Dicehunter
Resident Newb
So AMD came out with a new beta driver, 20.5.1 HWS, Thanks to Neverbackdown for posting this in Quick News, Enabling hardware accelerated GPU scheduling on Navi GPU's, Sorry Vega and Polaris owners, No support just yet.
I uninstalled the previous driver, Booted into safe mode, Ran DDU and then installed the beta driver.
Upon boot I made sure all graphics options in the driver control panel were at default minus V-Sync which I made sure was off along with Freesync being off, In the Wattman section I left everything alone minus power target which I set to the max of 50%+ and the fan to 100%.
All system fans and CPU cooler fans were also set to 100% to make sure thermals were not an issue.
I ran each test 8 x times, 4 x with HWS enabled and 4 x with HWS disabled and took the average score.
System Specs -
CPU: AMD Ryzen 9 - 3950X - Precision Boost Overdrive Enabled
Motherboard: Asus ROG Crosshair Hero VIII X570 *latest bios*
Memory: 16GB *2x8GB* Corsair Vengeance RGB Pro @ 3200MHz
Storage: WD Black NVME SN750 with Heatsink - 1TB
GPU: XFX 5700XT Thicc III Ultra
OS : Windows 10 Pro - Fully up to date
Driver - https://www.amd.com/en/support/kb/release-notes/rn-rad-win-20-5-1-ghs-beta
Cryteks Noir RT benchmark - No HWS
Cryteks Noir RT benchmark - HWS Enabled
Star Control Origins Benchmark - No HWS
Star Control Origins Benchmark -HWS Enabled
Tom Clancy's The Division 2 Benchmark - No HWS
Tom Clancy's The Division 2 Benchmark - HWS Enabled
Unigine Valley Benchmark - No HWS
Ungine Valley - HWS Enabled
So in conclusion not really anything to shout about, The games/benchmarks "felt" smoother although this could be a total placebo, The only game/benchmark that gave a consistent improvement over the 8 runs, 4 x with HWS on and 4 x with HWS off, Was Unigine Valley, Across all tests its minimum FPS was 6FPS faster with HWS enabled, The Division 2 is the only game that exhibited a bit of a bug but it was a non issue when it came to performance, Apparently the games benchmark cannot read GPU usage correctly when HWS is enabled.
I uninstalled the previous driver, Booted into safe mode, Ran DDU and then installed the beta driver.
Upon boot I made sure all graphics options in the driver control panel were at default minus V-Sync which I made sure was off along with Freesync being off, In the Wattman section I left everything alone minus power target which I set to the max of 50%+ and the fan to 100%.
All system fans and CPU cooler fans were also set to 100% to make sure thermals were not an issue.
I ran each test 8 x times, 4 x with HWS enabled and 4 x with HWS disabled and took the average score.
System Specs -
CPU: AMD Ryzen 9 - 3950X - Precision Boost Overdrive Enabled
Motherboard: Asus ROG Crosshair Hero VIII X570 *latest bios*
Memory: 16GB *2x8GB* Corsair Vengeance RGB Pro @ 3200MHz
Storage: WD Black NVME SN750 with Heatsink - 1TB
GPU: XFX 5700XT Thicc III Ultra
OS : Windows 10 Pro - Fully up to date
Driver - https://www.amd.com/en/support/kb/release-notes/rn-rad-win-20-5-1-ghs-beta
Cryteks Noir RT benchmark - No HWS

Cryteks Noir RT benchmark - HWS Enabled

Star Control Origins Benchmark - No HWS

Star Control Origins Benchmark -HWS Enabled

Tom Clancy's The Division 2 Benchmark - No HWS

Tom Clancy's The Division 2 Benchmark - HWS Enabled

Unigine Valley Benchmark - No HWS

Ungine Valley - HWS Enabled

So in conclusion not really anything to shout about, The games/benchmarks "felt" smoother although this could be a total placebo, The only game/benchmark that gave a consistent improvement over the 8 runs, 4 x with HWS on and 4 x with HWS off, Was Unigine Valley, Across all tests its minimum FPS was 6FPS faster with HWS enabled, The Division 2 is the only game that exhibited a bit of a bug but it was a non issue when it came to performance, Apparently the games benchmark cannot read GPU usage correctly when HWS is enabled.
Last edited: