NVIDIA Introduces Tesla K80 With Two GK210 GPUs - World's Fastest Accelerator With 2.9 TFlops of Compute

In a conference regarding supercomputers, NVIDIA introduced their high-performance Tesla K80 accelerator featuring two Kepler GK210 GPU cores. Nosotros final saw Tesla K80 (Previously codenamed Atlas) three months ago where it was denoted past the GK210-DUO proper name and today, we tin can finally ostend the specifications of NVIDIA's latest Telsa accelerator with a double precision compute performance of 2.nine TFlops.

NVIDIA Intros GK210-DUO Powered Tesla K80 Computing Accelerator

Since the Tesla K80 is a dual-chip graphics card, it will feature two GK210 chips which are codenamed as the "GK210-DUO". The specifications will include 2 GK210 chips featuring 4992 CUDA cores (2496 CUDA Cores each), 416 TMUs and 96 ROPs. The card features a massive 24 GB GDDR5 VRAM featured along a 384-scrap bus interface, clocked at 5.00 Gbps. Information technology pumps out 480 GB/s bandwidth and features a maximum TDP of 300W that is 75W lower than the GeForce GTX Titan Z.

The core clock is maintained around 573 MHz and the bill of fare also features a boost clock of 875 MHz. The memory is clocked at five GHz for each GPU. For cooling, the card can adopt a passive heatsink design if the thermals are under control. The large heatsink shroud comes in green and black color with a silverish lining which gives a pleasant view, although the menu would never be seen when packed within a dumbo server. Large exhaust vents push button the hot air out of the example which is supplied from the cooling in the server environment.

A slide posted in NVIDIA'southward HPC oriented Heterogeneous Computing Solutions presentation shows off the Tesla K80 "GK210-DUO" graphics card in a performance slide which provides a top bandwidth of 480 GB/s and total double precision functioning rated at ii.9 TFlops. The single precision operation for the menu is rated at over 8 TFlops which is impressive from a HPC perspective. The card will retail for a high toll of around $7000 US or even more merely for users who want to stick something close to the Tesla K80 tin just grab the GeForce GTX Titan Z for $3000 US. Even the GeForce GTX Titan Z is getting major toll cuts through arrangement builders with a few OEM builders offering a $k US price cut on the card dropping it down to $2000 US which withal is a huge price for a graphics card simply a lot cheaper than its original MSRP.

While AMD has only W9100 in the competition which features 2.62 TFlops of double precision compute which is close to NVIDIA's dual card offering, they do have plans to offering a dual chip FirePro variant in the near future which tin easily overtake their green team rival. NVIDIA Tesla K80

NVIDIA Tesla K80 Specifications:

	NVIDIA Tesla K80	NVIDIA Tesla K40	NVIDIA Tesla K20X	NVIDIA Tesla K20	NVIDIA Tesla K10
CUDA Cores	2496 ten ii	2880	2688	2496	1536 ten 2
Cadre Clock	573 MHz	745MHz	732MHz	706MHz	745MHz
Heave Clock	875 MHz	810MHz, 875MHz	N/A	North/A	North/A
Memory Clock	5GHz GDDR5	6GHz GDDR5	five.2GHz GDDR5	5.2GHz GDDR5	5GHz GDDR5
Memory Bus	384-bit ten 2	384-chip	384-bit	320-fleck	256-bit
VRAM	24GB	12GB	6GB	5GB	8GB
Single Precision	~8.5 TFLOPS	4.29 TFLOPS	3.95 TFLOPS	3.52 TFLOPS	4.58 TFLOPS
Double Precision	2.ix TFLOPs	ane.43 TFLOPS (1/3)	i.31 TFLOPS (1/3)	one.17 TFLOPS (1/3)	0.19 TFLOPS
Transistor Count	fourteen.two Billion	seven.1 Billion	vii.1 Billion	vii.1 Billion	7 Billion
TDP	300W	235W	235W	225W	225W
Cooling	Active/Passive	Agile/Passive	Passive	Agile/Passive	Active/Passive
Manufacturing Procedure	TSMC 28nm	TSMC 28nm	TSMC 28nm	TSMC 28nm	TSMC 28nm
Architecture	Kepler GK210	Kepler GK110	Kepler GK110	Kepler GK110	Kepler GK104
Launch Price	TBC	$5499	$3799	$3299	$2500