PathScale – a well-known provider of one of the industry’s highest-performance 64-bit C, C++, and Fortran compilers – has added support for ARM and C++11 in their EKOPath 5 compiler suite. “We are pleased to provide the ARM HPC ecosystem with our new EKOPath compiler,” said Christopher Bergstrom, the PathScale Chief Technology Officer. “Our C/Fortran compiler and BLAS libraries optimized for the ARMv8-A architecture will help increase the pace of innovation for ARM partners working in HPC applications.” A free 15-day trial is available to qualified new customers (inquire here). Existing customers with a support entitlement can upgrade to the EKOPath 5 compiler suite. PathScale products are also available through Cray and other leading Linux cluster solution providers in North America, Europe and Asia.
Lakshmi Mandyam, the director of server systems and ecosystems at ARM said that , “[t]he availability of PathScale’s EKOPath Compiler underscores the continued growth of ARM’s HPC ecosystem and the demand for ARM-based servers from within the HPC community. This announcement enables our partners to accelerate the deployment of the ARM-powered solutions in the HPC world.” (Source ARM.com)
Performance-oriented programmers have to love the PathScale commitment to extreme performance, “Our philosophy on performance is simple, but serious. The best benchmark is your code. If your application isn’t the fastest when compiling with the EKOPath compiler it’s a bug and we want to know about it.”
PathScale is also beta testing a complete GPGPU and multi-core solution with ENZO, a product intended to quickly transform any existing C, C++ or Fortran codebase into a highly efficient parallel code for GPU or multi-core systems. Email firstname.lastname@example.org for more information about gaining early beta access. Note that PathScale is now part of the OpenACC standards committee and that the EKOpath compiler suite is OpenMP 2.5 compliant along with being compatible with the AMD Core Math Library.
ARM servers along with Intel Xeon Phi have been generating excitement in the HPC world due to their power efficiency. Following is the abstract from one paper, “Heterogeneous High Throughput Scientific Computing with APM X-Gene and Intel Xeon Phi“, that provides benchmark results:
Electrical power requirements will be a constraint on the future growth of Distributed High Throughput Computing (DHTC) as used by High Energy Physics. Performance-per-watt is a critical metric for the evaluation of computer architectures for cost efficient computing. Additionally, future performance growth will come from heterogeneous, many-core, and high computing density platforms with specialized processors. In this paper we examine the Intel Xeon Phi Many Integrated Cores (MIC) co-processor and Applied Micro X-Gene ARMv8 64-bit low-power server system-on-a-chip (SoC) solutions for scientific computing applications. We report our experience on software porting, performance and energy efficiency and evaluate the potential for use of such technologies in the context of distributed computing systems such as the Worldwide LHC Computing Grid (WLCG). (Source arxiv.org)
What’s new in the latest EKOPath 5 release?
- Significant improvements in performance and robustness
- Support for latest Intel® 64 & AMD64 processors
- Support for SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4A, AVX, FMA3, FMA4, XOP & AES
- Support for AMD Bulldozer, AMD Piledriver, latest Intel processors & more
- More Fortran 2003
- New C++ front-end based on clang
- New C++11 lambdas
- New C++11 Rvalue references
- New C++11 Variadic templates
- New C++11 auto-typed variables
- New C++11 Null pointer constant
- (almost all core C++11 and more)
- Optimzed AMD memcpy for newer processors
- Single file IPA (inter-procedural analysis)
- Improved inlining support
- Hundreds of bug fixes for C, C++ and Fortran