Citrix announces XenDesktop 4 details, including bundling with XenApp

Awesome stuff! The final strike to the competition will be XenClient.

This will greatly reduce the amount of confusion in the market. As you pointed out last week there were too many complications in the previous model – what we see here should simplify things going forward. We should all benefit from not having to explain which capabilities are included in which product.

From AppSense’s perspective it is also good to be the only recommended personalization solution too.

Martin Ingram (AppSense)

Is that per "named user" or just "per user" ? Named user suggest to maintain a list of authorized user that "per utilisateur" have not (token FIFO) ???

@Kata

No, it is not named user. There is no need to keep a user list. Instead, the FAQ reads that they want you to just cover your entire AD tree. This is exactly what I was talking about in my post - the change is user lisc will be a dramatic drag on Citrix in what otherwise would have been a great announcement. I think sales will take a near term hit while clients (and their sales folks telling the story) get used to this new model....or better yet, they realize it's a bad model and make another announcement to fix.

I really hope Citrix does offer a concurrent user license for XenDesktop to the education market, because if they don't, they may have just priced it way, way out of reach for most computer lab / training lab scenarios.

I manage about a thousand PCs for various computer labs on campus.  We've been working up a plan to rollout XD for our labs, with the thought of using it at first primarily for OS Streaming, then expanding at some point to hosted virtual desktops in order to allow students to access our software remotely.

Before this change, we'd be looking at at a pilot of 50-100 licenses, then 300 or so per year until we hit 1000, then we'd likely expand some after that to allow for hosted desktops.

But we have over 28,000 unique users per calendar year in our labs, so per named user licensing would be insane for us.

We were looking at XD Advanced (which doesn't seem to exist anymore) at what, $200 or so retail, although our price would be cheaper.  1000 x $200 = $200,000 investment.

Now that would be 28,000 x $225 (retail cost for Enterprise?) = $6,300,000.  

Yeah, not gonna happen.

Before this switch, it seemed like XD was a way to jump more fully into the Citrix way of doing things at a cheaper cost than going the XenApp route, mainly since we could transfer our Ghost image way of doing things to a XenDesktop streamed image.  Now, I don't know what to think.

Citrix denmark told mé thar the Price for converting from xenapp platinum to the new xendesktop platinum (per user) is 135$

The changing of the license model will be interesting for my organization.  We currently have XenApp Platinum and use it for our remote users.  We have close to 10k named remote users monthly of which no more than 1500 are concurrent.  We are also looking at expanding use of XenDesktop.  Management  has always felt that their tab to Citrix has been too high and the conversion from concurrent to essentially an enterprise license for up to 12k users is a daunting proposition.  Maybe now is a good time to retire before I have to tell them we need to spend even more money?

"Citrix denmark told mé thar the Price for converting from xenapp platinum to the new xendesktop platinum (per user) is 135$"

Say what?!!!  I paid more for XenApp Platinum than I did XenDesktop Platinum.  I would expect a conversion of no less than the 2-1 same for same as mentioned in the article

As I re-read this over and over, I'm feeling seriously screwed here.  First they want me to pay potentially $75 for my XenApp platinum to trade up for XenDesktop?  I paid $375ea GSA for my XenApp Platinum licenses and $275 for XenDesktop Platinum.  So as far as I am concerned, the conversion, plus a fee, plus change in license model does not seem balanced to say the least.

When MS bought SoftGrid, they switched to per user and we got something like a 4-1 conversion with no fees, and that was a same for same product.

Pricing will be USD $75 per named user for the Standard Edition, $225 for Enterprise, and $350 for Platinum, retail.

I thought that when you have a per user pricing model, that the price is significantly less than CCU pricing.  More like $75 for XD Platinum, retail is more realistic.

