GPU mining in an ESXi VM w/ pass thru?

[Ch33rios]

I decided to just fire up ccminer on my Windows 10 machine with a 1050ti passed through and at first it seemed to be working ok but it ended up error'ing out after about 15 minutes due to a memory error. I played with the intensity levels and still no luck. Within 15 minutes or so it fails.

I ensured the drivers were the latest Nvidia ones and I'm running ccminer-x64 latest from GitHub (tpruvot version 2.2.3). I'm now getting an aes_cpu_init error so I'm wondering if it is even possible to successfully do this on an VM?

This is using the x17 algorithm as well. Haven't tried anything else to see if it matters or not.

Any thoughts for a new guy would be appreciated.

 

[archangel.dmitry]

Not sure about ESXi but I have boot2docker vm running on Proxmox host for days. Though, I am running CPU based mining.

 

[Ch33rios]

Not sure about ESXi but I have boot2docker vm running on Proxmox host for days. Though, I am running CPU based mining.

Oh? How does this impact your other VMs from a compute standpoint? Assuming the docker host is pretty much fully utilizing the CPU cores given to it? My E3 v5 is very under utilized from a CPU standpoint most of the time and it'd be nice to possibly use that extra idle time to play with some mining efforts.

 

[archangel.dmitry]

I get a little over 2.1 kH/s on 12 vCPU vm.

 

[vhypervisor]

I tryed to mine with an esxi vm and pcie passthrou, after hours try to install the gpu driver i swaped to an baremetal ubuntu 16.04 und start mining 5 mh/s are realy okay for an rx460 core no extra power connector (maximum 75 watt)


Sent from my iPhone using Tapatalk

 

[Ch33rios]

Ok I figured it out. I hadn't disabled the VMware SVGA display adapter and that was causing something to go a bit sideways. Once I disabled it in device manager all has been working fine for the past few hours.