Technologies / NVIDIA Tesla GPU Solutions / NVIDIA Tesla GPU Servers

Filter By

Rack Height

1U (15)
2U (11)
3U (2)
4U (16)

Processor Manufacturer

Intel (39)
AMD (4)
IBM (1)

Number of Processors Supported

1 (5)
2 (37)
4 (2)

Processor Socket

LGA 2011-3 (21)
LGA 3647 (16)
SP3 (3)
LGA 2011 (2)
POWER8 (1)
G34 (1)

Processor Supported

E5-2600 v4 (21)
E5-2600 v3 (21)
Silver 41XX (16)
Platinum 81XX (16)
Gold 61XX (16)
Gold 51XX (16)
Bronze 31XX (16)
Platinum 81XXF (3)
7000-Series (3)
POWER8 with NVLink (1)

Maximum Memory

1 TB (18)
1.5 TB (13)
768 GB (4)
512 GB (3)
2 TB (3)
384 GB (2)
738 GB (1)

Number of Double-Wide Card Slots

4 (14)
8 (8)
2 (6)
10 (3)
3 (3)
6 (2)
1 (2)
5 (1)

Storage Drive Form Factor

2.5" (33)
M.2 (11)
3.5" (11)
U.2 (1)

Number of Drives Supported

2 (3)
3 (2)
4 (8)
5 (1)
6 (7)
8 (4)
9 (2)
10 (6)
12 (1)
14 (1)

Network Technology

10BASE-T (42)
100BASE-TX (42)
1000BASE-T (42)
10GBASE-T (17)

Number of Network (RJ-45) Ports

4 (3)
2 (37)

Maximum Power Supply Wattage

800 W (1)
1000 W (20)
1000W (1)
1100 W (3)
1200 W (11)
1400 W (7)
1600 W (8)
1620 W (3)
1800 W (16)
1800W (1)

Green Compliance Certificate/Authority

80 Plus Platinum (25)
80 Plus Titanium (17)
80 Plus Gold (1)

NVIDIA Tesla GPU Servers

NVIDIA® Tesla® GPU Servers

Exxact GPU Servers utilizing the NVIDIA Kepler architecture Tesla K80 GPU Computing Accelerators represent the bleeding edge of supercomputing offering unmatched performance providing up to 14.96 Teraflops double precision and 44.80 Teraflops single precision performance. In comparison to servers equipped with only Sandy Bridge CPUs, Exxact servers, with NVIDIA Kepler GPUs, generally deliver performance gains greater than 10x.

NVIDIA CUDA parallel programming model has changed the high performance computing (HPC) landscape as an extension to industry standard languages like C, C++, and Fortran. When Tesla K80 GPU Accelerators are added to servers with Haswell CPUs, CUDA-enabled applications are typically accelerated up to 10x. The K20X benchmark below shows single node performance of leading applications in various science domains.