XenApp Platinum for 3000 CCU is 375 GSA or 1.13m

XenDesktop Platinum for 3000 CCU is 275 GSA or 825k

Now under XenDesktop4 per user with 15k users, is 5.2m

Converting 3000 XA Platinum is going to cost me 225k and yield me 6000 per named users.  I then need upward of another 4000 licenses minimum, or 3.15m (at 350ea retail) just to keep the same access level I have now.  Yes this adds XD to my functionality, but Citrix is the one changing the model on me, so the way I see it, its going to cost 3.5m to remain in the game in the new model.

Since XenApp has historically been more expensive than XenApp <---- ha someone made a mistake.

I have been listening to people complain about citrix pricing for many years... and yet they still use the product. Amazing how that works :)

"and yet they still use the product." <-- as opposed to what other product that does 'same for same'?  And was around 5-8 years ago or longer when organizations commited to the SBC model?

Basically Citrix has now given you an architecture that makes delivering the applications in whatever format suites yuor environmnet. This makes it a nice solution to design around, as opposed to a bunch of point products. They have a huge lead technology-wise, now let's see how they do with that lead. Adoption will depend on whether they understand that issue this licesne model poses to both their installed base as well as potential new deals. The clients will be faced with truly evaluating the value of what this solution offers, versus cobbling together point solutions and hoping for the best.

Unless they fix the lisc issue fast, I see near-term (ie this quarter and next) getting hit pretty dramatically. Time will tell whether we look back on this announcement with amazement for being great, or a flop. Either way, it will be a fun quarter. Less rules...more fun.

This is a good announcement from a product direction standpoint and as Brian points out what many of us have been hammering on Citrix about a product portfolio that differed from thier marketing direction on a single managed virtual infrastructure.

That being said the pricing announcement and the lack of an announcement on when the client hypervisor will be released was disappointing.  We all really need to know for sure what functionality will be included in the initial release of the client hypervisor and when.

The biggest upset however is the licensing change.  Citrix is carefully trying to market this by providing this 2 for 1 switch out.  This may work in some vertical industries, however I see push back in the very near future.  In the Healthcare industry (as I'm sure in Banking, Retail, Manufacturing, etc) we have way more people than we do devices that we manage.  We have over 40K users and under 16k devices to manage with the user count growing and the device counts remaining about the same.  I see a big pricing problem where there in industries where there is a 3 to 1 or 4 to 1 person to device ratio.

Try a 28 to 1 ratio.  :)

tjbruni13 is right.  Government organizations like mine where there are a ton of shift workers sharing devices will have to shell out a fortune for this new model. For new customers, this is great.  They pay for XenDesktop and they get XenApp for free. What a deal.

The one area that I raised to Citrix as being a huge problem for the new licensing model is work from home users.  I have a client that has 14k employees across 22k devices.  I was planning on trying to get them to buy 6k XenDesktop licenses to cover both Work from Home usage as well as pandemic scenarios.  Now I'll have to sell them on 14k licenses and they just aren't going to bite on that.  This is a severe issue and I really wished they had solicited customer, partner and CTP feedback on this because I really think it's a big mistake in the near term.

Shawn

To add to my previous comment, they've added a ton of great technology to XD4 and I'm thrilled that they are moving forward with one product license.  However, the implementation is going to leave a lot of my customers to keep their XenApp as CCU license and to skip out on any upgrades to XD4 or expansion of XD usage.

Shawn

We use XenApp for internal and remote access.  If I understand this to use XenApp it has to be from a XenDesktop.  This seem like a lot more overhead.  Now the user that had outlook, has to connect to XenDesktop to get Xenapp outlook.  I think we are going to stay with XenApp and now with VM hosted apps I don't see the need for 90% of the users to have a desktop.

Nick

Shawn brought up an interesting point...Pandemic Planning...something I was thinking about leveraging XenDesktop for....I have 3000 users, maybe 500 concurrent in the case of a pandemic outbreak but it would be shift based...Now I need 3000 XenDesktop licenses? That's insane.

@Brian,

Thanks for setting me straight.  I wanted to make sure there wasn't a catch to the XenApp.

So the only thing that could be a problem is the concurrent users vs. named users.

