As part of the ongoing work to ensure glTF meets the needs of the developer community the Khronos 3D Formats working group is working on a new glTF compression extension. The goal is to greatly improve transmission efficiency of texture assets while providing efficient, cross-platform transcoding into a wide range of GPU hardware-accelerated texture formats. There are many ways to get involved in helping glTF evolve to meet your needs. If your company wants a seat at the 3D Formats Working group, you are welcome to join Khronos. Also, anyone is welcome to follow, and contribute to, technical discussions on the public glTF GitHub repo. Learn more about this call for participation.
Codeplay has a very good write-up today on machine alternatives that don't use Neural Networks. The included code, SYCL-ML was developed as a proof of concept to show what a machine learning application using heterogeneous computing can look like and has been published as an open source project. The project was developed using SYCL and ComputeCpp, which is an implementation of SYCL developed by Codeplay.
Remograph, providers of products and services for the computer graphics, visual simulation and 3D modeling markets, announced the release of Remo 3D v2.7. Remo 3D is an effective OpenGL-based tool for creating and modifying 3D models intended for realtime visualization. The primary file format is OpenFlight. Remo 3D is currently available for Microsoft Windows 10/8/7 and Linux. This new version 2.7 of Remo 3D brings brings a greatly improved tool for modifying a texture UV mapping, support for new OpenSceneGraph 3.5.9 and other various fixes. The full list of new features and improvements can be found in the release notes on our website.
The Khronos Group announces the release of the Neural Network Exchange Format (NNEF™) 1.0 Provisional Specification for universal exchange of trained neural networks between training frameworks and inference engines. NNEF reduces machine learning deployment fragmentation by enabling a rich mix of neural network training tools and inference engines to be used by applications across a diverse range of devices and platforms. The release of NNEF 1.0 as a provisional specification enables feedback from the industry to be incorporated before the specification is finalized — comments and feedback are welcome on the NNEF GitHub repository. The goal of NNEF is to enable data scientists and engineers to easily transfer trained networks from their chosen training framework into a wide variety of inference engines. A stable, flexible and extensible standard that equipment manufacturers can rely on is critical for the widespread deployment of neural networks onto edge devices, and so NNEF encapsulates a complete description of the structure, operations and parameters of a trained neural network, independent of the training tools used to produce it and the inference engine used to execute it. Learn more about NNEF 1.0 Provisional Specification in the press release, or on the NNEF homepage.
Portable Computing Language (pocl) 1.0 has been released. One of the bigger highlights of this release is that most of the OpenCL 1.2 standard conformance tests pass with the CPU backend. There are some caveats though to this listed in the documentation. Pocl is a portable open source (MIT-licensed) implementation of the OpenCL standard (1.2 with some 2.0 features supported).
The 3DMark benchmark app for Android smartphones and tablets now includes a Vulkan benchmark test. Games typically make thousands of draw calls per frame, but each one creates performance-limiting overhead for the CPU. Vulkan reduces that overhead, which means games and apps can draw more objects, textures, and effects to the screen to create richer, more immersive experiences. You can now compare scores from Android devices using OpenGL ES 3.1 and Vulkan with scores from Apple iOS devices using Metal.
LunarG is pleased to announce the release of a new option in spirv-opt, which was designed to reduce SPIR-V size. This new option, -Os, allows developers to reduce SPIR-V size without specifying individual passes of spirv-opt. Learn more about this option at LunarG. For those developers who wish to have finer control over which passes are invoked, read more details about the available SPIR-V optimization options in the white paper by Greg Fischer of LunarG, titled SPIR-V Shader Size Reduction Using spirv-opt_v1.1.
Geeks 3D has a great overview of the new OpenGL and Vulkan features in the AMD Adrenalin 17.12.1 release. AMD has added some OpenGL 4.6 extensions (GL_ARB_gl_spirv) but not all, so Adrenalin 17.12.1 is still an OpenGL 4.5 driver with OpenGL 4.6 features. This driver exposes 319 OpenGL extensions (GL=294 and WGL=25) for a Radeon RX 470 on Windows 10 64-bit. Adrenalin 17.12.1 also exposes Vulkan 1.0.65 which is one of the latest specifications. Vulkan support has been added to Radeon Overlay, Radeon Relive, Enhanced Sync and Frame rate target control.
Neil Trevett, Khronos Group President and Radhakrishna Giduthuri, Software Architecture and Compute Performance Acceleration at AMD, spoke at two Khronos related events this past week. Neils presented was an update on the Khronos Standards for Vision and Machine Learning which covered Khronos Standards OpenVX, NNEF, OpenCL, SYCL and Vulkan. Radhakrishna presented Standards for Neural Networks Acceleration and Deployment covered Khronos Standards OpenVX and NNEF. The slides from both presentations are now online.
There is a wide range of open-source deep learning training networks available today offering researchers and designers plenty of choice when they are setting up their project. Caffe, Tensorflow, Chainer, Theano, Caffe2, the list goes on and is getting longer all the time. This diversity is great for encouraging innovation, as the different approaches taken by the various frameworks make it possible to access a very wide range of capabilities, and, of course, to add functionality that’s then given back to the community. This helps to drive the virtuous cycle of innovation.
Earlier this year, Microsoft acquired Simplygon, a leader in 3D model optimization based in southern Sweden. Today,Microsoft announces the launch of Simplygon Cloud on Azure Marketplace. Simplygon reduces complexity in the creation and extensibility of 3D models through optimization. Simplygon supports glTF, FBX and OBJ file types for ingestion; rendering engines including Unity 3D and Unreal Engine; and all major mixed reality platforms, including Windows Mixed Reality, iOS and Android.
Join Patrick Cozzi and his Penn graphics students final project presentations in a live stream December 11th at 5pm EST. There will be 19 4-minute presentations in total. Topics include rendering clouds, terrain, ocean, forests, weather, VR, voxels, ray marching, and more. A complete list of projects can be found online here.
The Khronos Group announces the ratification and public release of the finalized SYCL 1.2.1 specification. SYCL for OpenCL enables code for heterogeneous processors to be written in a “single-source” style using completely standard modern C++. The multi-vendor SYCL 1.2.1 standard is available royalty-free for industry use, and the full specification together with details about the SYCL open-sourced conformance test suite and Adopters Program are online.
The first two Khronos meet-ups coming up will both be in London. First up is the Khronos UK Game Dev Social on December 7th: "Meet developers working on Vulkan within Arm, Google, Imagination, Samsung and many more! Share your experiences with others in the graphics & game tech industries." The second is a Khronos London Chapter meet-up on December 12th: Come and hear speakers from Away3D, Arm, KDB, Unity and Intel talking about Vulkan, SPIR-V, OpenCL and of course WebGL. Back in the US on December 13th is the Khronos Boston Chapter meet-up covering the latet on Khronos APIs. Finally over in Sydney Australia will be the Khronos Sydney Chapter meet-up "Graphics XMAS Meeting in Pub (Art + Science + Design + Engineering)." All events require registration.