AMAX is a leading HPC solution provider specializing in GPU cluster solutions. AMAX's GPU cluster building block solutions are powered by the NVIDIA® Tesla™ 20-series GPU computing platforms, delivering up to 57,344 GPU cores, 131.84 Teraflops of single precision performance and 65.92 Teraflops of double precision performance per 42U cluster. They support advanced features for the technical and enterprise computing space, including ECC memory for uncompromised accuracy and scalability.
 |
As an NVIDIA® Tesla™ Preferred Provider, AMAX offers our customers the opportunity to try Tesla GPU Computing Processors prior to placing an order. Under this program, qualified customers may remotely access our in-house Tesla Personal Supercomputer or GPU cluster and experience first-hand the revolutionary computing power enabled by the NVIDIA CUDA™ parallel computing architecture.
|

For more information, visit AMAX's Tesla GPU Computing Solutions
|
Supercomputing Performance
Experience cluster level computing performance at 1/20th the power consumption and 1/10th the cost of traditional x86-based clusters.
AMAX's Tesla™ high performance computing (HPC) solutions enable the necessary transition to energy efficient parallel computing power. These solutions scale to solve the world's most important computing challenges-more quickly and accurately. They are designed with exclusive features to maximize performance for computing professionals.
- Full double precision floating point performance
- Faster PCIe communication
- Higher performance on technical applications with large data sets
- Faster communication with InfiniBand using NVIDIA GPUDirect™
- Higher performance CUDA driver for Windows OS
Accessible
CUDA programming environment for Windows or Linux, with broad support of programming languages and APIs, including C, C++, OpenCL, DirectCompute, or Fortran. The NVIDIA® CUDA™ parallel computing architecture enables developers to utilize C with NVIDIA GPUs to create computationally-intensive applications. CUDA is easy to learn and has been widely adopted by thousands of application developers worldwide.
|