures so you can effectively match the appropriate workload to each compute device.
[1:02:04]
September 2021
Find out how the Flow Graph Analyzer feature in Intel® Advisor shows you how applications written in DPC++/SYCL* map algorithms to each device.
[59:06]
September 2021
Get an introduction to the Intel® oneAPI Video Processing Library, a single video-focused API for encoding, decoding, and video processing across CPUs, GPUs, FPGAs, and more.
[56:55]
Training & Inference
November 2021
By replacing the stock scikit-learn and TensorFlow* included in Databricks Runtime for Machine Learning with Intel-optimized versions, training and inference were significantly improved.
November 2021
The stock scikit-learn library is inherently slow on hardware without additional software optimizations. Good news: Intel® Extension for Scikit-learn* is here. This webinar details how to take advantage of it.
[1:00:47]
November 2021
This case study describes the optimizations that were applied to a deep learning-based model for recommendations introduced by Facebook*, and the performance improvements achieved.
August 2021
This article explains how to train ridge regression models using the Intel-optimized scikit-learn, and to do it by altering a single line of code.
Offloading
December 2021
Build and compile an OpenMP* application that's offloaded onto an Intel GPU, and then run analyses with GPU capabilities and examine the results.
November 2021
Learn how to analyze and tune software for optimal performance after it is ported to your target post-release GPU using Intel® VTune™ Profiler.
[40:53]
November 2021
Can your application run faster if you offload some parts of the CPU code to a GPU? The Offload Advisor feature in Intel Advisor lets you know. Fast.
[3:06]
High-Performance Libraries
November 2021
Learn streamlined methods for implementing a 1D Fourier correlation to perform complex mathematical operations that require multiple kernel functions.
[47:37]
September 2021
Learn how to boost the performance of R—a language for statistical computing—by using the Intel® oneAPI Math Kernel Library and without modifying code.
Continuous
Download this open-access modern guide to Threading Building Blocks (TBB) written by leading engineers.
July 2021
Simplify FPGA development by building a shared library in almost any programming language. This presentation includes a demonstration of a fast Fourier transform (FFT) algorithm offload to an FPGA platform.
[58:40]
Resources
oneAPI Application Catalog
Explore hundreds of applications that take advantage of elements of the Intel oneAPI cross-architecture programming model.
Fellow Travelers & Partners
See the ecosystem support for oneAPI from a growing, global list of companies, universities, and institutions.
Forums & Support
Find answers to technical questions, plus links to community and priority support and forums for Intel oneAPI Toolkits and individual tools.
oneAPI Projects
See the growing collection of heterogeneous-distributed applications your peers and colleagues are creating using Intel oneAPI tools.