Virtualbox 7 so slooowwwwww.


cereberus

Well-known member
Guru
VIP
Local time
11:20 AM
Posts
5,687
OS
Windows 10 Pro + others in VHDs
Just done a comparison of Hyper-V with Virtualbox 7 (VB7 was in installation with no Hyper-V enabled), and also KVM from Linux Host

CPU - similar, worse for KVM
2D - VB awful but KVM slightly better
Memory - similar in all 3
Diskmark - VB awful, KVM much better but still 50% of Hyper-V


In my opinion, this shows how much more efficient a type 1 hypervisor is when it directly shares video and disk resources rather than type 2 emulation. On balance hard to say which is a better type 2 VB or KVM but not much in it.



Hyper-V on left, VB7 on right

BMHV.pngBMVB.png

Edit - results from KVM (Linux as Host OS) - weird results CPU worse, graphics marginally better, disk much better.
Still nowhere up to Hyper-V.

1676840855117.png
 
Last edited:

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
I agree with your opinion, this is indeed one of the differences between type-1 and type-2 hypervisors. Hyper-V FTW! But I know some people don't like it. VMware is cool too.
 

My Computer

System One

  • OS
    Windows 11
Just done a comparison of Hyper-V with Virtualbox 7 (VB7 was in installation with no Hyper-V enabled)

CPU - similar
2D - VB awful
Memory - similar
Diskmark - VB awful

In my opinion, this shows how much more efficient a type 1 hypervisor is when it directly shares video and disk resources rather than type 2 emulation.



Hyper-V on left, VB7 on right

View attachment 51453View attachment 51454
Hyper-V is far better than VBox or VMware -- significantly better -- no question -- but you can improve most VM's by using "Native I/O" rather than the default VM I/O which although more compatible is a cost on performance. Disk I/O performance these days on modern machines with decent CPU and RAM (and usually graphics too) is far more likely to degrade performance than almost anything else.

Most VM systems have facilities for attaching "physical disks" rather than having to create Virtual disks and store your program and data there.


cheers
jimbo
 

My Computer

System One

  • OS
    Windows XP,7,10,11 Linux Arch Linux
    Computer type
    PC/Desktop
    CPU
    2 X Intel i7
Hyper-V is far better than VBox or VMware -- significantly better -- no question -- but you can improve most VM's by using "Native I/O" rather than the default VM I/O which although more compatible is a cost on performance. Disk I/O performance these days on modern machines with decent CPU and RAM (and usually graphics too) is far more likely to degrade performance than almost anything else.

Most VM systems have facilities for attaching "physical disks" rather than having to create Virtual disks and store your program and data there.


cheers
jimbo
Sure - this would probably help a bit but I do not like attaching physical disks directly to vm.

As an aside, I thought I should be able to attach an existing vhdx file though but vb7 STILL does not support vhdx files but does fully support .vhd files (allegedly).

I cannot believe after all these years .vhdx files are still not supported (they can be read from but not written to)!

Poor show for VB7.
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
Just done a comparison of Hyper-V with Virtualbox
My only experience with Hyper-V and VirtualBox is in conjunction with Macrium Reflect's viBoot. After trying out both for a while, I used the latter for some time because it performed somewhat better on my machine. Then, out of curiosity, I tried the former again and it outperformed the latter by some margin, so I switched to the latter until today.

Heavens know why the performance changes happened.
 

My Computer

System One

  • OS
    Windows 10 Pro
My only experience with Hyper-V and VirtualBox is in conjunction with Macrium Reflect's viBoot. After trying out both for a while, I used the latter for some time because it performed somewhat better on my machine. Then, out of curiosity, I tried the former again and it outperformed the latter by some margin, so I switched to the latter until today.

Heavens know why the performance changes happened.
Yeah - that is a bit odd. Something must have been slowing Hyper-V down though.

The only thing I can immediately think of is perhaps you less RAM to Hyper-V than to VB. The default in Viboot is only 2 GB, and maybe you assigned 2GB to Hyper-V but more to VB. In this case the Hyper-V instance may have been doing a lot of pagefiling?

