| Appliance Platforms |
| |
 |
x86 OEM Appliances |
| |
|
CDN Server Appliances |
| |
|
GPGPU Appliances |
|
|
|
| Questions? |
|
|
|
Call 1.800.800.6328
|
|
|
|
|
|
| Case Study |
McAfee Speeds to Market with AMAX OEM Appliance Solution
McAfee
As a premiere achiever in an elite group of top McAfee global supply partners, AMAX understands the rapidly changing enterprise software market environment and works very closely with the McAfee product team to constantly introduce new generations of Intel-based server appliance solutions to the marketplace. |
|
|
|
 |
|
 |
AMAX leverages open architecture and integrates best-in-class computing technologies to customize your GPGPU appliance platform based on your desired form factor, performance specifications, lifecycle and compliance requirements.
AMAX’s GPGPU OEM appliances ensure time to market acceleration, long-term component availability, maximum energy efficiency, and lower total cost of ownership, without sacrificing the overall quality and reliability.
|
| |
| Model |
# of GPUs |
GPU Cores |
Floating Point Performance (Peak) |
CPU Type |
Xr-4801Gk2
 |
4U 8x Tesla K20
|
21,504 Cores |
31.6 Teraflops (SP)
10.56 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
Xr-2401Gk2
|
2U 4x Tesla K20
|
10,752 Cores |
15.80 Teraflops (SP)
5.28 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
Xr-1301Gk2
|
1U 3x Tesla K20
|
8,064 Cores |
11.85 Teraflops (SP)
3.96 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
Xr-1201Gk2
|
1U 2x Tesla K20
|
5,376 Cores |
7.90 Teraflops (SP)
2.64 Teraflops (DP)
|
Single Intel Xeon E5-2600 series processor |
Xr-2401Gk1
|
2U 4x Tesla K10
|
12,288 Cores |
18.32 Teraflops (SP)
.76 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
Xr-1301Gk1
|
1U 3x Tesla K10
|
9,216 Cores |
13.74 Teraflops (SP)
.57 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
PSC-6p
|
4U/Tower 4x Tesla K20/K20X/K10 with Passive Heatsink
|
10,752 Cores |
15.80 Teraflops (SP)
5.28 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
PSC - 6s
|
4U/Tower 4x Tesla K20/K10 with Active Heatsink
|
9,984 Cores |
14.08 Teraflops (SP)
4.68 Teraflops (DP)
|
Dual Intel Xeon E5-2600 series processors |
|
AMAX’s GPGPU appliances are powered by the NVIDIA® Tesla™ GPU computing platforms, based on NVIDIA’s Kepler GPU Computing Accelerators, the world’s fastest and most efficient high performance computing (HPC) companion processor, delivering 3 times higher performance per watt than the previous “Fermi” compute architecture, enabling hybrid computing dramatically easier, and applicable to a broader set of computing applications. AMAX’s Tesla-based GPU solutions deliver the best performance and power efficiency for applications across many industries.
AMAX’s Tesla-based GPGPU appliances are based on two different Nvidia Kepler GPUs:
Tesla K10 GPU Computing Accelerator – Optimized for single precision applications, delivering up to 2x the performance for single precision applications compared to the previous generation Fermi-based Tesla M2090 in the same power envelope.
Applications: Seismic, signal, image processing, and video analytics
|
Tesla K20 GPU Computing Accelerator – Designed for double precision applications and the broader supercomputing market, delivering 3x the double precision performance compared to the previous generation Fermi-based Tesla M2090, in the same power envelope.
Applications: Climate and weather modeling, CFD, CAE, computational physics, biochemistry simulations, data analytics, satellite imaging, and computational finance
|
KEY FEATURES:
- Powered by NVIDIA’s Kepler GPU Computing Accelerators, the world’s fastest and most efficient high performance computing (HPC) companion processor.
- Delivers 3 times higher performance per watt than the previous “Fermi” compute architecture, enabling hybrid computing dramatically easier, and applicable to a broader set of computing applications.
- Supports SMX (streaming multiprocessor) design, delivers up to 3x more performance per watt compared to the SM in Fermi, delivering 1 petaflop of computing in just 10 server racks.
- Supports Dynamic Parallelism capability, enabling GPU threads to automatically spawn new threads, greatly simplifying parallel programming and enables GPU acceleration of a broader set of popular algorithms, like adaptive mesh refinement (AMR), fast multipole method (FMM), and multigrid methods.
- Supports Hyper-Q, enabling multiple CPU cores to simultaneously utilize the CUDA cores on a single Kepler GPU, dramatically increasing GPU utilization, slashing CPU idle times, and advancing programmability.
- Supports GPUDirect with RDMA, eliminating CPU bandwidth and latency bottlenecks using direct memory access (DMA) between GPUs and other PCIe devices, resulting in significantly improved MPISendRecv efficiency between GPUs and other nodes.
|
|
