Published in Graphics

Nvidia introduces five new Quadro cards

by on13 August 2014

Kepler and Maxwell designs

Nvidia has revamped its Quadro professional graphics line-up with a total of five new cards, two of which are based on the company’s latest Maxwell architecture.

The new cards should start shipping in September. Nvidia has not released any pricing info so far.

Maxwell goes Quadro


The first new card is the Quadro K420, a Kepler-based card with 192 CUDA  cores clocked at 780MHz. It features 1GB of DDR3 memory on a 128-bit bus, clocked at 1.8GHz. The card has a TDP of 41W and it churns out 0.3TFLOPs.

quadroK2200

The Quadro K620 is a Maxwell design. It has 384 cores clocked at 1GHz, backed by 2GB of DDR3 clocked at 1.8GHz. The TDP stands at just 45W, but the card delivers 0.8TFLOPs, proving once again that Maxwell offers vastly superior efficiency.

The Quadro K2200 is a bit more serious. This mid-range professional solution packs 640 CUDA cores running at up to 1GHz. It uses 5GHz GDDR5 memory on a 128-bit bus and there’s a lot more of it – 4GB to be precise. The TDP is 68W and the card can pump out 1.3TFLOPs (single precision).

Kepler still powers 100W+ Quadro cards


The Quadro K5200 and K4200 are Kepler cards with a beefier 256-bit memory bus. The Quadro K4200 comes equipped with 1344 CUDA cores clocked at 780MHz. It has 4GB of GDDR5 clocked at 5.4GHz effective. The TDP stands at 105W and the card 2.1TFLOPs.

quadroK5200

The K5200 packs 2304 CUDA cores and it can deliver 3.1TFLOPs. It has 8GB of GDDR5 clocked 6GHz effective. However, the GPU clock is somewhat lower at 650MHz. Its TDP is 150W.

It looks like an interesting alternative to the mighty Quadro K6000, which is a $5,000 card with 2880 CUDA cores, or a "full GK110" implementation as some buffs prefer to call it.

Of course, Nvidia is not the only player in this segment. In fact AMD has made great strides in professional graphics and it is going strong. AMD also used Siggraph to announce four professional cards

Rate this item
(0 votes)

Read more about: