antihelten
Golden Member
- Feb 2, 2012
- 1,764
- 274
- 126
http://www.anandtech.com/show/7166/nvidia-announces-quadro-k6000
Quadro K6000 launched with full 2880 CUDA cores enabled.
http://www.anandtech.com/show/9096/nvidia-announces-quadro-m6000-quadro-vca-2015
http://www.anandtech.com/show/10179/nvidia-announces-24gb-quadro-m6000
Quadro M6000 launched with full 3072 CUDA cores enabled.
Quadro P6000(?) will launched with all 3840 CUDA cores enabled and hopefully with 8 stack HBM2 for 32GB of RAM in mass production, since you don't want to ship a 16GB Quadro flagship a year after you shipped the Quadro M6000 24GB.
Quadro K6000 used the GK180, which is just a slightly tweaked revision of GK110, and it launched 8 months after the first GK110 equipped Tesla card
Quadro M6000 launched a few days after Titan X, but then again GM200 was never a compute focused chip, and as such is probably a poor example to compare with in regards to GP100.
That means we can probably expect at least an 8 month gap between a Tesla GP100 and a fully unlocked Quadro GP100, which would put it around Q4 2017. At that point you're getting awfully close to Navi and Volta.