GPU

GPU passthrough with ESXi on the Apple 2019 Mac Pro 7,1

12/23/2020 by William Lam 12 Comments

The expandability of the Apple 2019 Mac Pro (7,1) has been the primary reason VMware customers have been so excited for this new platform for virtualizing macOS on ESXi. The most common request that I hear from customers is for GPU passthrough.

Although VMware does not officially support GPU passthrough, even for the existing Apple hardware systems on the VMware HCL, this has been a topic I have been keeping an eye out on, especially from what the VMware community is doing in this space.

My intention for this blog post is to provide a resource for the community on capturing the success and failures when attempting GPU passthrough on a 2019 Mac Pro. For those interested and have capable hardware, you may want to start with the VMware HCL for GPU passthrough devices listed under Virtual Dedicated Graphics Acceleration (vDGA). This may be your best chance to successfully passthrough a GPU that will be recognized by either a macOS or Linux/Windows guest operating system.

If you would like to share your experiences, feel free to leave a comment or reach out by filling the contact form.

Disclaimer: Although ESXi installs and runs on the Apple 2019 Mac Pro 7,1 it is currently not certified on the VMware HCL. There are no timelines on the certification due to challenges with COVID-19.

Passthrough of Integrated GPU (iGPU) for Apple Mac Mini 2018

06/25/2020 by William Lam 29 Comments

After successfully enabling and persisting the passthrough of the iGPU for the latest Intel NUC 10 (Frost Canyon), I thought it was worth experimenting with the Apple Mac Mini 2018 to see if the same could be accomplished with its iGPU, which is an Intel UHD 630. The biggest benefit in addition to unlocking the iGPU for general use is support for Apple's Metal API which gives developers access to the underlying GPU when building and testing MacOS and iOS applications. This is also quite timely as the Apple Mac Mini 2018 was just added to the VMware HCL!

My initial attempt failed when using the latest ESXi 6.7 Update 3 release. After enabling passthrough of the iGPU and rebooting the ESXi host for the change to take affect, the system would get stuck during boot up when loading the dma_iommu_mapper module. After speaking with Engineering, the issue is probably not related to dma_iommu_mapper module but some other module shortly after but without serial console output or ability to see terminal screen, it would be very difficult to debug the issue.

About to give up, my last attempt was try ESXi 7.0 and to my surprise the ESXi host fully booted up after enabling passthrough of the iGPU. It is still not clear on what might be causing the problem for 6.7 but at least 7.0 works!

Note: To be able to successfully power on a MacOS VM running on ESXi 7.0, ensure you have applied the recent ESXi 7.0b patch. You will need to go to the VMware Patch Portal siteto download and apply the update.

Step 1- Enable passthrough of the iGPU using the vSphere UI and then reboot for changes to take affect.

Step 2 - Navigate to Configure->Hardware->Graphics->Host Graphics and change the default graphics type to "Shared Direct"

Passthrough of Integrated GPU (iGPU) for standard Intel NUC

06/18/2020 by William Lam 20 Comments

Earlier this week I found out that it was possible to passthrough the Integrated GPU (iGPU) for standard Intel NUC which was motivated by a question I had saw on the VMware Subreddit. I have written about iGPU passthrough for Intel NUCs before but only for the higher end models which were the Hades Canyon NUC at the time.

Neat! Just found out you can actually passthrough the iGPU of standard Intel NUC. The trick looks to be enabling passthrough using ESXi Embedded Host Client UI & then you can assign it using vSphere UI#Homelab pic.twitter.com/NwuxbXwUMj

— William Lam (@lamw) June 15, 2020

To be honest, I never thought about trying this out with a standard NUC as I figured the iGPU may not be powerful enough or warrant any interests. After sharing the news on Twitter, I came to learn from the community that not only is this desirable for various use cases but some folks have also been doing this for some time now and have shared some the benefits it brings for certain types of workloads.

Can’t take credit. It was one of our collegaues that pointet me to it. Hw transcoding went up a factor of almost x 20. So for specefic workloads the nuc is suddently a lot more capable than before.

— Robert Jensen (@rhjensen) June 15, 2020

I’ve been doing this forever, when I need to crack passwords but don’t need the full 7 gpu rig - all Supermicro and 1080ti GPUs these dayshttps://t.co/GJGRV5eu8f

— Rob VandenBrink (@rvandenbrink) June 15, 2020

seems like this would be great for ESXi + Plex hardware transcoding

— Will Beers (@willbeers) June 15, 2020

Below are the instructions I used to enable iGPU passthrough on an Intel NUC 10 (Frost Canyon) with vSphere 7.0. These instructions should also be applicable for other NUC models and earlier versions of vSphere including details around passthrough configuration persistency which I know some folks have ran into which I was able to figure out as part of this experiment.

GPU Passthrough of Radeon RX Vega M in Intel Hades Canyon

01/15/2019 by William Lam 67 Comments

With the latest Intel Hades Canyon now being able to run ESXi, a number of folks have been interested in taking advantage of the integrated GPU that is included in the system. There are two models of the Hades Canyon, NUC8i7HNK which is the lower end system with Radeon RX Vega M and the NUC8i7HVK which is the higher end system with Radeon RX Vega GH. One of the first thing I had attempted after getting ESXi working on the Hades Canyon was to try to enable passthrough of the iGPU into a Windows GuestOS but in all my attempts, it resulted into a PSOD'ing the ESXi host once you start installing the AMD Drivers from Intel.

A few days ago, one of my readers, Chris78 shared an update where he was able to prevent the ESXi host from PSOD'ing by adding a VM Advanced Setting but he he was still having issues where the Windows GuestOS would now BSOD. This sounded promising, I figure it would not hurt to gave it a try and to my surprise, I was able to successfully passthrough the iGPU to a Windows 10, Windows Server 2016 and 2019 system from my limited testing. After reporting the success back to Chris78 who was still having issues even after using the settings I had used, his conclusion was there may be a difference between the HNK and HVK models, with the latter having BSOD issues. For now, it seems like iGPU can only be passthrough if you have the NUC8i7HNK model.