Datasheet

For more information : www.pny.eu/tesla
FT77-B7015 Datasheet
4U DUAL-SOCKET SMP SERVER IMPLEMENTATION FOR GPGPU APPLICATIONS
TESLA™ MODULES - GPU SPECIFICATIONS
Dual LGA1366 sockets support up to (2) Intel® Xeon® processors
5500/5600 series (codenamed Nehalem-EP/Westmere-EP)
(4) GbE LAN
Support QPI 6.4/ 5.86/ 4.8 GT/s
(4) Internal 2.5’’ SATA HDDs w/ RAID 0, 1, 10, 5 support by Intel® Matrix
Storage Technology
Dual Intel® 5520 IOH Onboard BMC w/ IPMI and iKVM support
Intel® ICH10R South Bridge
(2+1) Redundant, ERP1U high-efciency 2,400W PSU (@220V) w/ PFC, 80-
PLUS SilverCertied
(18) 240-pin DDR3 DIMM slots support up to 144GB R-DDR3
1333/1066/800MHz w/ECC memory
PART NUMBERS:
FT77-7015 + 1 x TESLA M2075 B7015F77V4R-N1275
FT77-7015 + 2 x TESLA M2075 B7015F77V4R-N2275
FT77-7015 + 3 x TESLA M2075 B7015F77V4R-N3275
FT77-7015 + 4 x TESLA M2075 B7015F77V4R-N4275
FT77-7015 + 5 x TESLA M2075 B7015F77V4R-N5275
FT77-7015 + 6 x TESLA M2075 B7015F77V4R-N6275
FT77-7015 + 7 x TESLA M2075 B7015F77V4R-N7275
FT77-7015 + 8 x TESLA M2075 B7015F77V4R-N8275
Tesla M2090 Tesla M2075
PEAK DOUBLE PRECISION FLOATING
POINT PERFORMANCE
665 Gigaops 515 Gigaops
PEAK SINGLE PRECISION FLOATING
POINT PERFORMANCE
1331 Gigaops 1030 Gigaops
CUDA CORES
512 448
MEMORY SIZE (GDDR5)
6 GigaBytes 6 GigaBytes
MEMORY BANDWIDTH (ECC off)
177 GBytes/sec 150 GBytes/sec
4U GPGPU
SERVER PLAFORM
FT77-7015 - ONLY B7015F77V4R
FT77-7015 + 1 x TESLA M2090 B7015F77V4R-N129
FT77-7015 + 2 x TESLA M2090 B7015F77V4R-N229
FT77-7015 + 3 x TESLA M2090 B7015F77V4R-N329
FT77-7015 + 4 x TESLA M2090 B7015F77V4R-N429
FT77-7015 + 5 x TESLA M2090 B7015F77V4R-N529
FT77-7015 + 6 x TESLA M2090 B7015F77V4R-N629
FT77-7015 + 7 x TESLA M2090 B7015F77V4R-N729
FT77-7015 + 8 x TESLA M2090 B7015F77V4R-N829
M2075
CHASSIS ONLY
M2090
Rear view
The GPU accelerates applications running on the CPU by ofoading some of the compute-intensive and time consuming portions of the code.
Designed for high performance in massively parallel computing environments , the PNY product portfolio of TYAN server and NVIDIA GPUs easily
meet users’ complex workloads requirements while accelerating general computing capability.
Target Applications
- Bio-Informatics, Life Sciences and Computational Chemistry
- Electronic-Design Automation
- Computational Finance and Computational Fluid Dynamics
- Computer Vision and MATLAB Acceleration
- Data mining, Analytics and Databases
- Imaging, Molecular Dynamics, Weather, Climate Modeling and
Atmospheric
- Oil, Gas and Petroleum Exploration
- Physic and Scientic Research
ERP1U (2+1) 2,400W board
(@220V) RPSU
(4) Internal 2.5” SATA-II
HDDs
Integrated front panel
board
System FANs (5+1)
12038Rear
S7015GM4NR
(2) LGA1366 sockets
(18) R-DDR3 1333/1066/800MHz
Max.144GB
(4) GbE ports
Up to (8) Double-deck,
Full-size (Gen.2) x16
GPU/Graphic adapters