Recent content by BoMbY

GFXbench and CompuBench are mixing up results sometimes. At least I'm very confident about the ID:

66AF is Radeon VII.

Not necessarily, simple RT could be implemented without many changes: http://diglib.eg.org/handle/10.2312/hpg.20141091.029-040 Of course dedicated circuits would be better. They could also do it like in the paper, and then reserve a fixed amount of CUs do to RT in parallel to the normal...

This is new: // Pre-GFX10 target did not benefit from loop alignment if (!ML || DisableLoopAlignment || (getSubtarget()->getGeneration() < AMDGPUSubtarget::GFX10) || getSubtarget()->hasInstFwdPrefetchBug()) return PrefAlign; // On GFX10 I$ is 4 x 64 bytes cache lines. // By...

These numbers are plain and simple wrong.

Btw. XFX lists Crossfire as supported feature: http://www.xfxforce.com/en-us/products/amd-radeon-vii/amd-radeon-vii-16gb-3xdp-hdmi-triple-fan-rx-vegma3fd6

Probably just the usual fear-mongering.

ROCm is a pretty alive and well "HSA Compliant Runtime and Driver for AMD RADEON GPU’s"

I just wish they would do a 64 CU LC version of Vega20.

An now they claim they spoke to someone from AMD who told them FP64 would not be unlimited: AMD’s Radeon VII GPU Will Not Support Uncapped Double-Precision (FP64)

And did you by any chance ask why?

They didn't announce any Radeon Professional cards so far, and it would be stupid not to limit the DP of a gaming card to 1/8 (like Hawaii was limited), if they plan to sell Pro cards.

Someone on another forum had the idea they may just announce/market it as PCIe 3.0 at first, because some people might get confused, and may think PCIe 4.0 is not backwards compatible, or might complain because it isn't using PCIe 4.0 on PCIe 3.0 platforms.

And what scenarios do you think currently saturate a PCIe 3.0 x16? I think there are very few where you actually need a constant shared memory access?

Okay, right. Radeon VII is also the first to brake the three month Linux driver rule of thumb. They added the device IDs only two month before the release. I think that is a first. Edit: I mean, "may be the first". I think we don't know the actual device ID yet?