Red Hat makes the Qumranet SPICE protocol open source. A free alternative to ICA/PCoIP?

Listen to this podcast

Yesterday Red Hat used what was arguably the worst-titled press release ever to announce that they are open sourcing their SPICE remote display protocol.

Yesterday Red Hat used what was arguably the worst-titled press release ever to announce that they are open sourcing their SPICE remote display protocol. SPICE was developed by Qumranet a few years ago and made a huge splash at BriForum in 2008 when they demoed the software-based SPICE protocol on a client with multiple monitors running high-def video, audio, and games. (Here are videos of their sponsored breakout session from BriForum 2008 and DEMO Lab product demo from BriForum 2009.)

Later that summer, Qumranet hired Gabe and I to perform an independent analysis of the performance of the SPICE protocol as it compared to RDP and ICA. We wrote a paper of our findings, although I don't know if they ever published that since a few months later they were acquired by Red Hat. (I think our main contact at the time had bigger items on his plate than our little protocol analysis.)

We wrote about that acquisition on, essentially saying that we thought the main reason Red Hat acquired Qumranet was for the KVM hypervisor (which would compete against the open source Xen hypervisor), and we weren't really sure whether Red Hat cared about desktop virtualization at all.

Fast forward to today: Red Hat has is evolving Qumranet's VDI product into Red Hat Enterprise Virtualization for Desktops (currently in beta), and they're converting SPICE to a full-on open source remoting protocol.

So SPICE is vendor-neutral independent remoting protocol that has the advantage of actually existing. What impact will it have? I guess that depends on how good it is.

How good is SPICE?

Is SPICE better than RDP? Will it give Citrix a run for their money in HDX/ICA? Did VMware just waste a lot of money on PCoIP? Is Net2Display even more worthless?

As part of the analysis we did for Qumranet in 2008, we recorded a bunch of videos comparing the performance of running a set of user scripts via three protocols: SPICE, ICA, and RDP. Our videos show SPICE in action against ICA and RDP, each for three use cases (single display, multiple displays, and multimedia apps).

One thing about the videos that's important to note is that our lab was a non-bandwidth constrained environment, so the bandwidth consumption data is not really valid. (Remember that any protocol will take up as much bandwidth as it can when not capped, so we were just trying to compare the architecture of the protocols back then—not do a full bandwidth analysis.) That said, SPICE is actually fairly advanced when it comes to bandwidth. It will make determinations of client capabilities, network characteristics, and other parameters to automatically change its behavior to provide the best user experience possible. (So in some cases it might be sending raw graphics commands to the client which are processed there, and in other cases it might send what amounts to screen bitmaps to the client.)

Leveraging host-side hardware and special hypervisor capabilities: the future is now!

One important thing you need to understand is that SPICE is architected a bit different than ICA and RDP. While ICA and RDP are made of up two components (a remote software component that runs in the OS of the Windows host you're connecting to, and a client), SPICE is actually made up of three components:

  • Remote guest component: A virtual graphics adapter running in the VM, just like ICA/RDP.
  • Client component; The SPICE client software, just like ICA/RDP.
  • Remote host component: A virtual graphics device which the hypervisor makes available to the VM. (This is different than today's ICA/RDP.)

In other words, because SPICE has a hypervisor component, it will only work when your remote hosts are VMs.

Last month we wrote an article asking whether the future of remoting protocols is going to be based on three-tiered architectures (such as SPICE). While not everyone thought it was a good idea, there's actually a lot of evidence that the industry is going down that path anyway. Microsoft has already told us that one of the ways Calista technology will make it into RDP will be via Hyper-V extensions that essentially provide virtual GPUs to guest VMs. And Citrix's HDX 3D leverages CUDA-enabled host-side Nvidia GPUs to provide specialized capabilities for encoding 3D graphics.

And then there's VMware's software-only implementation of Teradici's PC-over-IP protocol. Today's software PCoIP is only two-tier (just like traditional ICA and RDP), but that's really because VMware needed to get something out the door pretty fast. I wouldn't be surprised if we also saw some kind of ESX-based processing capability exposed to their VMs to really accelerate what they could do with PCoIP and View.

What impact could an open source SPICE have?

There are two possible things that could happen from SPICE being open source.

First, the actual SPICE protocol itself could get better which would lead to more support for Red Hat Enterprise Desktop Virtualization. This is a no-brainer and something I'm sure will happen. Will this lead to more people buying Red Hat? Probably not, because I don't think people choose desktop virtualization platforms based on protocol anymore.

Second, and the question that's on everyone's mind, is whether SPICE will make it into other desktop virtualization products out there, and if so, whether it will matter.

