Earlier this week, I was at Dell’s analyst update on its thin client business and it got me thinking about how thin clients were supposed to replace personal computers. An experience that was instant on/off, that embodied simplicity and reliability and that not only had the performance of a PC but could be updated without any user impact was a compelling idea.
Unfortunately, the compromises were initially far too great. Today technologies like NVIDIA Grid, MU-MIMO and hosted services like Mainframe 2 have largely removed the impediments to making the change (they still lack and likely will continue to lack on airplanes), but I think the market is waiting for a trigger. Let me explain.
A couple things were amazing about thin clients: They tended to be much more reliable and they retained state. This was Scott McNealy’s favourite demo. He would take a smartcard and log into a machine and open a job, then pull the card, put it into another machine and pick up exactly where he left off. I think everyone who saw this wanted it. Problem was that the rest of the solution was no good.
Initially, thin clients had three big problems: They were expensive, their performance stunk and they weren’t backward compatible.
The real question wasn’t why they didn’t sell well. It was who would ever want to buy something like this. Over time, Oracle and Sun – who had pioneered the concept failed out of the market – and HP and Dell, pretty much took it over. They knew better what was needed and were far more realistic about setting expectations, but the market didn’t pivot, the products simply moved into areas ranging from digital signage to data entry. It wasn’t until the iPad that folks suddenly started talking about the PC being replaced by a far more limited device.
More recently NVIDIA Grid and Qualcomm’s MU-MIMO technologies came to market, which together should allow a thin client device to fully step into a PC’s role far more seamlessly. Grid provides a unique server designed from the ground up for PC-level loads at scale, and MU-MIMO provides a wired switch-like performance on a wireless network – both critical to addressing the performance problems that thin clients have traditionally had. But I think we are now waiting for someone to create a unique client device.
iPad as Appliance
I think the iPad penetration best showcased that people wanted a far more appliance-like experience. Unfortunately, Apple didn’t go far enough and the massive wave they created that had everyone arguing that the PC was dead appears to be dropping off and PCs are currently selling very well again.
The iPad showcased the potential but the product fell too far short of what people needed in a PC replacement and while 2-in-1s like Microsoft’s Surface tablets showed we could both have a PC and tablet experience in the same product they largely got there by being PCs running most software locally and not truly addressing the promise of an appliance.
The Dongle: lacking accessories
Dell and others came up with the idea of a thin-client dongle, but while the solution is very portable, a little larger than a USB stick, it isn’t complete in that you still need a monitor or TV, a keyboard and mouse. Instead of cutting back to nearly nothing you jump up to needing to carry at least two of these things and locating a monitor or TV in order to actually work. The dongle is working well for digital signage because that doesn’t have any of the problematic requirements that force the other accessories.
The smartphone is the device most of us carry today, and we can put a thin client instance on one. Smartphones, at least those in the high-end, can connect to monitors and Bluetooth mice and keyboards. In fact, with sensors, a phone could actually be the mouse or keyboard depending on its size, and could connect to a monitor or TV wired or wirelessly.
Some may recall that Apple initially seemed to be on this path, with the iPhone as its first iteration. Unfortunately, the technology wasn’t ready and neither was Apple, but I think Jobs’ idea had far more merit than it was given credit for.
I think once folks fully grasp that a thin client can effectively be anything, someone will step in and create something amazing.
Thin client lives on
I think the vision that Larry Ellison and Scott McNealy initially came up with was and still is compelling. Who wouldn’t want a PC that was as reliable as an appliance and not only booted up and retained status, but that you didn’t need to carry with you.
Over the last two decades the industry has addressed the problems with the back-end of this technology, but now we need to rethink the client and make it more complete for users who have both business and entertainment needs. Being connected to cloud services like Mainframe 2 for business and OnLive for Grid Gaming for entertainment is possible now and generally provides better separation between consumer and business use cases today.
We still need to rethink the client based on today’s capabilities and needs, and those cloud services need to step up to the full set of requirements for all of the personal and business use cases. Whether it is Google, Microsoft, Dell, Apple or someone we’ve never heard of, someone will figure this out in the next five years and I’ll bet the market flips again.
Currently I think the perfect combination will be a smartphone-like device backed up with something like Amazon Web Services providing a blend of services to both consumers and businesses, but with a hard barrier between them and a user experience that is competitive with Apple or Microsoft (or Apple and Microsoft). The result should transform the market.
I don’t know about you, but I can hardly wait.