- Jan 8, 2017
- 6,660 (4.19/day)
|System Name||Good enough|
|Processor||AMD Ryzen R7 1700X - 4.0 Ghz / 1.350V|
|Motherboard||ASRock B450M Pro4|
|Cooling||Deepcool Gammaxx L240 V2|
|Memory||16GB - Corsair Vengeance LPX - 3333 Mhz CL16|
|Video Card(s)||OEM Dell GTX 1080 with Kraken G12 + Water 3.0 Performer C|
|Storage||1x Samsung 850 EVO 250GB , 1x Samsung 860 EVO 500GB|
|Display(s)||4K Samsung TV|
|Case||Deepcool Matrexx 70|
Tenor cores are now compliant with accelerating accelerate IEEE-compliant tensor FP64 computations
Fixed it. Tensor cores do tensor operations. That's why they are called tensor cores.
Meaning each SM is now capable of 128 FP64 op per clock which achieves 2.5X the tensor FP64 throughput of V100
Fixed it. Also, every SM can do 64 FP64 ops meaning it can do a total of 9.7 TFLOPS of PF64. Also in your whitepaper which isn't a whitepaper by the way it's just a god damn blog post they specify different performance metrics for the two because they represent different workloads. Are you correcting the words of a billion dollar company trying to sell millions of dollars worth of equipment who would just write a bunch of worthless crap in their arch whitepaper?
Case reopened and closed.
You are so wrong, stubborn and unintelligent, you've exceed all my expectations from past discussions with you. Anyway I thought you "rested" your case many comments ago, why are you still here ?
EgO mUcH ? Remember if you don't want to deal with me anymore then don't tell I'm wrong when I'm not. It's that simple, otherwise we can go on forever, I have all day as I said.