Microway is celebrating its 30th year serving the HPC industry. Meet with us at SuperComputing 2011 in Seattle (November 14 - 17) at Booth 2606:
Microway SC11 Booth Highlights
Visit our Booth 2606 to see the newest processor architectures from AMD and Intel, as well as the latest GPUs from NVIDIA and AMD. We are showcasing award-winning, flexible, high-density configurations, as well as our quiet WhisperStations that HPC users love to keep in their offices.
- NumberSmasher QuadNode with Future Xeon E5 CPUs and InfiniBand running MATLAB
- WhisperStation - Tesla PSC running GPU Accelerated VMD paired with active-shutter 3D technology
- The latest Microway Navion platforms with new Opteron 6200 Series CPUs with 16 cores and the Bulldozer core architecture.
- A Laser Wakefield Acceleration simulation from Tech-X, VORPAL
- Navion 4P Servers with 64 Cores and 2 GPUs in 2U
- WhisperStation with SimpleCL and CLIDE for OpenCL kernel development
Are you an OpenCL developer or considering OpenCL development? Stop by to learn more about Microway's OpenCL Toolkit - simplifying OpenCL adoption and benchmarking many hardware architectures supported by OpenCL.
CUDA GPU Programming Webinars and New, Easier to Use CUDA Developer Zone
For those programming in CUDA on NVIDIA Tesla GPUs, you may wish to view one of the many CUDA GPU Programming webinars. The CUDA webinars provide detailed coding tips, deep-down insight into advanced Tesla features, and some application specific information. They are great resources to beginning and advanced CUDA programmers. Microway's tech team suggests the following:
- Unified Virtual Addressing and P2P Copy
CUDA 4.0 allows users to set up a unified virtual memory address space. This allows users to reduce the pains of programming across GPUs and moving data to/from system memory. Send the program to a memory address, and CUDA automatically uses a pointer to refer to the proper location. P2P copy allows a direct transfer between GPUs without copying data back to system memory.
- CUDA Libraries
There are numerous pre-ported CUDA accelerated libraries available. These can be implemented with fairly little coding effort, and they often have dramatic effects on performance.
- GPUDirect (in cooperation with Mellanox)
The GPUDirect feature allows for faster data transfers between InfiniBand HCAs and Tesla GPUs. It removes the latency of an additional memory copy when transferring data off IB and to the GPU.
Partner SC11 Booth Highlights
This year's SC11 schedule is busier than ever. Our partners have some great product announcements and demonstrations. Learn more at their events and then have Microway design a configuration that brings them all together for you:
- Accelereyes (Booth 143)
Learn about easy acceleration of MATLAB using CUDA and OpenCL compute devices.
- AMD (Booth 823)
See Microway quiet WhisperStations accelerating a number of scientific applications and compilers utilizing the new 16-core Opteron 6200-series processors. http://server.amd.com/LP=238
- Bright Computing (Booth 2734)
View a live demonstration of Bright Cluster Manager, a powerful cluster administration tool. Learn about the new cloud bursting feature for on-demand addition of compute nodes from Amazon's cloud. http://www.brightcomputing.com
- EM Photonics (Booth 244)
Learn more about CULA, an easy-to-integrate GPU-accelerated Linear Algebra library (now with Dense and Sparse packages). http://www.culatools.com/
- Intel (Booth 2121)
Join Intel's Fellow Travelers Tour for a guided tour of Intel's best partner booths on the SC11 floor. Stops include a view of the Future Xeon E5 platform running MATLAB at Microway's Booth 2606.
- Mellanox (Booth 522)
See the latest in industry-leading FDR 56Gb/s InfiniBand solutions. http://mellanox.com/sc11/
- NVIDIA (Booth 2719)
Over three days and featuring 40 presentations, don’t miss the opportunity to visit NVIDIA’s GPU Technology Theater at SC11. Here an impressive line-up of HPC luminaries and NVIDIA experts will be talking about how GPUs are accelerating codes across many scientific disciplines. They also will share helpful tips and tricks to development on GPUs. http://www.nvidia.com/object/sc11.html
New Books, Free CUDA GPU Computing Gems Book with Tesla Purchase
In celebration of our 30th anniversary, we are providing a free copy of GPU Computing Gems Emerald Edition to any Microway customer purchasing an NVIDIA Tesla GPU, WhisperStation or Tesla cluster. Take advantage of this offer on orders through March 31, 2012.
- GPU Computing Gems Emerald Edition
GPU Computing Gems: Emerald Edition is the first volume in Morgan Kaufmann's Applications of GPU Computing Series. Graphics processing units (GPUs) can do much more than render graphics. Scientists and researchers increasingly look to GPUs to improve the efficiency and performance of computationally-intensive experiments across a range of disciplines. GPU Computing Gems: Emerald Edition brings their techniques to you, showcasing GPU-based solutions From Elsevier Press
- GPU Computing Gems Jade Edition
This is the second volume of Morgan Kaufmann's GPU Computing Gems, offering an all-new set of insights, ideas, and practical "hands-on" skills from researchers and developers worldwide. Each chapter gives you a window into the work being performed across a variety of application domains, and the opportunity to witness the impact of parallel GPU computing on the efficiency of scientific research. GPU Computing Gems: Jade Edition showcases the latest research solutions with GPGPU and CUDA. From Elsevier Press
- OpenCL Programming Guide
Using the new OpenCL (Open Computing Language) standard and framework, you can write applications that access all available programming resources: CPUs, GPUs, accelerators, and even external processors. Already implemented by Apple, ATI, NVIDIA, and other leaders, OpenCL has outstanding potential for PCs, servers, handheld/embedded devices, high-performance computing, and even cloud systems.
From Safari Books. Please visit AMD in Booth 823 and be eligible for a daily drawing for this book.
- Heterogeneous Computing with OpenCL
Heterogeneous Computing with OpenCL teaches OpenCL and parallel programming for complex systems that may include a variety of device architectures: multi-core CPUs, GPUs, and fully-integrated Accelerated Processing Units (APUs) such as AMD Fusion technology. Designed to work on multiple platforms and with wide industry support, OpenCL will help you more effectively program for a heterogeneous future.
From Elsevier Press. This book is the product of Microway customers at Northeastern University.
Microway HPC Tech Tip
Monitoring Hard Drive and RAID Health
Abstract: Generally, you won’t find out that one of your hard drives has failed until the data is gone. Even if you are using a software or hardware RAID, it will only continue to function if you replace failed drives. Some RAIDs run in degraded mode for months or years until additional drive failures have ruined any chance of data recovery.
Drives and operating systems are designed to work around issues as best they can until absolute failure. However, that doesn’t mean that you can’t monitor the situation and receive an alert as soon as the first problem develops.
We explore two utilities to help manage disks and stay on top of any potential problems.
Read the full article
In our next issue: GPU Monitoring with NVIDIA SMI
|Visit the Tech Team at SC11 in Seattle, Washington. Booth 2606 November 14-17, 2011.
Eliot Eshelman at 508-732-5534
Brett Newman at 508-732-5542
To opt out of future e-mail from Microway, please click here.