@Nick

Also worth noting...you can still purchase XenApp by itself using the concurrent user model. So if you just want to use server-based computing (which includes streamed apps to clients), then you can stay with XA concurrently. The more things change, the more they stay the same....at least in some clients.

I do think this will cannibalize top line revenue in the short term, which will also negatively impact stock price (once the analysts figure out what just happened to them), but long term ( a quarter or two out) the ship will right itself. I guess if you're gonna take a hit, this would be the year to do it, right?

This has some settling to go yet !!!

Not sure why Citrix wants to adopt a license model that has come under extreme criticism through other vendors. I honestly thought that people found the concurency model a good/fair and equitable way to do it.

I disagree with Brian when he says it "makes more sense".

"In-line with Microsoft VECD"..... why in the hell would you want to "line-up" with that ???? It would be great if MSFT lined up with a concurrency model !!

I am really hoping the dual hoping ICA issues are fixed by this whole new "XenApp fully supported on XenDesktop" If that is not the case, I don’t see how both can fully co-exist.  Cheers to all the virtual geeks out there.

Will this drive more peope to XenApp as a better solution overall?  I mean If I have to buy PER USER @ $350 list for PLAT plus VECD, How close does that get me to XA pricing and the ability to buy less to cover concurrent?  If capex costs were an issue for hosted desktop solutions already, what does this do?

@Brian,

How is per-named user in line with Microsoft?  Desktop OS are licensed per Device meaning many people can connect to a single computer over the coarse of a day.  Same with VECD,  VECD is licensed per device w/ SA not named user.

@Mike Stanely,

Citrix has already agreed to make an exception for education. Specifically students but not faculty.

Joe

Wow, I've been out of the loop for a few days and finally a single product, which is really really really really good. I've been thinking about all the license concerns and here is my view.

- If you want all VDI in your environment and have lot's of devices then this is perfect. I expect for a lot of the large banks per user to make a lot of sense because most of them get that this is the desktop of tomorrow.

- Lot's of crying about CCU, but I wonder if it's as bad as people think. Firstly if a user is using it as a Desktop, I'd expect them to want access most of the time  from any device, especially during a disaster.

- If CCU really does matter, shift workers etc then I see some options.

1) Use XenApp for desktops deal with app compat and push MS to make 2008 R1 better, I don't see rapid adoption of R2 for a some time.

2) Delude yourself and try crappy View. I'd go Provision or Ericom (try the blaze stuff)

3) F it and just stick to apps. community.citrix.com/.../Mending+broken+hearts+with+XenDesktop+4 the Citrix blogs comments section certainly have some people who think this.

4) Talk to your Sales guy and get a deal.

@Shawn if I were in your shows, I'd be asking for a volume deal to get close to the price, but upsell the value of having 14,000 users license if things really hit the fan. I am sure if Swine flu broke out, they would that ability, which is worth some premium above the 6K users. I won't argue with you on why not XA for BCP, that's a very specific client call.

That said do I think XD needs a CCU option? Maybe, but I question how they figured the 2:1 upgrade. Looks like somebody reached into their behind and pulled it out. Really, 2:1 is their believe on concurrency. I can certainly see a balance between wind bags trying to get something for free and saying I have 100 users per server and a more real number like 5-10 which Citrix will never do to protect prices.  

Let me ask this question. If XD CCU with XA bundled was $500-600 per CCU who would rush out and buy that today? New customers I think would not get it, but existing customers would. Citrix's problem then would be the old it's $$$$ issue, which may not matter. I also wonder how the license server can enforce per user? Perhaps why they want you to cover all users :-)

Anyway, seems like this spoiled a really good announcement which we should not loose site of, so congrats there to Citrix, and there seems to be a glimmer of hope that things may change for some. That said, unless they offer features like HDX Connect, and other advanced features, it's kind of moot, So far the XD product team have been too stupid to do anything about it, and their broker REALLY sucks. Looks like they just did a broader cluster F with licenses without talking to customers who are impacted. A pattern with the XD product team. Maybe I will hate them as much as the MDOP team. I hope not, I would hate to see Citrix let a team with crappy understanding of our market screw up the business.

