Virtualization and Cloud executives share their predictions for 2013. Read them in this VMblog.com series exclusive.
Contributed article by Amir Husain, CEO, VDIworks Inc.
2013: The Year to Make Good on VDI Promises
VDI vendors have fallen too far behind the curve
and must deliver in some important areas to maintain customer
interest in Desktop Virtualization
For the past four years the virtualization community has heralded
each successive year as the "Year of VDI" (Virtual Desktop
Infrastructure). While adoption has definitely increased and many enterprises
are actively considering desktop virtualization for their environments, growth has
not quite been what many analysts and industry watchers had imagined. Many in
the industry will recall Gartner's prediction of 49 million VDI users in 2013.
Those actually operating in the field will tell you that we are way off that
number when accounting for true VDI seats. Let's not even get into the strange
number fudging and the odd "buy an app virt license but let me
call it VDI" practices being followed by a few major vendors. Suffice to say,
as an industry we are not close to 50 million honest to goodness virtual
desktop users yet!
But why have things played out this way? Was the downturn to
blame? Was it a case of unjustifiably optimistic analysts looking to make
the case for the next big market? Perhaps, and perhaps. But technology vendors are not so
easily absolved of blame. Let's face it: Desktop
virtualization infrastructure has not been able to evolve as rapidly
as it needed to. As a consequence, it has taken way too long to iron out
performance, stability and functionality kinks that most customers faced with
their desktop virtualization proof of concept deployments.
Meanwhile, the larger IT industry hasn't stood still
these past four years. While VDI customers were grappling with remote desktop
performance and trying to validate vendor claims about the density ASICs
and protocols would enable on host hardware, ARM processors were coming of age,
tablets took off and smartphones evolved into devices far more powerful than a
Thin client circa 2008. At the same time, installed capacity in the Cloud for
storage and compute (application hosting) went up dramatically, spurring
application migration to a web-enabled model.
Too many customers share their frustration at the high cost,
complexity and rigid choices VDI vendors subject them to. It is not typical for
vendors to dictate choices to customers, "You must use this protocol
to get a decent experience, but it will only work with my Zero client and my
broker!". More so than traditional desktops, VDI requires a greater
number of infrastructure components. When each vendor tries to lock the entire
stack down, the customer is left computing ridiculously high per-seat licensing
costs, using obnoxious formulas and wondering how much of his datacenter he
will have to rip out if he ever went with a different desktop strategy.
To be specific, here are just a few of the areas where
VDI vendors have fallen behind:
1) A cost effective means of
delivering a high density of users per server is not really being offered by
the leading vendors. You can run a high number of users when your graphics
requirements are limited, but when you need productivity PC-like graphics
across a couple of monitors, your user density drops tremendously. You're
supposed to buy specialized ASIC-based accelerator cards - which only
work with a particular vendor, of course - and do all other kinds of
voodoo. Frankly, it is a pain to deal with. There are no acceleration standards;
there are no industry-wide benchmarks for end-to-end graphics
performance, recognized by all vendors. There are just too many holes here.
2) There are too many options
and not enough standardization around VDI storage architectures. In order
for the market to scale past the knee of the curve, the technology needs to be
easy to provision and rapidly scalable. This is not the case today. Each
VDI deployment is treated as a bespoke deployment. Contrast this with
taking a box cutter to a box labeled Lenovo, removing the enclosed PC
and plugging it into a wall socket and hitting the
power button. PC deployed. In contrast, a lot more needs to happen before
a VDI VM comes online. Storage is a big part of this. There are literally
dozens of storage vendors each selling a different solution to the same
problem, while all the major virtualization vendors have their spin on
things which essentially amounts to not committing to anything in particular.
Does the customer have to be a rocket scientist to figure out VDI storage?
Today, it's pretty close to that.
3) VDI Management continues to
be fragmented. Are we really telling customers that they have to implement
a separate endpoint management solution provided by the Thin
client vendor, while using VM management tools from a Hypervisor vendor, a
desktop inventory and management solution from a Systems
Management vendor and storage management software from a storage
company? Not to mention brokering, load balancing, power management,
remote control and other elements that belong in the mix. Do we
expect customers to recruit brain-surgeon help desk personnel who
will go through half a dozen different console applications to figure out why
someone's VDI system is having issues? This borders on the ridiculous, but is
also an unfortunate example of vendor arrogance. We will make it easy for
ourselves, regardless of whether it is hard for the customer. This won't
work any longer.
These are just a few examples, but there many other
shortcomings which have been known for a while now and have gone
In 2013, vendors in the VDI ecosystem really need to get with the
program and deliver on the promises made over the last several years. My
prediction is this: if we fail in doing so, as an industry we will
be confronted with uniform disbelief when we once again proclaim the
coming year to be, "The Year of VDI".
About the AuthorAmir Husain is the President and CEO of VDIworks (www.vdiworks.com), an Austin, TX based developer of VDI
management software. He holds over a dozen filed and awarded patents in
Virtualization and Cloud Computing. Amir was the CTO and currently sits on the
Board of ClearCube Technology, the world's first developer of PC Blade and
Connection Brokering technology. Amir is also a Board member at Pepper.pk, the maker of 3 World #1 Mobile Applications
and Wheel InnovationZ, a Texas based stealth startup focused on mobile Cloud