To answer that, it's important to first keep in mind that as of today, SPICE can only connect to remote hosts running on KVM-based hypervisors. I guess the idea is that it's now open source, that will change, but remember that since the hypervisor also has to provide a virtual graphics device to the VM, this isn't as simple as popping a SPICE agent in a View or XenDesktop VM. Using SPICE on a Xen, Hyper-V, or ESX-based VM will require additions to the hypervisor. When will we see those? (Or will we ever?) Who will make them and why?

So let's think about the vendors who could do this. Citrix doesn't need to, since you already get HDX/ICA with all their products, so they have no incentive to support another protocol. Microsoft already has plans for something like this with Calista, and they're probably already almost done with it, and it will probably be part of RDP, so they don't really need this. I guess that just leaves VMware, but now that they spent the money on PCoIP, I'm not sure if they have an incentive to add another protocol, especially if it means making changes to the hypervisor.

Of course we might just see a port of SPICE over to Xen-based hypervisors too, although I have no idea whether that's a simple thing or something that will require months of re-engineering?

The bottom line is that I want to love SPICE and think that it's going to be everywhere. But I think the reality now is that everyone who needs a remoting protocol has one. I'm guessing an open-source Xen-based SPICE is highly probable, along with SPICE getting better on KVM. But other than that, who knows?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.


Good article as usual. I do think that your statement "any protocol will take up as much bandwidth as it can when not capped" would be better phrased as "any protocol will take up as much bandwidth as it can use when not capped". Your statement implies that all remote display protocols will suck up all available bandwidth. This is not correct - even in uncapped environments different protocols will consume different amounts of bandwidth. For example, even in uncapped environments our Blaze RDP accelerator will cause RDP to consume less bandwidth than regular RDP (I work at Ericom).

Regarding protocols that utilize a component outside the VM vs. those that don't (three-tier vs. two-tier): running a component outside the VM is an optimization. It should be possible to implement any such protocol using two tiers (only having a component inside the VM). As a result, it should definitely be possible to implement SPICE in a hypervisor independent manner. Whether or not anybody will do this, or how such an implementation will compare to RH SPICE in terms of performance are different questions.

Also, anybody thinking about implementing SPICE must take into account that most of the SPICE code is distributed under GPL v2:


@Brian "I don't think people choose desktop virtualization platforms based on protocol anymore."

Sorry dude, I totally disagree with you on that statement. It is 100% why I pick one solution vs. the other. It is why I *** about HDX Connect and not ask for a better broker or yet more complexity in my environment. The protocol is everything and well it runs on the WAN is real world.

As for SPICE, YAWN.......... How does anybody make any money. The myth of open source is that people do it out of the goodness of this hearts. That's BS. There has to be $$$$ that somebody can really make to get the investment needed to make the protocol fully featured. Who's going to do that with MS muscle behind RDP, Citrix the clear leaders, and VMWare jumping into the game.

The Hypervisor lock-in is what this is all about. SPICE was going nowhere, RedHat is not a desktop virt company and will no invest in this space. They want KVM everywhere, and this is just a carrot to try and stimulate demand. With all the other big players this really is a big YAWN.


My first though was that they should've done it a year ago instead of waiting 15 months. But if you look, most of the protocol developments from the big guys happened before Red Hat even bought Qumranet. Microsoft acquired Calista in January of 2008, VMware and Teradici started their partnership in September of 2008, and while HDX was only announced in Feb of this year, I doubt Citrix was going to be the company to jump on SPICE.

So if not Microsoft, VMware, or Citrix, who?  We can look into the next level to find some answers.  If what Dan says is true, about being able to implement a three-tier protocol in a two-tier manner, there might be a few people that could get involved:

Symantec - Currently no protocol or hypervisor.  This could give them a solution for people that don't like RDP and don't want the added expense of Citrix's products on top of their Symantec products.

Quest - Although not likely, since they're pretty tight with Microsoft, I figured I should at least mention them.

Ericom - Making optimizations to RDP can only get you so far, and since they're not extremely close to Microsoft (and I think willing to try something new), I think there's a good chance that Ericom will at least explore this opportunity.

There's a bunch of really little companies out there that might find a way to use this as well.  HOB, Jetro, ProPalms, GraphOn.  If they can find a way to make it work, it would give them a leg up among their peers and maybe an opportunity to jump up a tier.  Sadly, a protocol doesn't fix management or scalability problems, so much more would also have to happen.


Oh, and AppDetective - KVM is already everywhere.  It's been built into the Linux kernel since early 2007..  In fact, now that SPICE is open, anyone running a Linux kernel version 2.6.20 or newer with KVM compiled in can now use SPICE.

It comes down to the management and support of the hypervisor - that's what RedHat wants.


SPICE's impact in the open source arena will definitely be positive and OSS developers will be quick to take advantage of it.  It's impact on the current proprietary VDI market leaders will be minimal in the short term.  But SPICE has the potential to to be 'everywhere'; the ubiquity of Linux on so many devices along with it's integrated KVM hypervisor and now the open sourcing of the SPICE protocol _could_ provide for entirely new use cases.