I guess we will probably never know, but I am glad Hyper-V is now performing better (or I suppose VB worse), but like for like, I would always expect Hyper-V to be more efficient in general.

The one big weakness with Hyper-V is the limitations when in basic mode (no sound, hard to share drives etc) rather than enhanced mode.

The reason for this is historic - Hyper-V was originally designed for Windows Server versions, designed to run multiple Windows vms in an Azure type environment. It was later ported to Windows 8/10 for the domestic market, but no real attempt was made to make it more non-Windows OS flexible.

Now this would be a major feature improvement rather than emoji 15 (I never even used emoji 1 - 14 lol).

As a minimum, you would think they could make Windows 10 Home work in enhanced mode (actually you can by installing an RDP server but not strictly allowed).
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
Yeah - that is a bit odd. Something must have been slowing Hyper-V down though.

The only thing I can immediately think of is perhaps you less RAM to Hyper-V than to VB. The default in Viboot is only 2 GB, and maybe you assigned 2GB to Hyper-V but more to VB. In this case the Hyper-V instance may have been doing a lot of pagefiling?

I guess we will probably never know, but I am glad Hyper-V is now performing better (or I suppose VB worse), but like for like, I would always expect Hyper-V to be more efficient in general.

The one big weakness with Hyper-V is the limitations when in basic mode (no sound, hard to share drives etc) rather than enhanced mode.

The reason for this is historic - Hyper-V was originally designed for Windows Server versions, designed to run multiple Windows vms in an Azure type environment. It was later ported to Windows 8/10 for the domestic market, but no real attempt was made to make it more non-Windows OS flexible.

Now this would be a major feature improvement rather than emoji 15 (I never even used emoji 1 - 14 lol).

As a minimum, you would think they could make Windows 10 Home work in enhanced mode (actually you can by installing an RDP server but not strictly allowed).

Hi there
Confirms it runs like "Greased lightning" on Windows server 2022. Even the standard edition with desktop GUI. If you want to try it you can get free 180 day evaluation -- (and you can via slmgr rearm get another 5 extensions of the 180 day period !!!!)

Note though things like Macrium etc won't run on Server (unless you buy server edition) --Office LTSC 2021 runs OK on it. However for simply testing HYPER-V and comparing performance with the same VM running on W11 could be an interesting exercise.

Cheers
jimbo
 

My Computer

System One

  • OS
    Windows XP,7,10,11 Linux Arch Linux
    Computer type
    PC/Desktop
    CPU
    2 X Intel i7
I use Virtualbox on Windows11 and have no problem. But then again I have 16gb Ram too.
 

My Computers

System One System Two

  • OS
    Windows 11 Professional (Via VMware)
    Computer type
    PC/Desktop
    Manufacturer/Model
    HP
    CPU
    AMD Athlon(tm) II X2 220 Processor × 2
    Motherboard
    HP
    Memory
    4GB
    Graphics Card(s)
    Advanced Micro Devices, Inc. [AMD/ATI] Caicos PRO [Radeon HD 7450] (prog-if 00 [VGA controller])
    Sound Card
    Nvidia High Definition System Audio
    Monitor(s) Displays
    Phillips TV
    Screen Resolution
    32" TV
    Hard Drives
    222GB SSD
    Keyboard
    Generic Wireless Keyboard
    Mouse
    Custom Wireless Mouse
    Internet Speed
    400 mbps
    Browser
    Google Chrome
    Antivirus
    Webroot SecureAnywhere
  • Operating System
    Windows 11 Professional
    Computer type
    Laptop
    Manufacturer/Model
    Letsung
    CPU
    Intel Celeron N4020C
    Motherboard
    Intel
    Memory
    6GB DDR4 Ram
    Graphics card(s)
    Intel Celeron
    Sound Card
    Nvidia Geforce
    Monitor(s) Displays
    24" TV
    Screen Resolution
    24"
    Hard Drives
    500G NVME M2
    Internet Speed
    400mbps
    Browser
    Google Chrome
    Antivirus
    Webroot Secure Anywhere Internet Security Complete
Interesting results.
Some were expected but others were a complete surprise.
Thanks for adding KVM!
Those are indeed pretty low scores for VBox.

