1. Computers
  2. Display Drivers
  3. Graphics Cards
  4. Memory
  5. Motherboards
  6. Processors
  7. Software
  8. Storage
  9. Operating Systems


Facebook RSS Twitter Twitter Google Plus


Phoronix Test Suite

OpenBenchmarking.org

Portable OpenCL 0.7 Improves On OpenCL 1.2

Compiler

Published on 09 January 2013 12:02 PM EST
Written by Michael Larabel in Compiler
1 Comment

Version 0.7 of POCL, the Portable OpenCL implementation targeting OpenCL 1.2 compliance, has been officially released.

Portable OpenCL aims to be open-source, very portable, and improving performance through compiler optimizations and reducing target-dependent manual optimizations. Portable OpenCL was released in 2011 and released last August was Portable OpenCL 0.6 that began to implement the OpenCL 1.2 specification. POCL is built around the LLVM compiler infrastructure.

Portable OpenCL 0.7 introduces support for LLVM 3.2 (the latest LLVM release from last month), support for generating the work group functions using simple/parallel loop structures, fixes for POCL on PowerPC32/PowerPC64/ARMv7, and initial Cell SPU support. The Cell SPU back-end is still very experimental and meant as an example of a heterogeneous POCL device driver, though with LLVM 3.2 the Cell back-end was dropped.

In terms of the OpenCL 1.2 support, Portable OpenCL 0.7 doesn't yet implement the full specification and there are known bugs. However, POCL 0.7 is ready for wider-scale testing and is passing OpenCL tests from ViennaCL, Rodinia, Parboil, and the OpenCL Programming Guide samples as well as those from the AMD APP SDK.

The Portable OpenCL 0.7 release announcement can be found on the LLVM mailing list. The POCL project is hosted on SourceForge.

Interestingly, the development of the Portable OpenCL 0.7 release was sponsored by Nokia, namely the Radio Implementation Research Team from Nokia Research Center.

About The Author
Michael Larabel is the principal author of Phoronix.com and founded the web-site in 2004 with a focus on enriching the Linux hardware experience and being the largest web-site devoted to Linux hardware reviews, particularly for products relevant to Linux gamers and enthusiasts but also commonly reviewing servers/workstations and embedded Linux devices. Michael has written more than 10,000 articles covering the state of Linux hardware support, Linux performance, graphics hardware drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated testing software. He can be followed via and or contacted via .
Latest Linux Hardware Reviews
  1. Even With Re-Clocking, Nouveau Remains Behind NVIDIA's Proprietary Linux Driver
  2. The Power Consumption & Efficiency Of Open-Source GPU Drivers
  3. AMD R600g/RadeonSI Performance On Linux 3.16 With Mesa 10.3-devel
  4. Intel Pentium G3258 On Linux
Latest Linux Articles
  1. Nouveau vs. Radeon vs. Intel Tests On Linux 3.16, Mesa 10.3-devel
  2. KVM Benchmarks On Ubuntu 14.10
  3. X.Org Server 1.16 Officially Released With Terrific Features
  4. Ubuntu With Linux 3.16 Smashes OS X 10.9.4 On The MacBook Air
Latest Linux News
  1. GCC As A Just-In Time Compiler Is An Interesting Project
  2. Age Of Wonders III Is Still Being Ported To Linux
  3. Git 2.1 To Further Mainline Windows Support Patches
  4. Debian 8.0 Jessie Is Settling For Linux 3.16
  5. Meson: A Next-Gen Build System Showing Promise
  6. Linux 3.16-rc7 Calms Things Down For The Linux 3.16 Kernel
  7. Open-Source AMD Users Report Hawaii GPU Acceleration Is Working
  8. Linus Torvalds On GCC 4.9: Pure & Utter Crap
  9. Cauldron 2014: GCC & LLVM Will Look To Collaborate More
  10. GCC Receives ACM Programming Languages Software Award
Latest Forum Discussions
  1. Porting Mesa to the Playstation 2
  2. Updated and Optimized Ubuntu Free Graphics Drivers
  3. ASRock AM1H-ITX: One Of The Best AM1 Mini-ITX Motherboards
  4. Debian + radeonsi
  5. Open-source drivers on ATI R7 260X
  6. Table test
  7. How To Setup Radeon DPM On Ubuntu Linux
  8. New build, first Linux PC, what could go wrong? ;)