Intel's Intelligent Systems Lab Organization released Open3D 0.19 as the latest iteration of this open-source library for 3D data processing in Python and C++. Open3D is an open-source project from Intel for dealing with 3D data structures, various 3D data processing algorithms, and also a visualization viewer with support for physically based rendering and more. Significant with the new Open3D 0.19 release is introducing experimental cross-platform GPU support using the SYCL single-source C++ abstraction layer. For those with SYCL GPU support this can work across vendors and provide much faster 3D data processing.

The SYCL Working Group has announced the release of Revision 11 of the SYCL 2020 Specification, introducing eight powerful new extensions alongside numerous specification clarifications that demonstrate the Working Group's continued commitment to advancing the specification for the benefit of both developers and implementers.

Intel has been working on SYCL support within LLVM and various related efforts like the LLVM SPIR-V back-end as part of their oneAPI ambitions and Data Parallel C++ across their spectrum of hardware. The latest hitting upstream LLVM is libsycl as a SYCL run-time library. Libsycl will be ready for the LLVM 22 release due out around March of next year.

The Khronos Group is celebrating a significant milestone as SYCL marks its tenth anniversary. First ratified and published in May 2015, SYCL has grown from an ambitious specification to a cornerstone of heterogeneous computing, enabling developers to write single-source C++ code for a wide range of accelerators, including CPUs, GPUs, FPGAs, and specialized AI hardware. 

SYCL’s provisional spec was introduced in March 2014 with the first demo at SuperComputing 2014, and the initial SYCL specification was ratified and published in 2015. Since then, SYCL has evolved significantly to address the growing demands of heterogeneous computing. What began as a C++ programming model for OpenCL has transformed into a robust, vendor-neutral standard that powers some of the world's most demanding applications across high-performance computing, artificial intelligence, scientific research, and desktop applications

UXL Foundation's Language SIG was the first group to meet earlier this year and there was a presentation from Intel’s Tom Honermann on the progress being made to upstream the DPC++ SYCL compiler project to the LLVM Clang project. This was an interesting insight into the technical decisions and work involved in integrating this SYCL compiler into the Clang compiler project.

In this Codeplay blog, they examine the limitations of the current SYCL images API design and demonstrate how the bindless images extension addresses these challenges. They explore the extension’s core concepts, walk through a practical example, showcase its key benefits, and discuss our plans for the future of bindless images in SYCL.

SYCL continues to grow as developers contribute and provide feedback, shaping the future of this portable programming model. Recently, the SYCL Working Group announced the release of Revision 9 of the SYCL 2020 Specification and exciting upcoming features. The release of SYCL 2020 Revision 9 enhances specification readability and eases implementation.

SYCL Working Group members will be at Supercomputing (SC24) to showcase recent and roadmap developments and offer in-depth, hands-on experience in SYCL programming for high-performance computing.

Khronos Group’s ANARI and SYCL will each be represented Supercomputing 2024 in Atlanta, Georgia on November 17-22. ANARI will have a Birds of a Feather (BOF), and SYCL will also have a BOF along with a hands-on tutorial. Come talk to us!

PyTorch 2.4 now supports Intel® Data Center GPU Max Series and the SYCL software stack, making it easier to speed up your AI workflows for both training and inference. This update allows a consistent programming experience with minimal coding effort and extends PyTorch’s device and runtime capabilities, including device, stream, event, generator, allocator, and guard, to seamlessly support streaming devices. This enhancement simplifies deploying PyTorch on ubiquitous hardware, making it easier for you to integrate different hardware back ends.

In a world where AI, HPC and Safety-Critical acceleration is shifting toward heterogeneous architectures that integrate processors with different architectures from multiple vendors, the need for seamless interoperability and shared open standards has never been more critical. That’s why the UXL Foundation (Unified Acceleration) and the Khronos Group have entered into a liaison agreement to help accelerate the evolution of open accelerated heterogeneous programming.

Intel announced that the oneAPI DPC++/C++ Compiler has completed conformance testing and is the first compiler officially supporting the full SYCL 2020 specification from the Khronos Group. Having it as a complete reference implementation is a major milestone not only for Intel, but for the entire accelerated computing software industry.

IWOCL 2024 will be hosting a SYCL Hacking Festival on Monday April 8 for teams interested in spending a full day hacking with SYCL, the cross-platform abstraction layer that enables code for heterogeneous and offload processors to be written using modern ISO C++. The HackFest provides a great opportunity for the community to get together and hack on projects that they feel they could progress within a single day. By hacking together teams will have the opportunity to discuss general code development, specific project issues, optimisation techniques, the latest SYCL features, and more. All participants will be supported by mentors that are world-leaders in SYCL development.

Biagio Cosenza from the University of Salerno / CINECA Supercomputing Center pens this blog on the SYnergy research project that enables efficient C++ based heterogeneous parallel programming with the Khronos SYCL API. The blog describes the SYCL-based SYnergy API’s ability to map high-level energy semantics to vendor APIs, and by running it on a large-scale compute cluster, it shows scalable energy savings up to 64 GPUs.

The Khronos Group announces the launch of the SYCL 2020 Adopters Program to enable implementations to become officially conformant to the latest-generation SYCL specification—and to ensure that SYCL is consistently implemented by multiple vendors to create a reliable platform for developers. SYCL 2020 is the open industry standard for single source C++-based heterogeneous parallel programming. The SYCL Working Group has released an updated Conformance Test Suite (CTS) as part of the SYCL 2020 Adopters Program, making it available as an open source GitHub project to enable broad industry participation in implementation testing and ongoing test suite improvements.