@cereberus
What are the VM specs?
I assume you used default - (disk) formats - vm settings - same cpu count / ram - per hypervisor.
 

My Computers

System One System Two

  • OS
    Windows 11 Pro
    Computer type
    PC/Desktop
    Manufacturer/Model
    Intel NUC
    CPU
    i3 8109U
    Motherboard
    Intel
    Memory
    16GB DDR4 @2400
    Graphics Card(s)
    Intel Iris Plus Graphics 655
    Sound Card
    Intel / Realtek HD Audio
    Monitor(s) Displays
    LG-32ML600M
    Screen Resolution
    1920x1080
    Hard Drives
    Intel SSD 250GB + Samsung QVO SSD 1TB
    PSU
    Adapter
    Cooling
    The usual NUC airflow
    Keyboard
    Logitech Orion G610
    Mouse
    SteelSeries Rival 100 Red
    Internet Speed
    Good enough
    Browser
    Chromium, Edge, Firefox
    Antivirus
    Windows Defender
  • Operating System
    CentOS 9 Stream / Alma / Rocky / Fedora
    Computer type
    Laptop
    Manufacturer/Model
    TOSHIBA
    CPU
    Intel i7 4800MQ
    Motherboard
    TOSHIBA
    Memory
    32GB DDR3 @1600
    Graphics card(s)
    NVIDIA Quadro K2100M
    Sound Card
    Realtek HD Audio
    Monitor(s) Displays
    Built-in
    Screen Resolution
    1920x1080
Just tested
Interesting results.
Some were expected but others were a complete surprise.
Thanks for adding KVM!
Those are indeed pretty low scores for VBox.

@cereberus
What are the VM specs?
I assume you used default - (disk) formats - vm settings - same cpu count / ram - per hypervisor.
Yep same settings.
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
Hi folks

VBOX is indeed Sloooooooooooooooow - it's slower than a one legged dog stuck on a perfectly smooth icepond or a dragonfly trying to swim through a sea of molasses -- HOST --i7 64G machine -- running W11 pro insider build (dev).

Screenshot 2023-02-25 142737.png

The only advantage I can see of using VBOX is that it's free and seems simple to use.

Is there way to get iso cdrom /optical device to be SATA rather than IDE and to optimise disk I/O -- otherwise any attempts to improve performance on this VM platform will be doomed to failure.

VMware workstation is far superior -- but it's of course a paid for product while HYPER-V is great too there are a number of considerations for running Non Windows Guests on it.

Cheers
jimbo
 

My Computer

System One

  • OS
    Windows XP,7,10,11 Linux Arch Linux
    Computer type
    PC/Desktop
    CPU
    2 X Intel i7
Hi folks

VBOX is indeed Sloooooooooooooooow - it's slower than a one legged dog stuck on a perfectly smooth icepond or a dragonfly trying to swim through a sea of molasses -- HOST --i7 64G machine -- running W11 pro insider build (dev).

View attachment 53637

The only advantage I can see of using VBOX is that it's free and seems simple to use.

Is there way to get iso cdrom /optical device to be SATA rather than IDE and to optimise disk I/O -- otherwise any attempts to improve performance on this VM platform will be doomed to failure.

VMware workstation is far superior -- but it's of course a paid for product while HYPER-V is great too there are a number of considerations for running Non Windows Guests on it.

Cheers
jimbo
You can use free version of VMWare.

Also turning off Hyper-V hypervisor is recommended if using vmware or virtualbox.
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
For performance issues with VirtualBox on a Windows Host System the first check should be to review the VBox.log for the VM affected for the entry "Attempting fall back to NEM: VT-x is not available" for which there is a Tutorial on the VirtualBox Forums at Board index ‹ General ‹ Howtos and Tutorials ‹ Windows Hosts plus some advice in the VirtualBox User Guide.
Some other issues can arise with other 3rd-party software which attempts to control/make exclusive use of VT-x/AMD-V)/hardware acceleration features e.g. some 'security' applications if hardware acceleration enabled, Docker Desktop etc.

