Yesterday we reported on a story that was circulating regarding the performance of RadGameTool’s Bink Video 2.4. During the story we reported that RadGameTool’s own website reported the performance numbers of producing a frame of 4K video on both the PS4/X1 along with a PC.
“it can play 4K video frames (3840×2160) in 4 ms on PCs and 11 ms on Sony PS4 or Xbox One using the CPU only (or 1.4 ms and 2.3 ms using GPU acceleration)!” said their own website. There were several conflicting reports regarding this, and we noticed that the numbers seemed to differ a little compared to those on the right hand side of the website. RadGameTools promptly responded to our request to clear up the confusion and we received this email from Jeff.
“…1.4 ms PC, 2.3 ms for both PS4 and XBox One.
Bink on GPU uses both the CPU and the GPU. Those times are the worst time for each (because you can overlap the CPU with the GPU, but the max speed can never be more than the slowest). So the specific times are:
PC (2.8Ghz Core i5 with 4 cores and AMD R9 290x):
CPU: 1.3 ms. GPU: 1.4 ms.
CPU 2.3 ms. GPU: 1.6 ms.
CPU 2.3 ms. GPU: 2.3 ms.
So, the GPU time on the PS4 is less (because there are more compute units, of course), but the best case throughput time is still limited by the CPU on **BOTH** the PS4 and XBox. So, the effective speed is 2.3 ms on both.”
So indeed, the PS4’s GPU is as you’d expect considerably faster – but for Bink video, its performance is rendered to be the same due to the slower CPU in both machines. It’s also quite nice that we’ve got the actual model of the CPU used in the PC (along with its GPU) as we’ve got a (slight) comparison now to the consoles.