Intel Y Processors and Virtual Box


#21

Checking in with hyper V results:

  • starting up is very slow, like 5 mins but I hear that’s a hyper V problem
  • after starting, everything feels reasonably OK - as in, no real complaints about performance
    • using powershell remotely and running simple commands
    • using edge to browse and download programs
    • installing program (sql server and visual studio)

Havent actually tried doing anything with sql server yet, will check back when/if I do.

I might just install powerbi desktop and try some stuff on it - it creates an analysis services instance so I guess should give some kind of idea how the vm will handle some queries

P.s. vm with default settings and win 10 installed from iso.


#22

Update and amendment: running a Win 10 Pro VM on Hyper-V on my i7 is buttery smooth. The startup delay i mentioned earlier was due to a misconfiguration in the network switch.

To summarise, i have zero complaints to report regarding virtualisation on the Y-processor.

I think anyone else who does so will feel the same, as long as they understand that they are running a VM on a dual core CPU.


#23

How big or complex is your VM environment? I’m also interested in seeing if Eve-V is capable of running a small VM test lab.


#24

Hey @Wickedly,
Wow, a fellow BI practitioner! I work in BI (Cognos mostly, trying out some PowerBI). Still waiting for my LB i7 16gb but I was wondering how yours is handling the analytic services, do you have any insights you can share?
Thanks in advance!
Fred


#25

*ONE* CORE.

it can handle up to exactly that much complex.

I’m not actually a BI practitioner, i’m just an excel user that picked up the languages behind those tools really quickly and use them a lot. I’d probably be more efficient in life if i stopped using them and picked up python or something.

Then when i thought of testing a VM (which i don’t use very often), running VS + an SASS instance + querying it seemed like a decently crash-inducing way to test it, lol.

Specifically for a PowerBI-type workload (as in developing and testing models), working with Power BI and the AS instance it creates in the background is fine, buti don’t work with datasets that would large enough to be a problem for it.
I can’t imagine at which point it would choke either, because the main limitation of the V is multi-core, and the (primary) bottleneck with the AS engine is the single threaded process which kind of works out. or really stupid code - the V won’t help with that at all tho.

I didn’t install sql server on the VM so i don’t know if its any different, but running VS to use the integrated workspace server AND THEN trying to query it with Power BI was… not good. lol.


#26

That’s all interesting, thanks for your reply @Wickedly, I can’t wait to test it on my V… One day soon, hopefully!


#27

@Eriol_Ancalagon just to give a precision to what you said at that time for future consultation.
And if you ask the question on stackoverflow you will be punished just because you dare ask this question, you NEVER installed more than one hypervisor in any given machine.
One hypervisor should always be lonely. If you really need several hypervisor then you can always use nested virtualization with some of the hypervisor. I know it’s technically possible to have several hypervisor installed but it’s not because something works as it is than it should be working that way. We don’t authorize several hypervisor for very specific security reasons, not for our pleasure.
An hypervisor is a layer that should have control and can monitor the hardware. Hardware can not and SHOULD not be handled by several control point at the same time. If you do that then you are open to many holes in your security and ultimately data corruption.