Announcement

Collapse
No announcement yet.

VMware Fusion Stuns VirtualBox In CPU Tests

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • schmidtbag
    replied
    Originally posted by bulletxt View Post
    How many legends my friend. i have more than 45 vm running with vbox from over 1 year and they never hanged,always worked perfectly and serve mysql ,apache and java services perfectly. all this with a gpl software. how you dare say oracle doesn't give a shit of vbox?have seen all feateures it supports? have seen there are packages for all major distros? have you seen new features coming with vbox 4.2? this article is just trying to say vmware is great while vbox is crap. he should do the same with kvm, but he wont because red hat is behind it. michael hates oracle,thats the problem here .
    I'm sure you do all those things just fine - experiences vary between setups. The only gripe I have with virtualbox is there is no GPU passthrough, otherwise I think its a nice program. Performance could be better but it isn't bad. Virtualbox supports the least amount of features of all modern x86 VMs. Xen is probably the best VM out there of all, the only problem is it's very user-unfriendly. it too is open source. The article isn't explicitly saying vmware is great and vbox is crap, it is saying that according to those tests, vmware runs better, and it does. As I said before, run the tests yourself and you can't deny that. Even if you weren't to prove this yourslef, your argument is still invalid; Michael seems to dislike Microsoft more than he dislikes Oracle (if he in fact dislikes Oracle at all) and he's proven several times that Windows generally performs better than linux in gaming, and a few other tasks. L4D2 is the 1 exception.

    There are plenty of things out there where a paid closed-source version is better. Oracle does care more about Virtualbox than OpenOffice, but they didn't buy out Sun for Virtualbox. Even when Sun gained ownwership of VB, it was still open source - if Oracle really cared about it, they'd have collaborated with Sun to improve it for their needs. There may have only been 3 significant releases involving more than just bug fixes since Oracle bought out Sun. They have a lot of work to do to catch up with the others. I have no problem with Virtualbox - it is my go-to VM all the time, but I can still admit it has its problems.

    Leave a comment:


  • Kano
    replied
    @Vadi

    That is not logical in several ways for vbox because the 3d support for d3d games is done via wine libs. Then you can use wine directly much faster. Using a vm for games is somehow weird. Maybe in a dedicated setup with vga paththru.

    Leave a comment:


  • bulletxt
    replied
    Originally posted by schmidtbag View Post
    Holy crap, when are people like you going to stop making this same lame, annoying, unproductive statement? I am so tired of people whining and rumoring that a free content provider is being paid by a company to endorse a product. Virtualbox is free and vmware fusion has a trial - if you really think he's making up all this stuff then why don't you test it for yourself?

    T
    I'm not surprised virtualbox did worse. Oracle doesn't really give a crap about virtualbox, or OpenOffice for that matter. Virtualbox is starting to become the most unmaintained modern x86 emulator. Not only does the VMware company focus on pretty much only emulators/virtualizers but they ask to be paid for most of their better stuff. Set performance issues aside and VMware is still better, you don't need a sponsor to know that.
    How many legends my friend. i have more than 45 vm running with vbox from over 1 year and they never hanged,always worked perfectly and serve mysql ,apache and java services perfectly. all this with a gpl software. how you dare say oracle doesn't give a shit of vbox?have seen all feateures it supports? have seen there are packages for all major distros? have you seen new features coming with vbox 4.2? this article is just trying to say vmware is great while vbox is crap. he should do the same with kvm, but he wont because red hat is behind it. michael hates oracle,thats the problem here .

    Leave a comment:


  • kgonzales
    replied
    Originally posted by droidhacker View Post
    Isn't apache performance tied to the network? That would tend to indicate that the networking layer of vmware is a *little* more efficient than virtualbox under apple. Quite frankly however, you'd have to be completely insane to run a virtualized web server for serving static pages. The *smart* way to do it is to run apache virtual hosts.
    Two issues here:

    1) Many, many enterprises virtualize 50+% of their workloads. Web services are often the earliest one to move to a virtualized environment. Or a cloud environment, which is virtualized.

    2) Given that we are talking desktop virtualization, these results matter when you are running a small test environment on your laptop.

