@Intel
’s hub for all things
#oneAPI
— an open source model for
#developers
looking to accelerate and optimize heterogeneous workloads across architectures.
The latest issue of The Parallel Universe is out! Inside you can explore how far
#oneAPI
has come since 2018, and how Intel’s new built-in AI acceleration engines are ushering in a new era of accelerated
#AI
on Intel CPUs:
What does
#AI
have to do with hunting for dinosaur bones? AI Software Product Manager,
@SusanKahler3
, explains how AI can be used to aid in finding Jurassic-period bones. Read more:
#oneAPI
@IntelAI
#ICYMI
:
@GoogleAI
and
@intel
teamed up to enable the
#oneAPI
oneDNN library as the default backend CPU optimization in
@TensorFlow
2.9! Millions will benefit from AI performance acceleration, driven by open software optimizations, without code changes.
Intel Arc Pro GPUs is a next step in providing open, end-to-end solutions to scale performance like the Intel Open VKL plugin for
@pixarsrenderman
, the Intel Open Path Guiding Library (OpenPGL), and more. Learn more:
#SIGGRAPH2022
#oneAPI
@Blender
Curious about the process of converting
#CUDA
apps to
#SYCL
?
@codeproject
provides a hands-on view of the steps to simplify code migration for cross-platform development.
#oneAPI
We are thrilled to announce that
@Altair_Inc
has moved Radioss to
#opensource
using the
#oneAPI
Base and HPC Toolkits. Learn more about this exciting development here:
Achieve fast and efficient code development with a complete set of cross-architecture tools making heterogeneous development easier. Explore
#oneAPI
today:
#SYCL
#AI
#HPC
#DevTools
Intel's Open Path Guiding Library (OpenPGL) has now been integrated into
@Blender
3.4 Cycles helping reduce noise in scenes where finding a path to light is difficult for regular path tracing. See how to access OpenPGL:
#oneAPI
#rendering
#raytracing
Grab a cup of coffee & spend your morning with
@intel
's VP of AI & Analytics, Wei Li, &
@Accenture
's Applied Intelligence CTO, Jean-Luc Chatelain (
@InformationCTO
) as they discuss strategies to scale AI everywhere at the
#oneAPI
DevSummit for
#AI
on 7/12.
Have you seen the Optimize Applications for Intel GPU with Intel
#VTune
Profiler User Guide, targeting Intel GPUs for computation, media client and server applications? Check it out.
#oneAPI
#GPU
#VTuneTuesdays
We’re making big commitments — a software-first approach, built on core values of staying open and secure and delivering the best developer tools and experience. Learn more about our software-first approach.
#opensource
#oneAPI
#crossarchitecture
The
#oneAPI
DevSummit at
@ISChpc
'22 focuses on the evolving and complementary forces of
#HPC
and
#AI
. Take a peek at the agenda, full of brand-new content and speakers and make sure to register today:
#ISC22
#SYCL
@SYCLstd
See how
@Lenovo
is using the
#oneAPI
#AI
Analytics Toolkit to deliver pre-tuned AI frameworks for many data science workflows that can help drastically increase speed and efficiency.
Deploy more efficient
#deeplearning
models with the Intel Neural Compressor. It improves productivity with a lighter binary size, a new quantization accuracy feature, and much more. See the latest:
#oneAPI
#AI
#DL
Intel
#VTune
Profiler 2022.3 release is out! This release introduces support for DirectML API to pinpoint call inefficiencies and introduces new memory metrics to optimize memory-bound GPU code. Learn more:
#oneAPI
#VTuneTuesdays
Experience how Intel Arc and Data Center GPUs are supported through
#oneAPI
. This session will show how to use a single codebase in
#SYCL
across multiple CPU and GPU architectures for GPUs in
@Blender
.
@BlenderDev
@BlenderArtists
Using a real-world application, see a step-by-step tutorial on improving your GPU workload performance by running and applying guidance from the Intel VTune Profiler—an analysis tool powered by
#oneAPI
.
#CPU
#GPU
#SYCL
Intel Embree 4.2 delivers significantly faster rendering speed on Intel GPUs vs. CPUs & supports easier rendering using a single
#SYCL
codebase. See how experts create real-time
#rendering
for next-gen innovations in
#raytracing
,
#scivis
, &
#3D
.
