Over the last week I gave some thought to getting a decent gaming PC. I also have a PS4 that I play most of my non-Mac games on. My Alienware Alpha handles the general games that Windows is the only way to play the game, or I need the PC multiplayer.
I then started thinking about just getting en eGPU for my MacBook Pro, and running Bootcamp. I have an i7 15" from 2016, with 16gb of ram. After reading the egpu.io forums, getting Bootcamp and an eGPU to work on machines with a discrete GPU (like mine), is a pain in the ass. You basically need to hack a few things, and shut off PCI lanes on the MBP.
Then I stumbled on this Reddit post about a user’s experience with Parallels and gaming. He claimed it was largely impossible to tell the difference between Bootcamp and Parallels. This is because Parallels uses Metal 2 to handle all the graphics stuff in between Windows and the display.
So, I decided to do some tests. I used my 2016 15" with the Radeon 455 with Parallels running on an external drive, and my Alpha R1. I recorded the benchmarks as best I could. If the game had a built in benchmark, I used that. Otherwise I used Steam to just keep an eye on FPS and ballpark the averages. It wasn’t scientific, but I could handle some margin of error. If the game had a macOS version, I also used that as a control. My benchmarks are at the end of the post.
Even without the eGPU. yeah, he was correct. There were two odd things I noticed: I have no idea why the Alpha benchmark for Tomb Raider was so high, other than maybe it defaulting to a different graphics setting; and any online game (Destiny 2 and WoW) really struggled on the VM. This is a non-issue for me since I was just using those games to collect data. It was interesting to see there really wasn’t a hell of a difference between all of them.
I am likely going to get the eGPU this summer. I still don’t think I will buy a lot of AAA-style games even on PC with it. Those, I will just use my PS4 and soon to be PS5 for. My main goal is to try and get as functionally close to a minimal one-device setup as I can. But, for some of my games this will work great.
Ya, VM is Parallels. I didn’t end up doing a Bootcamp test because I won’t be using it due to the issues with eGPUs. I did a few tests with Bootcamp and the Witcher 3 came in at the same frame rate as Parallels. I meant to mention that in the original post.
With an i7, and 16gb of ram, I am basically towards a decent rig;
With the exception of a handful of games, I either play it on the Mac, or PS4, and then a few others on PC. Those are lighter games;
The enclosure will also give me a USB hub, and a one-cable solution;
Seeing the Parallels numbers made the handful of windows games I play more viable.
I have two computers: a 2016 15" MBP, and an Alienware Alpha R1 (i3, 8gb, Nvidia 860). When I did the tests, I wanted to compare 3 things: How well did the Alpha handle the game; how well did the Mac natively handle the game; and how well did Parallels handle it. Mostly, I was curious if Parallels wouldn’t have the issues I ran into running games before on a virtual.
What I found is for what I want to do now, it’s viable, even without the eGPU. The benchmarks on the 3 test subjects indicated even without the eGPU, there really wasn’t much of a difference in FPS. Which tells me when I do get an GPU, it will do very well.
I don’t think it is that cut and dry myself (but it does depend on the game), and it comes down to what is playable as well, especially when you talk about adding a eGPU.
Its been a long time since I tried the Bootcamp v Parallels v macOS myself, but when I did cooling seemed to be vastly different. Running a game on MacOS (VM or natively) and the fans wouldn’t kick in. Bootcamp and the same game the fans were going full speed. But performance didn’t seem that much greater. I remember being frustrated with the whole thing.
Yeah, and the tests – except for where there was an official benchmark – have a decent margin of error. I pretty much just ran around the place I had my save game in the Witcher.
I am not likely to be playing intensive PC games. Maybe I might get Metro Exodus, but that would be it. Most of the AAA games are going to get offloaded to the PS4.
@Mark_Crump Did you ever pull the trigger on this eGPU? I’ve been thinking about this option.
Basically, I’m stuck on a potato work laptop, with my last desktop scrapped for parts by friends as I couldn’t get back to it (border closures during pandemic). May have to move multiple times in the next year, so hesitant to build a full size tower which I may have to ship soon after. But also getting eager to catch up on some gaming as winter approaches.
So, considering an external monitor/mouse + very fast CPU/basic GPU laptop that can also double as my on-the-road PC (instead of the work potato), with an eventual eGPU, probably in 2021. Seems like it could be a sweet spot in terms of bridging the gap between my impatience, the gap between now and when the 30xx cards are available, and concerns around moving. Of course there are some cost considerations versus a strictly desktop setup, but the real comparison is against a basic laptop + beefy desktop combo, and with mobility of my main machine. My main concern is how much of a performance sacrifice there is of the eGPU combo versus a comparable desktop setup (if any).
