KitGuru KitGuru.net – Tech News | Hardware News | Hardware Reviews | IOS | Mobile | Gaming | Graphics Cards
A web-site has published what it claims to be specifications of Nvidia’s next-generation Quadro M6000 graphics adapter as well as GM200 graphics processing unit. If the specifications are correct, then we should expect the GM200 to be a very powerful graphics solution.
VideoCardz web-site claims that it has found GPU-Z validation results of Nvidia’s next-generation high-performance professional graphics card called Quadro M6000, which is based on the GM200 graphics processing unit. The graphics card reportedly carries 12GB of GDDR5 memory, the amount that is expected to be featured on Nvidia’s forthcoming professional graphics card.
The GPU has 17F0 device id, which has never been used before. However, it is impossible to verify whether the device id is correct.
According to the publication, the GM200 has 3072 stream processors, something that has been revealed by multiple rumours. However, it is unclear whether 3072 is maximum configuration or not. If the graphics solution has 3072 SPs and Nvidia has not changed architectural configuration of Maxwell GPUs, then it also features 192 texture units, 96 raster operations pipelines and 384-bit memory bus.
Previously it was reported that graphics cards based on the GM200 will be 30 per cent faster than the GeForce GTX 980. Since the GM200 features 33 per cent more stream processors than the GM204, then it is logical to assume around 30 per cent performance difference.
Nvidia is expected to release its GM200-based graphics solutions sometimes in February or March.
Nvidia did not comment on the news-story.
Discuss on our Facebook page, HERE.
KitGuru Says: 3072 stream processors is a logical number of SPs for GM200. However, we cannot be sure that all of them will be enabled on commercial products. The chip is pretty large and its yields cannot be as high as yields of the GM204, therefore, some stream processors may be disabled for redundancy reasons, at least on first products based on the GM200.
The gaming version of this is in my sights.
We’ll see.. I hope it’s a little more than 30% faster.. Either way I am excited for Nvidia’s next flag ship and even more excited for Amd’s 390x. This year is going to rock!!
Keep in mind that Maxwell also overclocks significantly higher than Kepler. So while stock vs. stock may only be 30% higher, with the overclock difference, OC vs. OC would result likely in the 45-55% difference region.
Lower power consumption, and 50% higher performance (since anyone buying this card would be looking to OC). That’s not too bad for 28nm Maxwell, to be honest. That having been said, I’m not planning to upgrade my Titan setup until 16nm in late 2016.
GM200 features 50% more stream processors not 33%. 3072 on the full chip.
yep so 50% more performance
stupid Pre-voc maths peasants
The reason for lower 30% performance despite having 50% more cores is most likely due to the low base clock of 988MHz (Not cfm yet).