#oneAPI
Employing a universal
#AI
platform provides the flexibility to run every AI code, scope to empower developers, & scale to enable AI everywhere.
@VentureBeat
discusses Intel's AI platform, including the AI Analytics Toolkit—powered by
#oneAPI
.
#IntelXeon
Tools like
#oneAPI
remove code barriers allowing developers to increase productivity and get to market faster.
@GregL_Intel
details why embracing an open ecosystem allows for greater community engagement.
The Intel Open Volume Kernel Library (OpenVKL) — part of the
#oneAPI
Rendering Toolkit — improves performance and memory efficiencies with additional support for VDB volumes and the latest Intel processors.
#rendering
#raytracing
#opensource
#DevTools
Achieve state-of-the-art video processing with the Intel
#oneAPI
Video Processing Library (oneVPL). It provides a single, video-focused API for encoding, decoding, and video processing across a range of accelerators.
#Rendering
#CPU
#GPU
Recently,
@ServeTheHome
took a tour of the
#IntelDeveloperCloud
and looked at the hardware and systems it takes to provide a development cloud-based solution to test and run
#AI
workloads on the latest Intel architectures. Watch the video here:
Get the fastest & most-used math library for Intel-based systems with the Intel
#oneAPI
Math Kernel Library (oneMKL). This library is optimized to help accelerate math processing routines, increase application performance, & more.
#openecosystem
#OpenMP
Don't get spooked by proprietary lock-in. With
#oneAPI
, get the freedom to use a single code base across multiple architectures. Discover the toolkits:
#Halloween
#HPC
#Rendering
#AI
Transitioning to heterogeneous computing is the future. Empowering developers to target multiple XPUs with a single code base is a game changing initiative. Check out the latest from
@VentureBeat
on the
#oneAPI
initiative.
@SYCLstd
@IntelSoftware
#SYCL
Join experts from
@codeplaysoft
,
@intel
and other companies to learn how to apply Intel's work on open, standards-based accelerated computing models. Register for Intel Innovation 2022 and maximize your computing performance!
#oneAPI
#IntelON
#SYCL
In the
@Blender
3.6 LTS release, content creators and
#rendering
experts can access the high-performance
#raytracing
library, Intel Embree. This can decrease rendering times significantly with considerable performance increases. Learn more:
#oneAPI
AI developers rejoice—
@PyTorch
2.0 is here! With this latest release, get access to four features contributed by
@intel
that will help
#AI
developers optimize performance for their inference workloads. Take a deep dive here:
#oneAPI
#AI
We’re developer-centric and committed to a software-first approach. oneAPI contributes to an open and trustworthy ecosystem across vendors and architectures. Learn more here.
#opensource
#oneAPI
#crossarchitecture
We are excited to announce that we have joined the
@PyTorch
Foundation as a Premier member! This is a significant step in the joint mission to continue to drive the adoption of
#AI
and
#deeplearning
through an
#openecosystem
. Learn more:
#oneAPI
Accelerate your
#AI
software performance with the latest 2023.1 AI Analytics Toolkit release. Developers & data scientists can more effectively & efficiently accelerate end-to-end training & inference of their AI workloads.
#oneAPI
@TensorFlow
@PyTorch
The
#oneAPI
2023.1 Toolkit release is now available! With updates to the oneAPI DPC++/C++ Compiler, Fortran Compiler,
#VTune
Profiler, and more. Explore the latest:
#openecosystem
Get hands-on with Intel's library of tools for end-to-end data science workflows and AI accelerators at this year's Intel Innovation. Register for your spot here:
#oneAPI
#InteON
#AI
#datascience
Using the Intel
#oneAPI
open multiarchitecture programming & tools, the team at
@UKAEAofficial
gained code portability, performance, and productivity working on a fusion powerplant in the United Kingdom. See how it was done:
#AI
#HPC
#OpenEcosystem
The Intel Extension for
@TensorFlow
is an extension of the stock,
#opensource
TensorFlow library optimized for high-performance computing on Intel architectures. In this article, get a detailed overview, plus get started with a free code sample!
