Please find my Geekbench OpenCL and Metal benchmarks. There is improvement in Geekbench, but in Luxmark and GFXBench Metal the results are predictably the same as before because they were already optimal.
MacBook Pro 15" Mid 2018. Radeon 5700XT Anniversary Edition. Razer Core X.
@craftsman, is this consistent or one-off?
purge-wrangler ✧ tbt-flash ✧ purge-nvda ✧ set-eGPU
Insights Into macOS Video Editing Performance
Master Threads:
2014 15-inch MacBook Pro 750M
2018 15-inch MacBook Pro
2019 13" MacBook Pro [8th,4C,U] + RX Vega 64 @ 32Gbps-TB3 (Mantiz Venus) + macOS 10.14.6 & Win10 [build link]
@mac_editor, consistent with the usual margin of error. This is on external display only and clamshell mode, just in case anyone wonders about the previously reported issues with multiple displays when running Geekbench.
So Geekbench GPU scores on macOS seem to be related to the resolution of the external monitor, which makes absolutely no sense since the tests are performed off screen in memory only.
Big Sur beta 3
5700XT with Eizo 1200p display - circa 60000 points
5700XT with internal retina display only - circa 42000 points
5700XT with external 4K display - circa 33000 points
As the onscreen resolution increases the scores go down.
2018 Mac Mini i5 / 6-core 3.0-4.1 Ghz / 32GB / 256GB / Intel UHD 630 / PowerColor Gaming Station / PowerColor dual-fan RX 5700 XT / Mac OS 10.15.6 Catalina / Windows 2004 😏 / Dual LG 27" 4K displays
Big Sur beta 4
Geekbench the same.
GFXBench has lost 30% performance in offscreen Aztec Ruins. I'm getting around 130fps now. Used to get circa 200fps in Mojave and Catalina.
Unigene Valley has lost 10%. Now getting about 2950 points. Used to get about 3300 points.
DisplayPort output suddenly stopped working. Big Sur will detect the display and then turn it off after a second before it displays an image there.
Solution if this happens to you : Choose the reset option in the display's settings. Clear everything.
Big Sur Beta 9. Radeon 5700 XT.
GFXBench Metal is 30% slower than Catalina.
Shadow of the Tomb Raider is 15-20% slower than Catalina.
It looks like 3D performance has gone down the toilet and it's not coming back. If you're thinking of buying an eGPU for macOS gaming forget about. Use a PC or a console. Big Sur finally confirms what people have always been annoyed about for 20 years. macOS is not a gaming platform and the GPU drivers will always be the slowest of all platforms.
@craftsman, thank you for sharing. That's disappointing to hear. If I have time, I'll try and run a couple of A/B tests as well.
purge-wrangler ✧ tbt-flash ✧ purge-nvda ✧ set-eGPU
Insights Into macOS Video Editing Performance
Master Threads:
2014 15-inch MacBook Pro 750M
2018 15-inch MacBook Pro
2019 13" MacBook Pro [8th,4C,U] + RX Vega 64 @ 32Gbps-TB3 (Mantiz Venus) + macOS 10.14.6 & Win10 [build link]
Big Sur Beta 9. Radeon 5700 XT.
GFXBench Metal is 30% slower than Catalina.
Shadow of the Tomb Raider is 15-20% slower than Catalina.
It looks like 3D performance has gone down the toilet and it's not coming back. If you're thinking of buying an eGPU for macOS gaming forget about. Use a PC or a console. Big Sur finally confirms what people have always been annoyed about for 20 years. macOS is not a gaming platform and the GPU drivers will always be the slowest of all platforms.
Pretty sad to ear but I have still hope. Often GPU drivers in macOS gets a little better with pain releases. Anyway we will discover Big Sur actual performance soon.
MacBook Pro 2020 Touch Bar i7 quad-core 2.3Ghz - 16GB RAM - 1TB PCIe SSD
my awesome Radeon VII eGPU
my Mantiz Venus extreme mod with Sapphire Nitro+ RX Vega 64
2018 13" MacBook Pro [8th,4C,U] + Radeon VII @ 32Gbps-TB3 (Mantiz Venus) + macOS 10.15 [build link]