Jump to content

AMD announces GPUOpen initiative


DKT27

Recommended Posts

  • Administrator

AMD's Radeon Technologies Group (RTG) has announced an initiative called GPUOpen. Following up the information released last week concerning hardware plans, such as FreeSync over HDMI, we now are free to publish information concerning a software initiative where AMD will share the full source code to effects like TressFX, LiquidVR and ShadowFX plus tools like the CodeXL static analyser and libraries like LiquidVR. These GPUOpen resources will give developers unprecedented access to AMD GPUs and go live in January via GitHub. Another important announcement from AMD concerns the development of the AMDGPU unified Linux driver.

 

EbSCmQV.jpg

 

GPUOpen

 

AMD sees an 'industry problem' with divided console and PC development and so called 'black box' solutions, which probably is code-speak for Nvidia GameWorks. With AMD hardware at the core of the big two console systems and optimised for those platforms it should mean that moving games to PCs and smoothly making the most of PC-based AMD GPU features would result. AMD wants to provide open access to its GPU, many of its graphics API/SDK tools and follow industry standards to reduce friction in development across consoles and PCs.

 

UBmERa2.jpg

 

As AnandTech notes, "everything at the GPUOpen portal will be open source and licensed under the highly permissive MIT license". Under this license developers can utilise AMD's code from its tools and libraries in any open and closed source projects they like. This contrasts to Nvidia's closed source libraries and limited sharing.

 

AMDGPU Linux Driver

 

AMDGPU was first seen back in April. A few months later we don't have a lot more information but we know that AMD wants to make this a unified driver for consumers and professionals. The new RTG produced driver will be made available as both a fully open source and a mixed open/closed source driver to the distros that require it. Both drivers will have a core AMDGPU source but the proprietary one will include "user mode driver components such as their (typically superior) OpenGL and multimedia runtimes," reports AnandTech. The single code base will allow AMD to focus its Linux development resources onto a narrower point.

 

f76p1Y0.jpg

 

AMDGPU is an important part of the Boltzmann Initiative, announced about a month ago. This is a software initiative aimed at High Performance computing (HPC) and provides provides a Heterogeneous Compute Compiler (HCC) using C++, a headless Linux driver and HSA runtime, and the Heterogeneous-compute Interface for Portability (HIP) to port CUDA apps to C++. With AMDGPU's open nature HPC users and developers will have greater access to the code behind the driver than offered by rivals such as Nvidia.

 

We shall have to see what tangible benefits result from RTG's latest open software initiatives. Will the open source code access improve the adoption of and boost the benefits of AMD's previously proprietary tech?

 

:view:View: Original Article

Link to comment
Share on other sites


  • Replies 5
  • Views 988
  • Created
  • Last Reply
Last week, we posted a story about the AMD Radeon Technology Group’s plans to support FreeSync over HDMI and other upcoming display technologies like HDR and DisplayPort 1.3. In that story, we mentioned that AMD and the new RTG (Radeon Technologies Group) organization would have more news as the year drew to a close. Today, we can talk about their next major initiative in dirving GPU resources across a wider swath of new applications, known as "GPUOpen."
 
In a nutshell, AMD is releasing a slew of open-source software and tools to give developers of games, heterogeneous applications, and HPC applications deeper access to the GPU and GPU resources.
 
Let’s discuss gaming first. AMD and the RTG are looking for ways to ease game development, so games for PCs and consoles can be produced in a similar manner. To quote AMD, “As a continuation of the strategy we started with Mantle, we are giving even more control of the GPU to developers. As console developers have benefited from low-level access to the GPU, AMD wants to continue to bring this level of access to the PC space.” The AMD GPUOpen initiative is meant to give developers the ability to more easily leverage assets they've already made for console development. With GPUOpen, game developers will have direct access to GPU hardware, as well as access to a large collection of open source effects, tools, libraries and SDKs, which are being made available on GitHub under an MIT open-source license, which essentially gives users unrestricted access to the software.
 
AMD hopes that GPUOpen will enable console-style development for PC games through this open source software initiative. But the company doesn’t stop there. GPUOpen also includes an end-to-end open source compute infrastructure for cluster-based computing and a new Linux software and driver strategy.
 
A few weeks back, AMD announced its Boltzmann Initiative and an HSA (Heterogeneous System Architecture) software suite that includes an HCC compiler for C++ development, which will expand the field of programmers who can leverage HSA. One of the primary goals of Heterogeneous Systems Architecture (HSA) is to help ease the development of parallel applications through the use of higher level languages, like C++. The new HCC C++ compiler is a key tool in enabling developers to more easily leverage the resources of discrete GPU hardware in heterogeneous systems.
 
The HCC complier that’s part of the Boltzmann Initiative also allows developers to convert CUDA code to portable C++. According to AMD, internal testing shows that in many cases 90 percent or more of CUDA code can be automatically converted into C++ with the final 10 percent converted manually in the widely popular C++ language. An early access program for the “Boltzmann Initiative” tools is planned for Q1 2016.
 
AMD GPUOpen also introduces a new Linux driver model and runtime targeted at HPC Cluster-Class Computing. The new headless Linux driver addresses core high-performance computing needs, including low latency compute dispatch and PCI Express data transfers, peer-to-peer GPU support, Remote Direct Memory Access (RDMA) from InfiniBand that interconnects directly to GPU memory and Large Single Memory Allocation support.
 
It will take some time before the GPUOpen initiative bears fruit, but these moves appear to be heading in the right direction for AMD. Leveraging the open-source community to not only further software development in general, and give developers closer-to-the-metal access to GPU resources, could result in a larger pool of applications optimized to run on Radeon hardware. 
 
 
Link to comment
Share on other sites


  • 2 weeks later...

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...