#oneAPI
New in the
#oneAPI
#HPC
Toolkit:
- Enhanced
@OpenMP_ARB
performance with the oneAPI C++ Compiler, based on LLVM tech
- Additional offload support with the Fortran Compiler, based on
#LLVM
tech
- Improved capability support of the latest OpenMP 5.1
The Intel Optimization for TensorFlow has been updated to
@TensorFlow
2.9.1. This latest experience allows you to enable oneDNN optimizations on Linux-based Cascade Lake & newer CPUs by removing compiler requirements. Learn more:
#oneAPI
#AI
#DL
Hear Steve Hikida, Intel VP and GM of Compiler Engineering and James Brodman, Sr. Principal Engineer in Intel’s Technology discuss how compilers have evolved and what's in store for the future of compilers. Read the Q&A here:
#oneAPI
#compiler
@llvmorg
At the
#oneAPI
DevSummit, hear from experts on
@PyTorch
's recent endeavors and the project's future, as well as how TraffikAI is enhancing traffic systems efficiency and improving traffic management. Sign up here:
#AI
#DataScientists
#Developers
We are excited to announce our collaboration with
@GoogleAI
to enable
#oneAPI
oneDNN library as the default backend CPU optimization in
#TensorFlow
2.9!
@TensorFlow
users will now see AI performance acceleration, driven by our open software optimizations, with no code changes!
TensorFlow 2.9 has been released! 🎉
Highlights include:
👉 Performance improvements with oneDNN
👉 Release of DTensor, a new API for model distribution that can be used to seamlessly move from data parallelism to model parallelism
Learn more ↓
Enterprises & developers in
#AI
& machine learning can take advantage of the Intel AI Portfolio as a resource for productivity & performance at scale to allow you to accelerate your AI journey from the edge to the cloud.
#MachineLearning
#ML
#DataScience
.
@ebenbuild
provides patients suffering from Acute Respiratory Distress Syndrome (ARDS) with precision therapy based on personalized digital twins. As part of the
#oneAPI
startup program, they'll advance their mission using the toolkits.
See how the libraries and tools behind the Intel Distribution of
#Python
in conjunction with the Intel
#oneAPI
Math Kernel Library (oneMKL), can help developers achieve multithreading and multiprocessing to accelerate your
#AI
and
#ML
workflows.
Boost your
@scikit_learn
code with almost zero effort as part of the
#oneAPI
#AI
Analytics Toolkit. See how to run the same scikit-learn code much faster, where to look for resources and code samples, and go over examples with performance comparisons.
With the latest
#oneAPI
Toolkits 2023.2 release, developers can enjoy a simplified migration from CUDA to
#SYCL
,
#CPU
to
#GPU
development to
#Python
, and new optimizations on the latest
#IntelXeon
and GPU platforms. See the latest:
#AI
#HPC
The Intel
#oneAPI
DPC++ Library (oneDPL) allows developers to create heterogeneous solutions with accelerated
#SYCL
kernels on CPUs, GPUs, and FPGAs. This article gives you an overview of the tool, including code samples. Check it out:
@SYCLstd
#cpp
The latest Parallel Universe issue is now available! This issue covers generative AI, optimizing large language models with
#oneAPI
, aligning
#SYCL
with future C++ language concepts, and much more. Check it out:
Experts from
@googlecloud
accelerated their workloads by adopting tools & libraries from the Intel
#oneAPI
HPC Toolkit—like the Intel MPI Library & VTune Profiler—to boost performance & migrate workloads into the cloud. See how others are using oneAPI:
Get the most out of your
#rendering
projects with the latest
#oneAPI
Rendering Toolkit 2022.3 release! Developers will enjoy updates to Intel Embree, enhancements to Intel OpenVKL, and additional support for Intel OSPRay. Learn more:
#raytracing
Software engineers Sebastian Herholz of
@intel
and Dian Nikolov of
@ChaosGroup
sat down with
@tonymongkolsmai
to discuss how Chaos's V-Ray and the Intel Open Path Guiding Library (Open PGL) are making path guiding faster. Listen here:
#oneAPI
#raytracing
Now on-demand—learn how to migrate your CUDA code to
#SYCL
to enable developers to create portable applications that can be deployed on multiple architectures. Including code samples for SYCLomatic.
#oneAPI
@SYCLstd