@appdetective Great points.  So far  the only real god fit for the PER USER I have discovered is the case where all users are on no matter what,  Think Point of Sale where there are "defined" users that log in to the desktop and never log out from store open to store close then never use it "off-hours".  

I agree on the XenApp Desktops especially since 2008 "looks like" Vista.  Would you then suggest using VM Hosted apps to get the users the apps that will not run or stream to the server desktop?  Is it ready for prime time?

@AppDetective

- Lot's of crying about CCU, but I wonder if it's as bad as people think. Firstly if a user is using it as a Desktop, I'd expect them to want access most of the time  from any device, especially during a disaster. --

What you would expect for your environment versus what is the reality for other diverse environments, particularly those designed around the CCU model, is quite a different story.  If you service users of all types externally, ie mobile, teleworkers, panflu, etc. who are not always connected continuously, then CCU is the better choice.  

My 150 CCU XD project just went to hell in a hand basket, as I have 15k users, any of which could connect to this XD project environment at any time.  So what originally was $42k to start the project, could cost me 5m?!

So where does that leave us XenApp users who may want to move over? could we upgrade the licenses then continue to server xenapp to some and xendesktop to others during the migration?

When we change the licenses to XenDesktop will they continue to work for standalone XenApp use?

I agree with those who say per-user pricing is a bad idea. It is especially bad in the SMB arena and for those who have been using a 'Standard' version of PS 4/4.5 or XA 5.x.

This must be fixed somehow.

Thank you, tom

@Byran I think VM Hosted is early, but a good step. Quest does a better job, but then it's the HDX/EOP thing. I'd bet HDX any day up to a point. I think that model is fine, desktops and apps on 2008 (guess what MS has to make 2008 x32 R1 tier 1) for those who want CCU. x64 is a stretch I guess for the bulk of people.

@Lee. Yes cry babies, but I hear your pain. The real point is that per-user is the future of apps and desktops, makes the most sense in a virtual/cloud world. There are many many devices and will grow and that's a model that will cost you more in the end. MS screws us with this today. VECD should be per-user also. Where I empathize with you is on the CCU 2:1 conversion. This F's you I get it. So tell them to give you a better deal, or stick with XA desktops and apps.  I predict Citrix will be all over special deals to drive volume to make up for revenue. This is a brilliant move by them to win mindshare and shoot VMWare in the head. If they don't bend on price then they are missing the boat. Anyway time will tell, who knows what they will do, I am sure they are thinking about it more as the impact is understood to stock prices.

As people evaluate desktop virtualization with XenDesktop, I would encourage people to start by looking at all the scenarios where per-user licensing *does* work instead of focusing on the exceptions. How many of your customers have large user bases with a 1:1 device-to-user ratio--say 1,000 users using 1,000 PCs or laptops?  When determining the cost of the solution, you can't simply compare it against XenApp, which you might be using for a particular subset of applications or users.  Rather, you should compare it to the cost of managing distributed desktops and devices, which you are deploying to practically all of your users.  How much is deploying and managing desktops costing you now in terms of software licensing costs, PC/laptop hardware costs, and support costs?  How much time and effort are you spending on imaging and managing devices, locking them down, supporting them, etc.?  How long does it take you to re-image a user's device, or to onboard a new employee, or even onboard an entirely new set of users due to an acquisition?  

Hard dollar costs as well as soft costs around time and effort should be accounted for in a thoughtful TCO and ROI analysis.  If you can re-image devices simply by rebooting them, you've reduced support costs.  If you onboard acquisitions in days instead of weeks or months, you've realized faster ROI from that acquisition.  And like Lee E. said, you should factor in the other benefits of desktop virtualization.  Being able to access your work desktop from any device or location is a significant productivity benefit and can enable initiatives like bring-your-own-computer programs and telework.  

