Out of interest how many Durango GPU's would need to match the PS4 GPU's power (only counting raw flops).
Quick rundown (only counting raw flops)
PS4 GPU TFLOPS -- 1843200 (1152 ops/clock * (1 mul + 1 add) * 800 MHz = 1.84 TFLOPS)
Xbox GPU TFLOPS -- 1228800 (768 ops/clock * (1 mul + 1 add) * 800 MHz = 1.22 TFLOPS)
TFLOPS gap between Xbox GPU and PS4 GPU 614400 (614GFLOPS)
614400 * 2 = 1228800, which is the full performance of Durango GPU.
614400 * 3 = 1843200, which is the full performance of PS4 GPU.
To be pretty simplistic about it -- 1843200 / 10 = 184320 (This is 10% of PS4's total flops performance)
184320 * 5 = 921600 (50% of PS4's total flops performance aka 921GFLOPS)
184320 * 6.6 = 1216512 (Close enough to Durango's full performance)
More or less, Durango's GPU is around 66 to 67% of the PS4 GPU's power in raw flops. More or less it would take 1.33 Durango GPUs, which was already mentioned jokingly to poke fun at me by Mantis, but I wanted to show you how I arrived at most of the numbers.
Fine, then say the Durango is 33% less powerful. That would be correct. Saying the PS4 is 33% more powerful than the Durango because the Durango is 33% less powerful than the PS4 is absolutely wrong.
I understand that, but we aren't exactly trying to be scientific here. Just speaking in terms of raw flops. No way the raw flops on either of these systems will be the ultimate decider of end performance. There's a lot of other factors that come we have to factor, but we were just speaking purely in terms of raw flops. PS4 devs only having to worry about that single pool of GDDR5 should be important also.
