Classic Mac Pro vs Mac Mini eGPU (Gaming)?

[eksu]

I'm trying to come up with a plan on if I should stick with my cMP for another year+, or if I should use this next year to hack Catalina onto it and try to off-ramp (sell and buy) onto another platform.

I mostly use my Mac as a mac I can also occasionally code, transcode videos on and boot into Windows and PC game with. (Hoping one day with DXVK + MoltenVK I can just stay in MacOS).

I've seen some people say that a eGPU on the Mac Mini gets a 20% performance penalty, but I was wondering how the PCIe 3.0x4 link compares to the cMP's PCIe 2.0x16 link and if some of the performance latency was due to going through the thunderbolt controllers, and how the difference in CPU's and thermal throttling factors in.

So, if someone has a cMP AND a Mac Mini with an eGPU inclosure, I would be very interested in some gaming FPS comparisons, or if you've seen different people doing synthetic benchmarks with the similar graphics cards.

Could also be helpful to everyone else wondering if they should stay on the cMP platform or ramp off. Thought I would ask here first, before reaching out to someplace like eGPU.io or the mac mini forum.

Thanks,
EKSU

 

[handheldgames]

I'm trying to come up with a plan on if I should stick with my cMP for another year+, or if I should use this next year to hack Catalina onto it and try to off-ramp (sell and buy) onto another platform.

I mostly use my Mac as a mac I can also occasionally code, transcode videos on and boot into Windows and PC game with. (Hoping one day with DXVK + MoltenVK I can just stay in MacOS).

I've seen some people say that a eGPU on the Mac Mini gets a 20% performance penalty, but I was wondering how the PCIe 3.0x4 link compares to the cMP's PCIe 2.0x16 link and if some of the performance latency was due to going through the thunderbolt controllers, and how the difference in CPU's and thermal throttling factors in.

So, if someone has a cMP AND a Mac Mini with an eGPU inclosure, I would be very interested in some gaming FPS comparisons, or if you've seen different people doing synthetic benchmarks with the similar graphics cards.

Could also be helpful to everyone else wondering if they should stay on the cMP platform or ramp off. Thought I would ask here first, before reaching out to someplace like eGPU.io or the mac mini forum.

Thanks,
EKSU

Have you visited Barefeats.com? That site should have the comparison's your looking for. There are also some threads in the MR Mac-Mini forum forum detailing MacPro users who have made the jump to the new mini.

 

[eksu]

Barefeats is great.

They have three articles that get close, but mostly centering around professional applications (GPU Compute) or not comparing against the cMP.

If they had included the Mac Mini in their Radeon VII round up, I could compare against my machine, but unfortunately it wasn't included.

• Why the 2018 Mac mini really needs an eGPU
• Do Pro Apps Run Faster on the 2018 Mac mini with an eGPU
• Can the 2018 Mac mini with an eGPU compete with the "Big Dogs" running Pro Apps?

I think a sample of some real games would be a very useful benchmark because they may suffer from the lessened PCIe bandwidth, but they might gain from CPU, Memory improvements so long as they don't throttle to hard, so it's hard to guesstimate.

 

[tpivette89]

Barefeats is great.

They have three articles that get close, but mostly centering around professional applications (GPU Compute) or not comparing against the cMP.

If they had included the Mac Mini in their Radeon VII round up, I could compare against my machine, but unfortunately it wasn't included.

• Why the 2018 Mac mini really needs an eGPU
• Do Pro Apps Run Faster on the 2018 Mac mini with an eGPU
• Can the 2018 Mac mini with an eGPU compete with the "Big Dogs" running Pro Apps?

I think a sample of some real games would be a very useful benchmark because they may suffer from the lessened PCIe bandwidth, but they might gain from CPU, Memory improvements so long as they don't throttle to hard, so it's hard to guesstimate.

I have a 4,1 (flashed to a 5,1) with a Radeon RX480 and then swapped the same graphics card into the new Mini using a Razor Core X. Only did testing in Geekbench with the two cards, however.





I don't game or have Windows in either computer, so I've only done synthetic testing. Hope this helps.