@Appdetective: don't take offense, but based on your comments over the last year or so, I don't believe you have much, if any, experience with open source solutions or knowledge concerning the benefits of the open source development model.  

For example, the "MS muscle behind RDP"; until MS bought the tiny Calista Technologies company (at most 35 employees) the RDP protocol was developed by the largest and richest software company in the world, and it was the worst protocol of the bunch.

However, you are nearly correct about one thing: Red Hat wants a form of RHEL everywhere.  KVM and SPICE are both technology enablers to expand RHEL's market share.  Red Hat may not be a 'desktop virt company' but they are true believers in KVM's ability to be the platform to virtualize any use case.  

Interestingly enough, I have never chosen a solution for any of my employers based 100% on a single component within a large and complex product.


@Brian "the worst-titled press release ever"

Couldn't agree with you more.  


@gabe. No ESX is everywhere, and KVM is not in their Kernel. KVM is an OS component so purists can argue it's not a hypervisor.

MS will gain more traction with Hyper-V and KVM will go head to head with ESX more than they will Hyper-V. Xen will be out there as a player for a long time to come. Bottom line there will be more than one hypervisor. So getting back on track it means that multi hypervisor mgmt is going to be more important, good night ESX, what is RedHat going to do here? Xen has already open sourced some of their mgmt stack and Oracle who can't be ignored uses Xen. Clearly the Citrix strategy is manage Xen and hyper-V with their essentials offerings.  

SPICE is just a gimmick. No easy money to made here to extend a commodity item, the delivery protocol, The real $$$ q43 in virtual desktop mgmt, hence why the all the startups poping up, Atlantis, Unidesk, Wanova, Moka 5, Unidesk etc. Sure they may be some uses for something very unique, but mass protocol to enable broad desktop use cases across many devices and networks built out of kindness by opensource, dream on......


@Brian - "I don't think people choose desktop virtualization platforms based on protocol anymore."

I'm with AppD here and will disagree with you.  The display protocol and the WAN is what it's all about for large scale VDI, Work from home (Wfh), DR, Pandemic.  This is the chief reason why I recommended XenDesktop STANDARD edition for customers because it satisfied the basic needs of a good remoting protocol and everything else was just icing (expensive icing that I couldn't get to stick to the cake BTW).

If you pick a crappy display remoting protocol, you severely limit your abilities.



@Brian - "I don't think people choose desktop virtualization platforms based on protocol anymore."

I agree with the previous posters, the protocol is extremely important for desktop virtualization projects. Fortunately, in America bandwidth is relatively cheap, but in other areas of the world bandwidth can be extremely expensive. SPICE may provide the richest remote access display, but the traditional remote access environment has no need for HD display, nor will these environments have the bandwidth to support such protocols.


@Rodd Linux will not get traction for Desktops, at the OS or hypervisor level. MS will not let that happen, and let anything get under them. I have my teams add value my contributing to open source, not a lot but certainly I have toyed with open source for years. Without support, open source is a slow grind. It takes ages for things to get agreed to. I want fully supported products when it comes to running their production workloads. Open source only get's me so far. I like the fact it make some things commodity, just like Xen is doing to the ESX, but I don't use Xen in production in anger because the mgmt is still catching up. There is also a massive risk if I bet my career on open source and a community that has no incentive to stay together is dies, then I am f'd and fired for being naive about open source. I also have to be careful about licensing, distribution of products I build for my external customers for profit vs. open source. It's a lot of f'ing hassle. The linux dream has been around for years. Business apps run on Windows. HTML 5, web apps etc, years out, and Windows will adapt. RedHat has no reason to fight MS in at the desktop. They are better of focusing on VMWare, get friending with MS to do that and make sure WIndows runs on KVM. We'll see how long that strategy will play our for them.



Regarding your comment about Ericom's relationship with Microsoft - we have a very good relationship with them:

* Ericom is a Microsoft Gold Certified Partner

* Our product has been certified by Microsoft for 2008 and 2008 R2 - AFAIK the only third-party Connection Broker certified for these platforms

* Our offering was the only third-party product linked to on the Microsoft Windows Server 2008 RDS homepage

* I personally participated in a Microsoft TechNet webcast promoting our joint offerings

Microsoft obviously have good relationships with several vendors in this space. We are one of these vendors.

Now back to SPICE ...


@Scott, don't get me wrong I do believe open source has value just not as much  as people hype it up to have.

Many of my friends are still roasting marshmallows and singing Linux lullabies after preaching for years how now was the time for the free Limux desktop, OpenOffice was going to kill MS office, Open source systems management tool were going to be everywhere and replace the $$$ vendors that do this today. It hasn't happened to the extent they expected, why?