XenApp-based desktops are an option, but no matter how much WS 2008 looks like Vista/Win7, they are still server desktops, which must be locked down.  The consumerization trends in IT are driving corporate computing towards a more personal direction.  People want their corporate computing experience to mirror their home computing experience in the same way they want to use their iPhones for work purposes.  Companies that embrace this trend will be seen as innovators that use technology to enable vs. restrict.

XenDesktop and desktop virtualization are not replacements for XenApp; rather, XenDesktop is complimented by XenApp, which is why it's included with it.  Of course, there are cases where XenApp and the CCU-based model is working fine and can continue to work fine, such as in a hospital scenario with thin-client stations outside patient rooms that are used by rotating shift nurses.  

But what about everywhere else?  And even in the hospital scenario, where can XenDesktop work?  I'm sure doctors would love to access a complete work desktop from home or from a branch site.  And if you took those doctors out of the XenApp CCU pool and converted them to the per-user XenDesktop model using the 2-for-1 trade-up promotion, you could use the additional licenses to deliver virtual desktops to the hospital administration staff, which use PCs instead of thin clients.  If you have any licenses leftover, you can use them to deliver desktops to the fresh group of interns who split their time between the hospital and university.

The point is, if you start by thinking about where XenDesktop and the per-user model *will* work, you will start to think more creatively about other use cases and opportunities for the solution.

I should correct myself in that if you want to trade up only a subset of your XenApp licenses, you will get a 1-for-1 exchange for XenDesktop licenses.  That said, the hospital scenario I described could still be compelling even for doing a 1:1 conversion for just the doctors and administration staff.  Or, if your user-to-XenApp CCU ratio is 2:1 or less, you could trade up all your XenApp licenses for twice as many XenDesktop licenses without giving it a second thought, as you can continue using your XenApp implementation as-is.  Of course, you might find the aforementioned benefits of XenDesktop so compelling that you trade up all your XenApp licenses regardless of your CCU ratio. :-)

@brad, you are right that XD 4 is being compared to traditional desktops, which is fine when you are all VDI or moving there. I'll argue it's the exception for a long time.

It also needs to be compared to the CCU model to help people with partial adoption. It is naive the more I think about it for Citrix to just assume that people will just buy into go all VDI now that XA is bundled. Not going to happen for the vast majority. People simply should use XA with CCU if they are partial adopters, unless they really need a Desktop OS which is probably a niche use case. While I agree long term per user is a good thing for the industry it's too early to force it on people. Kind of like the mind set of gang bang a broker on the whole world when it's not needed. I think the XD team is pretty arrogant in what they do. I believe they are all CA based from what I can tell, perhaps that's the issue, new guys trying to make a mark and don't get the business. East coast vs. West coast and guess who's screwing up!!

Brad I also take with the casual nature that vendors are throwing out the enable install what ever the F you want mind share to look like a hera engineer. CRAP. Nothing to do with consumers, it's a pipe dream and a mess. Self Service is something Citrix talks about which I agree with, not creating an uncontrolled security mess. Single image means locked down and more secure and cheaper to mange. The flexibility to self sever on top is good, but the whole notion of user installed crap is just dumb and junior IT practice.

My initial research shows that Citrix apparently makes NO allowance for companies with PS 4.0 Standard to use this tradeup. I have asked for more information.

Thank you, Tom

If there are already going to make exceptions for one industry, why would they not make it available for all?  What ever they think a CCU is worth put a price on it.  Let customers decide if it is worth the cost.  More importantly, let customers use the product the way it make sense for them.  

Just my 2c

@Tom

Not true. We can convert all the way back to MetaFrame XP, with or without SA. The programs of course differ based on the level of Mf/CPS/XA you own, along with whether or not you have your SA current.

Drop me a line and I'll make sure you get the info you need.

-Robert

rmorris@agsi.us

PS. For those in the SouthEast - I have a team of technical presenters lined up, from multiple mfgs, to do a whole VDI -v- Desktop Deliver/Virtualization thing (which includes XD4 and it's lisc). www.advantec.us/events-dtv.htm

It's invitation only - drop me an email if you want to get an invite.