Hi all,
I did some research and I came across an interesting fact. If you're looking at the AMD website, they actually have already the data for the M395X online. Link is here.
It seems like the PC version of the R9 M395X is exactly the same as the R9 M295X.
M295X: 28 nm, 32 Compute Units, 2048 Stream Processors, 723 MHz, 4 GB GDDR5, 1250 MHz memory clock, 160 Gb/s memory bandwith, 256-bit memory interface.
M395X: 28 nm, 32 Compute Units, 2048 Stream Processors, 723 MHz, 4 GB GDDR5, 1250 MHz memory clock, 160 Gb/s memory bandwith, 256-bit memory interface.
If AMD is doing such a rebrand, we could at least expect some MHz boost when it comes to the engine, right? Or does that mean that the technicalities are the same, but the underlying architecture/logic would be more efficient?
GPUboss however paints another picture and differentiates the reference design of the M295X. There seems to be a difference when it comes to the MHz.
The clock speed in the Mac version of the R9 295X seems to be 850 instead of 723 MHz, the memory clock speed is 1362 MHz instead of 1250 and thus the floating-point performance and the pixel rate are higher.
I'm at a loss here - GPU wise it seems there's no point to wait for the next revision of the riMac. Or will Apple do some magic and turn up the MHz even higher? I'm completely confused now. Any thoughts?
I did some research and I came across an interesting fact. If you're looking at the AMD website, they actually have already the data for the M395X online. Link is here.
It seems like the PC version of the R9 M395X is exactly the same as the R9 M295X.
M295X: 28 nm, 32 Compute Units, 2048 Stream Processors, 723 MHz, 4 GB GDDR5, 1250 MHz memory clock, 160 Gb/s memory bandwith, 256-bit memory interface.
M395X: 28 nm, 32 Compute Units, 2048 Stream Processors, 723 MHz, 4 GB GDDR5, 1250 MHz memory clock, 160 Gb/s memory bandwith, 256-bit memory interface.
If AMD is doing such a rebrand, we could at least expect some MHz boost when it comes to the engine, right? Or does that mean that the technicalities are the same, but the underlying architecture/logic would be more efficient?
GPUboss however paints another picture and differentiates the reference design of the M295X. There seems to be a difference when it comes to the MHz.
The clock speed in the Mac version of the R9 295X seems to be 850 instead of 723 MHz, the memory clock speed is 1362 MHz instead of 1250 and thus the floating-point performance and the pixel rate are higher.
I'm at a loss here - GPU wise it seems there's no point to wait for the next revision of the riMac. Or will Apple do some magic and turn up the MHz even higher? I'm completely confused now. Any thoughts?