Hello guest   •   log in   •   shopping cart
Site Search:
Quote Lookup:
Quote ID
Conf #
Home > NVIDIA Tesla GPU Products


GPU (Graphics Processing Unit) computing offers unprecedented accelerated performance by offloading portions of compute-intensive applications to the GPU, while the remainder of the code runs on the CPU. By harnessing the processing power of both the CPU and GPU, applications run significantly faster.

Silicon Mechanics GPU servers and HPC (High-Performance Computing) systems are optimized with the latest GPU accelerator technology from NVIDIA, providing scalability and breakthrough power efficiency across a wide array of applications. Our configurable systems enable seamless integration of GPU computing for the world's most complex computational problems.

Now Featuring NVIDIA Tesla K40 GPU Accelerators!
Silicon Mechanics GPU servers and HPC Systems now feature Tesla K40 GPUs, the newest addition to the NVIDIA Tesla GPU family and the world's fastest accelerator!

Tesla K40 GPUs feature 12GB of memory, 2880 CUDA cores, and a new GPUBoost¹ feature. With the GPUBoost feature, users can select a sustained accelerated GPU Clock speed to convert power headroom into performance acceleration. Tesla K40 GPUs offer up to 40 percent higher performance compared to Tesla K20X GPUs and are the best solution for tackling the most demanding big data and HPC challenges.

Silicon Mechanics GPU Products









Rackform iServ R352.v4
-Up to 3x GPUs
-1U hybrid compute server
-2x Intel® Xeon® CPUs









Rackform iServ R354.v4
-Up to 4x GPUs
-2U GPU compute server
-2x Intel® Xeon® CPUs









Hyperform R2504.v4
-Up to 4x GPUs
-4U/tower personal supercomputer
-2x Intel® Xeon® CPUs









Rackform iServ R485
-Up to 4x GPUs
-5U rackmount server
-8x Intel® Xeon® CPUs









Rackform iServ R4550
-4U 4-Node High Density Server
-Up to 12x GPUs
-8x Intel Xeon CPUss
-8x 3.5-inch SAS/SATA Drives









Rackform iServ R4552
-4U 4-Node High Density Server
-Up to 12x GPUs
-8x Intel Xeon CPUs
-24x 2.5-inch SAS/SATA Drives

GPU Technology
Maximize your HPC performance with the NVIDIA Tesla family of GPUs, built on NVIDIA Kepler™ compute architecture and the CUDA parallel computing platform. This makes them ideal for delivering record acceleration and more efficient compute performance for big data applications in fields including seismic processing; computational biology and chemistry; weather and climate modeling; image, video and signal processing; computational finance, computational physics; CAE and CFD; and data analytics.

TECHNICAL SPECIFICATIONS
TESLA K40
TESLA K20X
TESLA K20
TESLA K10²
Peak double-precision floating point
performance (board)
1.43 TFLOPS
1.31 TFLOPS
1.17 TFLOPS
0.19 TFLOPS
Peak single-precision floating point
performance (board)
4.29 TFLOPS
3.95 TFLOPS
3.52 TFLOPS
4.58 TFLOPS
Number of GPUs
1 x GK110B
1 x GK110
2 x GK104
Number of CUDA cores
2,880
2,688
2,496
2 x 1,536
Memory size per board (GDDR5)
12 GB
6 GB
5 GB
8 GB
Memory bandwidth for board (ECC off)³
288 GB/s
250 GB/s
208 GB/s
208 GB/s
Architecture features
SMX, Dynamic Parallelism, Hyper-Q
SMX
System Type
Servers and Workstations
Servers
Servers and Workstations
Servers


Take a Free Test Drive with Silicon Mechanics!
Learn More

The innovative Kepler compute architecture design includes the following powerful technology features:
  • SMX (streaming multiprocessor) - Delivers up to 3x more performance per watt than the SM in previous-generation NVIDIA Fermi™ GPUs.
  • Dynamic Parallelism - Enables GPU threads to automatically pawn new threads. By adapting to the data without going back to the CPU, this feature greatly simplifies parallel programming.
  • Hyper-Q - Allows multiple CPU cores to simultaneously use the CUDA cores on a single Kepler GPU. This dramatically increases GPU utilization and slashes CPU idle times.

» More about NVIDIA Tesla GPU computing

» More about NVIDIA CUDA programming

» More about GPU Applications

1. For details on GPUBoost refer to the K40 Board spec on www.nvidia.com/object/tesla_product_literature.html | 2. Tesla K10 specifications are shown as aggregate of two GPUs. | 3. With ECC on, 6.25% of the GPU memory is used foe ECC bits. So, for example, 6GB total memory yields 5.625 GB of user available memory with ECC on.

Storage Solutions
Rackform Rackmount Servers
High-Density Servers
Storform Storage
Hyperform HPC
Racks & Accessories
Spares Kits
What's New
1U Servers
2U Servers
3U & Larger Servers
Battery Backup Modules
Warranties
Technical Support
Documentation
Discontinued Servers
Library
News & Press Releases
Events
Careers
Newsletters
Privacy Policy
About Us
Contact Us
My Account
Cart
Checkout

Copyright © 2014 Silicon Mechanics

Silicon Mechanics and the Silicon Mechanics logo are registered trademarks of Silicon Mechanics, Inc.
Intel, the Intel logo, Xeon, and Xeon Inside, are trademarks or registered trademarks of Intel Corporation in the US and other countries.
All other trademarks are the property of their respective owners.