SC11 in Seattle is just around the corner and with it comes news of exciting GPU and processor innovations. Microway is exhibiting at Booth 2606:
Tesla™ C2075/M2075 Hardware Now Shipping
NVIDIA has tweaked the Tesla C2070 GPUs with a few notable improvements in the new Tesla C2075:
- Reduced TDP of 225W for superior system cooling and quieter operation
- Better ECC reporting: visibility into ECC register files and GPU cache
- Dynamic power scaling in some platforms
- Tesla M2075 and M2090 for 1U GPU Servers are also available
Tesla C2075 is now replacing Tesla C2070 in all new orders. NVIDIA has also announced EOL on the 3GB Tesla C2050 and Tesla S2050 products
Sign Up for Advances in GPU Computing
Northeastern University, October 27, 2011
Learn more about the coding and speedups of GPU computing on all platforms. Microway is a co-sponsor of this event. Topics include:
- Coding for many architectures: high performance GPU, heterogeneous, and embedded
- Developing Portable, High Performance Kernels with Microway's OpenCL Tools and Services - A presentation by Microway GPGPU Unit Manager Michael Fried
- Portable OpenCL coding for many architectures and optimization tips
- Unique hardware and upcoming software tools to ease GPU computing implementation
An
agenda for the event can be found here. For more information, please visit the
event website
Microway's MD SimCluster is Available for Tesla M2090-based Cluster Benchmarking
- Preinstalled with AMBER and NAMD, so you may conveniently upload your model and run. Custom CUDA code is welcome with advanced notice
- 4 1U GPU Compute Nodes: each with 2 Tesla M2090 GPUs- 512 core, 6GB GDDR5 ECC
- 5.32 TFLOPs double precision GPU performance from 4096 CUDA cores (10.64 TFLOPs single)
- 12 Xeon X5675 CPU cores and 48GB per node, QDR InfiniBand switch
For more information and to register for a benchmarking slot, please click here
Accelerate Your Linear Algebra Codes with WhisperStation Plus Free CULA Promo Running Until December 31
Accelerate your scientific applications from 2-10x with CULA Premium R12. CULA's linear algebra GPU library runs on WhisperStation- Tesla and requires no CUDA expertise to implement. The following promo runs until December 31, 2011
CULA Premium accelerates all types of codes leveraging linear algebra. LAPACK, Factorization, Eigenvalues and more functions are included. For more information on CULA, please click here

Microway HPC Tech Tip
Managing a Linux Software RAID with MDADM
Abstract: There are several advantages to assembling hard drives into a RAID: performance, redundancy and capacity. Microway workstations and servers are most commonly outfitted with software RAID. This will prevent a single drive failure from destroying your operating system installation.
Linux provides a robust software RAID implementation which costs nothing and offers great performance for lower array levels (e.g, 0, 1, 10). It is flexible and powerful, but array monitoring and management can be opaque if you’ve not previously worked with a Linux software RAID.
We explore the most common scenarios of checking array health and replacing a failed hard drive.
Read the full article
In our next issue: Monitoring Hard Drive and RAID Health
Visit Microway at SC11 in Seattle, Washington. Booth 2606 November 14-17, 2011. |
 |
 |
Eliot Eshelman at 508-732-5534 Brett Newman at 508-732-5542 sales@microway.com |