Since the release of macOS 13.0 (Ventura) Mac computers also have a native Virtualisation Platform activated which prevents software from directly accessing VT-x, but IME so far this doesn't appear to have the same negative impact on 3rd-party virtualisation software performance that an active Hyper-V system does on Windows, suggesting Apple may have provided a more effective API for developers/users than Microsoft has.

For Linux Hosts a common problem seems to be trying to use VirtualBox with the KVM platform already active, which again blocks access to VT-x/AMD-v.

Personally, I haven't faced any real-world performance issues with VirtualBox, beyond the occasional software bug/regression, but I make sure to de-activate Hyper-V and other known 'blockers' on Windows PCs when using any 3rd-party virtualisation. However, differing use cases can, of course, produce different results.
 

My Computers

System One System Two

  • OS
    macOS (plus VMs: Windows XP, 7, 10 Home/Pro, 11 Home/Pro, Linux Distros)
    Computer type
    Laptop
    Manufacturer/Model
    a) Apple MacBook Pro (Intel) - 2019 b) Apple MacBook Pro M1 MAX - 2021
    CPU
    a) Intel i9 b) M1 MAX (ARM)
    Memory
    a) 16GB b) 32GB
    Hard Drives
    a) 1TB SSD + 256GB SD Card b) 1TB SSD (+ 1TB SD Card)
    Browser
    a) Safari/Vivaldi/DuckDuckGo b) Safari/DuckDuckGo
    Antivirus
    -
  • Operating System
    Windows 11 Pro (plus VirtualBox VMs: Windows 11 Pro & Linux Distros)
    Computer type
    Laptop
    Manufacturer/Model
    a) Microsoft Surface Book 2, b) HP Spectre X360
    CPU
    a) i7, b) i7
    Memory
    a) 16GB, b) 16GB
    Hard Drives
    a) 1TB SSD, b) 1TB SSD
    Browser
    a) MS Edge, b) MS Edge
    Antivirus
    a) Defender, b) Defender
vmware - better than kvm or virtualbox but still not as good as Hyper-V.

1677342245677.png
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
For performance issues with VirtualBox on a Windows Host System the first check should be to review the VBox.log for the VM affected for the entry "Attempting fall back to NEM: VT-x is not available" for which there is a Tutorial on the VirtualBox Forums at Board index ‹ General ‹ Howtos and Tutorials ‹ Windows Hosts plus some advice in the VirtualBox User Guide.
Some other issues can arise with other 3rd-party software which attempts to control/make exclusive use of VT-x/AMD-V)/hardware acceleration features e.g. some 'security' applications if hardware acceleration enabled, Docker Desktop etc.

Since the release of macOS 13.0 (Ventura) Mac computers also have a native Virtualisation Platform activated which prevents software from directly accessing VT-x, but IME so far this doesn't appear to have the same negative impact on 3rd-party virtualisation software performance that an active Hyper-V system does on Windows, suggesting Apple may have provided a more effective API for developers/users than Microsoft has.

For Linux Hosts a common problem seems to be trying to use VirtualBox with the KVM platform already active, which again blocks access to VT-x/AMD-v.

Personally, I haven't faced any real-world performance issues with VirtualBox, beyond the occasional software bug/regression, but I make sure to de-activate Hyper-V and other known 'blockers' on Windows PCs when using any 3rd-party virtualisation. However, differing use cases can, of course, produce different results.
All my tests were done on separate host windows without Hyper-V installed (except for Hyper-V test). Linux KVM was a totally separate installation as well. VT-x was available in virtualbox. I am fully aware if hyper-v hypervisor is on it affect performance of type 2 hypervisors. To be fair, the MS APIs that allow coexistance of type 2 vms with hyper-V being installed is not as bad as it used to be.

I have just tested vmware and clearly better than kvm or virtualbox but still not as good as Hyper-V.