Features. The open source for complex enterprises which most of them are, need advanced features. Take office for example. Macro's in Excel in Open Office vs. regular Excel just not there. That means in the real world two spreadsheet product to support. F that, $$ support costs, and at volume MS get's it cheap enough where I don't care.

I get that for home users, specific apps that consume open source components etc it makes a lot of sense and hooray for open source from that point of view. In the real world of business it's just not rich enough fast enough and there is not enough incentive to out my $$$ developers to accelerate features to then just give them away. Incremental contribution yes, but core focus for FREE F no.

Now go back to SPICE. It will not have the advanced features for many many years. There is ZERO community around it today. There is no evidence that it works well. Let RedHat have the balls to post Brian's findings from a while back. I'd love to see the bandwidth performance, huge sucking sound. By the time RedHat get their behinds in gear, a community forms and many basic features are added we'll be well into Desktop Virt. Switching costs will be high by that time. The motivated for profit guys will have jumped even further ahead with their stuff (just think about multi platform support) and SPICE will just be a better VNC/X for Linux. This is a Windows desktop world EOM.

Now that said the good news is here that this will hopefully cause some disruption. It will force the ahole XenDesktop mgmt team to wake up to the fact that that they have to allow HDX Connect, it's so obvious that they need to do this, WTF are they waiting for? If not one may just be forced to go to open source to ensure they do it as a stick and to drive prices down. That's the real value of open source to me, a stick and rattle rather than real world mass production use unless there is a matures vibrant community which with SPICE will not happen in any meaningful timeframe. YAWN.....


appdetective: How widely deployed and used is  How about Linux on the desktop?  I've been using Linux on the desktop since 1995.

Sure Microsoft has the lion's share of office sales and desktops and that will probably remain that way for some time to come.  Heck, RHEV will surely be used mainly to deploy even more Windows desktops running Microsoft Office, right?  I'm guessing we actually agree with each other quite a bit... it is just we differ in degrees.  

I'm not one of those who says that Open Source is going to put commercial products out of business.  I don't want to kill Microsoft.  I want multiple players in all markets so they offer each other competition.  Without Linux where would Windows be now?  That is a hard question because if Linux hadn't have come along many think that FreeBSD would be where Linux is now.  I think if Microsoft had a core meltdown, Linux development would slow down.

You keep saying that SPICE needs work.  For what RHEV uses it for, it is ready now or so they claim.  The SPICE will be adapted to become a general purpose remote desktop protocol separate from KVM and I don't think that is going to take as long as you think.

I too would like to see those comparison figures from the unpublished report.  I don't think there is anything fishy about the report not being published.  Qumranet probably just wanted the information for their own use... and to hand over to Red Hat.  Many companys have reports made that they have no reason to publish the results.  I challenge to do a new, independent (not funded by any of the protocol venders) report and publish the results.  Let techtarget pay them for that. :)

Anyway, I think 20-30 million Linux desktop users (the estimate I use for now) is a success.  That is a huge market for app makers... although I did hear an Adobe executive tell the audience I was in that they weren't interested in porting their desktop apps to Linux until they were sure they could sell 50 million copies.  Really?  You have to be able to sell 50 million copies of Photoshop before you can make a reasonable profit?  That's hard to believe.

For SPICE to be successful it doesn't have to crush all of the other protocols out there.  Many Open Source projects are a total flop... but there are a great number of successes.  Simply opening up SPICE doesn't mean much until a community appears that does something useful with that.  I have little doubt that will happen because us Linux users are missing such a protocol... as are any others who want an open protocol.


@Scott, it RedHat was acquired by IBM do you think the world would care as much? Nothing in life is free. If SPICE does work for Linux, get's traction etc, why wouldn't Citrix just open source ICA or at least part of it to enjoy exactly the same benefit. I know many people use Citrix for Unix and it's better than X.

Business apps for Linux are a huge question that makes my Ubuntu desktop a toy, and limited for anything real world. Much better than RedHat though :-)

Anyway we will see, I remain highly skeptical at the prospects of a commodity item becoming rich enough to serve real world business use cases. I just don't see SPICE a LINUX thing getting traction on WINDOWS desktops en mass anytime soon enough to matter. Good to see you posting here.


Brian at BriForum 2008, the presentation of SPICE used a WAN emulator, if my memory serves me correctly. However, this article says "One thing about the videos that's important to note is that our lab was a non-bandwidth constrained environment, so the bandwidth consumption data is not really valid." Could you clarify these discrepancies? Did the WAN emulator simulate other aspects of the WAN such as latency and other traffic, without limiting bandwidth?


No discrepency.. this is actually two different things. The "non-bandwidth contrained environment" was for the videos that Gabe and I shot while visiting the Qumranet office. So that video is a different  environment than the BriForum video.