Nvidia Announces PCI-E version of Tesla P100 to

Jun

20

2016

Nvidia PCI-E versions of the Tesla announced P100 accelerator. Nvidia unveiled the hpc card with Pascal GPU in April, but that version had a mezzanine connector developed by the manufacturer itself.

Nvidia will end this year, two versions of the PCI-E version of the P100 available : one with 16GB hbm2 and one with 12GB. The cards do not have the nvlink interconnect: for that interface to P100 cards can communicate Nvidia has developed its own mezzanine connector.

The cards have slightly lower boost clock speeds than the Mezzanine version and also the lower TDP: 250W instead of 300W. The 12GB version of the PCI-E-P100 also features a 3072-bit wide memory bus, opposite the 4096bit interface of the other two cards. Like the P100 card that Nvidia in April announced , these are accelerators for high performance computing as supercomputers with Pascal GPU and high bandwidth memory of the second generation.

Nvidia announced the P100 cards during the International Supercomputing Conference in Frankfurt.

Tesla M40 but Nvidia & # 039; s site of the P100 so same form factor?

Nvidia P100 PCI-E
Nvidia P100 PCI-E
Nvidia P100 PCI-E

Nvidia Tesla family
Tesla P100
(Mezzanine) Tesla P100
(16GB) Tesla P100
(12GB) Tesla M40
stream Processors 3584 3584 3584 3072
Core Clock function. 1328MHz ? ? 948MHz
Boost kloksn. 1480MHz 1300MHz 1300MHz 1114MHz
Kloksn memory. 1,4Gbit / s HBM2 1,4Gbit / s HBM2 1,4Gbit / s HBM2 6Gbps / s GDDR5
Memory 4096-bit 4096-bit 3072-bit 384-bit
memory Bandwidth 720GB / sec 720GB / sec 540GB / sec 288GB / sec
memory Amount 16GB 16GB 12GB 12GB
half Precision 21.2 TFLOPS 18.7 TFLOPS 18.7 TFLOPS 6.8 TFLOPS
single Precision 10.6 TFLOPS 9.3 TFLOPS 9.3 TFLOPS 6.8 TFLOPS
Double Precision 5.3 TFLOPS
(1/2 rate) 4.7 TFLOPS
(1/2 rate) 4.7 TFLOPS
(1/2 rate) 213 GFLOPS
(1/32 rate)
gpu GP100
(610mm2) GP100
(610mm2) GP100
(610mm2) GM200
transistors 15.3 billion 15.3 billion 15.3 billion 8 billion
TDP 300W 250W 250W 250W
Form Factor mezzanine PCI-E PCI-E PCI-E
Cooling AFTER passive passive passive
Process TSMC 16nm FinFET TSMC 16nm FinFET TSMC 16nm FinFET TSMC 28nm
Architecture Pascal Pascal Pascal Maxwell 2

Viewing:-78

In: Technology & Gadgets Asked By: [15187 Red Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »