• A Great friend to the HardForum with a great kid that he is trying to get a scholorship to continue his schooling. Please give hime a vote! Only 24 hours left! Thanks.
    If you have an VOTE FOR KEENAN!

Does 2600k HT matter for VMs?

blade52x

2[H]4U
2FA
Joined
Sep 9, 2006
Messages
3,230
So I'm running 3 virtual machines on my desktop pretty much at all times now.

1) Windows XP Mode
2) Vista Ultimate
3) Vista Ultimate x64

All together they use up 5.5gb of RAM. My only issue is video gaming has a bit of stuttering going on. And it's probably because the virtual machines use up three cores, leaving the rest of the system with one. For general desktop use, this isn't a problem - I don't notice a thing. It's just in video games where I notice a performance hit. RAM isn't the issue as I have 16gb in total (I even have a 4gb RAM disk).

So my question is: would upgrading my 2500k to a 2600k make any difference here with the 2600k having extra virtual cores? Or am just going to have to wait for 6-8 core processors later on?

Thanks.
 
Hyper-Threading would help, but it depends on how really important it is to you since the 2600K is $100 more expensive than a new 2500K. What do you have those VMs doing while you are gaming? If they are just idle I would think they would have very negligible effects on a quad-core processor especially a Sandy Bridge at that frequency. However, I do not claim to be a virtualization expert, I just dabble with them.
 
Last edited:
They run some automated tasks, and are pretty fully loaded. On system idle with the VMs running I'm at about 65-80% computer usage. I don't run idle VMs. Either they are doing something or they are shut down.

I was more or less curious if it was possible to assign a VM to a virtual core on the 2600k (however I would expect VM performance to be degraded, and possibly to the point where it would be unusable) and leave the real cores for the main operating system.
 
getting a 2600k isn't going to give you significantly more processing power, mabye 10%. (assuming you clock it the same as ur 2500)
You could try setting the vm's to low priority and your games to high priority.
 
getting a 2600k isn't going to give you significantly more processing power, mabye 10%. (assuming you clock it the same as ur 2500)
You could try setting the vm's to low priority and your games to high priority.

It's funny because I actually thought of that, except I set lower priority for the running processes inside of the virtual machines, and it worked. With all three running they now take up ~45% cpu. And everything still runs well within virtual machines. I love SB... :D
 
What are you running the VMs in? VMware? HyperV? VirtualPC?

This concept has intrigued me though, and I started up some of the VMs I have in VMware Workstation and noticed that you can assign the number of processors and the number of cores per processor, but I see no options for threads.

I'm curious as to how things are handled and am going to try and load up the VM and see how the host CPU usage is effected with different configurations.

BTW, I am running a 2600k.

Testing Results:
1 CPU, 1 Core:
VM = 100% CPU Usage
Host = 13% CPU Usage

1 CPU, 2 Cores:
VM = 100% CPU Usage
Host = 25% CPU Usage

So it appears, VMware Workstation is intelligent enough to utilize individual threads, even though the verbage in the settings isn't exactly perfect (CPUs and cores opposted to cores and threads). This would lead me to believe that a 2600K would help for virtualization because, if needed, you can dedicate a thread, instead of a full core to a given process, allowing more threads for the Host OS.

If you are able to handle your issue based off of affinity, then great. I guess I was assuming your VMs were more fully loaded, sounds like they are not.
 
Last edited:
So it does seem like a 2600k would be beneficial. But...

I'm curious as to how things are handled and am going to try and load up the VM and see how the host CPU usage is effected with different configurations.

I've noticed that when the VMs are in idle, they use up very little cpu. It was when they became loaded (100% load in the VM) that translated into 20-25% load on the 2500k. Since I dropped the affinity of what I was running in the VM, the load within the VM dropped to around 60% with actual load on the 2500k dropped to around 15%. I was very happy to see the specific program performance keep up within the VMs after doing that.

By the way, I use VMWare Player. For now I'll stick with the 2500k since I found a solution, and since I don't legally own any more Windows operating systems it's not like I can load up anymore. :p
 
Yeah, if you've found a solution, no need to spend more money. I'm just glad I saw this thread for my own knowledge :).
 
Back
Top