These tests all done on fresh Windows 11 installations, same size RAM (4GB), same storage (80 GB), 4 cores.

I have done similar tests before on my older laptop (different cpu, storage, graphics), and I always got similar results - Hyper-V is just so much faster.
 
Last edited:

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
I now am using Windows 11 Pro on VMware Professional. For my needs, the speed, accuracy, and overall results are better for Windows 11 using vmware. Rather than Virtualbox.
 

My Computers

System One System Two

  • OS
    Windows 11 Professional (Via VMware)
    Computer type
    PC/Desktop
    Manufacturer/Model
    HP
    CPU
    AMD Athlon(tm) II X2 220 Processor × 2
    Motherboard
    HP
    Memory
    4GB
    Graphics Card(s)
    Advanced Micro Devices, Inc. [AMD/ATI] Caicos PRO [Radeon HD 7450] (prog-if 00 [VGA controller])
    Sound Card
    Nvidia High Definition System Audio
    Monitor(s) Displays
    Phillips TV
    Screen Resolution
    32" TV
    Hard Drives
    222GB SSD
    Keyboard
    Generic Wireless Keyboard
    Mouse
    Custom Wireless Mouse
    Internet Speed
    400 mbps
    Browser
    Google Chrome
    Antivirus
    Webroot SecureAnywhere
  • Operating System
    Windows 11 Professional
    Computer type
    Laptop
    Manufacturer/Model
    Letsung
    CPU
    Intel Celeron N4020C
    Motherboard
    Intel
    Memory
    6GB DDR4 Ram
    Graphics card(s)
    Intel Celeron
    Sound Card
    Nvidia Geforce
    Monitor(s) Displays
    24" TV
    Screen Resolution
    24"
    Hard Drives
    500G NVME M2
    Internet Speed
    400mbps
    Browser
    Google Chrome
    Antivirus
    Webroot Secure Anywhere Internet Security Complete
Here is my earlier Hyper-V and new Host OS results.

It shows clearly Hyper-V performs well when compared with Host.

To me this proves Hyper-V is just far superior as a type 1 hypervisor. If I had to use a type 2 hypervisor, VMWare is far better than Virtualbox 7.

I am unlikely to use KVM but showed it for interest. I have seen people claim it is more efficient than Windows VMs due to lower Linux overheads but the evidence of my tests does not support that.

Of course, these results are based on my setup. Others may get different results of course.

All tests are based on default installations of vm software e.g. I have not tried to use more advanced passthrough techniques or use physical drives rather than virtual drives which may improve performance. Tests for the future.

1677434785437.png

1677434716890.png
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
@cereberus
That looks like an upgrade.
So the last two screenshots are Hyper-V for two different hosts.
It clearly shows that the better the host, the better the guest in overal performance.
And of course in that case with all those different machines out there people will get different results all across the board.
 

My Computers

System One System Two

  • OS
    Windows 11 Pro
    Computer type
    PC/Desktop
    Manufacturer/Model
    Intel NUC
    CPU
    i3 8109U
    Motherboard
    Intel
    Memory
    16GB DDR4 @2400
    Graphics Card(s)
    Intel Iris Plus Graphics 655
    Sound Card
    Intel / Realtek HD Audio
    Monitor(s) Displays
    LG-32ML600M
    Screen Resolution
    1920x1080
    Hard Drives
    Intel SSD 250GB + Samsung QVO SSD 1TB
    PSU
    Adapter
    Cooling
    The usual NUC airflow
    Keyboard
    Logitech Orion G610
    Mouse
    SteelSeries Rival 100 Red
    Internet Speed
    Good enough
    Browser
    Chromium, Edge, Firefox
    Antivirus
    Windows Defender
  • Operating System
    CentOS 9 Stream / Alma / Rocky / Fedora
    Computer type
    Laptop
    Manufacturer/Model
    TOSHIBA
    CPU
    Intel i7 4800MQ
    Motherboard
    TOSHIBA
    Memory
    32GB DDR3 @1600
    Graphics card(s)
    NVIDIA Quadro K2100M
    Sound Card
    Realtek HD Audio
    Monitor(s) Displays
    Built-in
    Screen Resolution
    1920x1080
@cereberus
That looks like an upgrade.
So the last two screenshots are Hyper-V for two different hosts.
It clearly shows that the better the host, the better the guest in overal performance.
And of course in that case with all those different machines out there people will get different results all across the board.
No - you misunderstand.