    Originally posted by droidhacker View Post
    Further, I see absolutely no relevance AT ALL in the results, since they are tested over APPLE. The performance of a virtual machine is VERY strongly tied to the host operating system.

    Try again, this time with a LINUX HOST.
    I see a GREAT amount of relevance to these tests as I run alot of Linux VMs on OSX, and check the performance of VMWare vs VirtualBox vs Parallels often.

    Leave a comment:


  • Vadi
    replied
    Originally posted by Kano View Post
    How really cares about 3d speed in a vm? The normal usecase is running win inside a vm from linux host and that usually not because of some games but other software that does not run with wine. It might be already enough to use aero.
    I would if I was trying to use it to play games. That's not really feasible yet... but it would certainly work better than Wine if it did.

    Leave a comment:


  • russofris
    replied
    Originally posted by Kano View Post
    How really cares about 3d speed in a vm? The normal usecase is running win inside a vm from linux host and that usually not because of some games but other software that does not run with wine. It might be already enough to use aero.
    Kano, I know it was probably a rhetorical question. The obvious answer is "Anyone running a 3D application under a VM cares". What many Mac users care about is whether the VM and virtualized executables run at full resolution, at their display devices sync rate, without tearing. It is arbitrary to me if an app/game/compositedDesktop runs at 100 or 200 fps. What matters is that it runs consistently at 60fps without tearing. To restate, the user shouldn't have to worry about whether or not they are running in a VM.

    A note regarding Fusion: I have noticed that Bridged mode networking tends to have problems. The virtual network adapter on the Ubuntu guest will work fine for a while, but stop working when I/O gets too high, requiring a guest restart or manually disabling/re-enabling the network adapter. I'm in a fortunate position where I can use NAT mode and remain unaffected.
    Originally posted by droidhacker
    Quite frankly however, you'd have to be completely insane to run a virtualized web server for serving static pages. The *smart* way to do it is to run apache virtual hosts.
    Perhaps you are simply being obtuse, but there are quite a few reasons to run your Apache httpd servers in a virtualized environment. Coming from the enterprise, cloud, and managed services background, I cannot think of a production Apache server that has not been virtualized in our DC. I personally tend to several hundred Apache httpd servers and a thousand Apache tomcat servers under ESX. The only bare-metal servers in our enterprise that come to mind are running Oracle RAC.

    F

    Leave a comment:


  • Kano
    replied
    How really cares about 3d speed in a vm? The normal usecase is running win inside a vm from linux host and that usually not because of some games but other software that does not run with wine. It might be already enough to use aero.

    Leave a comment:


  • schmidtbag
    replied
    Originally posted by bulletxt View Post
    Michael let's be honest, Vmware payed you to write such a stupid article.
    Holy crap, when are people like you going to stop making this same lame, annoying, unproductive statement? I am so tired of people whining and rumoring that a free content provider is being paid by a company to endorse a product. Virtualbox is free and vmware fusion has a trial - if you really think he's making up all this stuff then why don't you test it for yourself?


    I'm not surprised virtualbox did worse. Oracle doesn't really give a crap about virtualbox, or OpenOffice for that matter. Virtualbox is starting to become the most unmaintained modern x86 emulator. Not only does the VMware company focus on pretty much only emulators/virtualizers but they ask to be paid for most of their better stuff. Set performance issues aside and VMware is still better, you don't need a sponsor to know that.

    Leave a comment:


  • bulletxt
    replied
    Michael let's be honest, Vmware payed you to write such a stupid article. I'm really mad with you this time! But I accept it if this is what you need to continue making phoronix exist.

    Leave a comment:


  • droidhacker
    replied
    Isn't apache performance tied to the network? That would tend to indicate that the networking layer of vmware is a *little* more efficient than virtualbox under apple. Quite frankly however, you'd have to be completely insane to run a virtualized web server for serving static pages. The *smart* way to do it is to run apache virtual hosts.

    The rest of the tests show only MARGINAL differences, not the dramatic annhilation as the article would suggest.

    Further, I see absolutely no relevance AT ALL in the results, since they are tested over APPLE. The performance of a virtual machine is VERY strongly tied to the host operating system.



    Try again, this time with a LINUX HOST.

    Leave a comment:

Working...
X