I was just showing running passmark in Hyper-V was close to results of running it on Host OS i e. not in a VM.

It shows Hyper-V vms are very efficient on my pc compared with alternatives.

There is no point in comparing results across different host pcs other than boasting about ones "equipment" lol.
 

My Computer

System One

  • OS
    Windows 10 Pro + others in VHDs
    Computer type
    Laptop
    Manufacturer/Model
    ASUS Vivobook 14
    CPU
    I7
    Motherboard
    Yep, Laptop has one.
    Memory
    16 GB
    Graphics Card(s)
    Integrated Intel Iris XE
    Sound Card
    Realtek built in
    Monitor(s) Displays
    N/A
    Screen Resolution
    1920x1080
    Hard Drives
    1 TB Optane NVME SSD, 1 TB NVME SSD
    PSU
    Yep, got one
    Case
    Yep, got one
    Cooling
    Stella Artois
    Keyboard
    Built in
    Mouse
    Bluetooth , wired
    Internet Speed
    72 Mb/s :-(
    Browser
    Edge mostly
    Antivirus
    Defender
    Other Info
    TPM 2.0
No - you misunderstand.

I was just showing running passmark in Hyper-V was close to results of running it on Host OS i e. not in a VM.

It shows Hyper-V vms are very efficient on my pc compared with alternatives.

There is no point in comparing results across different host pcs other than boasting about ones "equipment" lol.

Indeed.
Thanks for the clarification! I missed second paragraph somehow. My bad.

That makes sense since the second screenshot has 3D score, it's the host. ;-)
Pretty good result for the VM compared.

The short conclusion is: Windows VM under Hyper-V performs best, excluding 3D tests which are not applicable.

One needs the patience for these kind of tests since all hypervisors need to run on the same machine each on its own optimal run.
Your results are welcome, so thanks again. I didn't tried this yet for practical reasons and time.
It's complicated already to run VirtualBox om my current setup: I have to disable Hyper-V but also memory integrity which makes this less attractive in this case. In the rare case I need to run some exotic VM that doesn't work in Hyper-V I might go with VirtualBox. Then for KVM: you need to install Linux and set it up on same host for conclusive tests, or attempt to run it from live USB, if possible.
I simply run KVM on a separate host and I am very pleased with it there.
 

My Computers

System One System Two

  • OS
    Windows 11 Pro
    Computer type
    PC/Desktop
    Manufacturer/Model
    Intel NUC
    CPU
    i3 8109U
    Motherboard
    Intel
    Memory
    16GB DDR4 @2400
    Graphics Card(s)
    Intel Iris Plus Graphics 655
    Sound Card
    Intel / Realtek HD Audio
    Monitor(s) Displays
    LG-32ML600M
    Screen Resolution
    1920x1080
    Hard Drives
    Intel SSD 250GB + Samsung QVO SSD 1TB
    PSU
    Adapter
    Cooling
    The usual NUC airflow
    Keyboard
    Logitech Orion G610
    Mouse
    SteelSeries Rival 100 Red
    Internet Speed
    Good enough
    Browser
    Chromium, Edge, Firefox
    Antivirus
    Windows Defender
  • Operating System
    CentOS 9 Stream / Alma / Rocky / Fedora
    Computer type
    Laptop
    Manufacturer/Model
    TOSHIBA
    CPU
    Intel i7 4800MQ
    Motherboard
    TOSHIBA
    Memory
    32GB DDR3 @1600
    Graphics card(s)
    NVIDIA Quadro K2100M
    Sound Card
    Realtek HD Audio
    Monitor(s) Displays
    Built-in
    Screen Resolution
    1920x1080
Back
Top Bottom