You know how you can think of 100 things to talk to someone about the next time you see them, but when the time comes you forget to bring up every single one of them? That’s how it’s been between Teradici and I the past few times I’ve run into CEO Dan Cordingly or CTO Randy Groves. I talked to both last week, and I learned about Teradici’s new Alta zero client, their Cloud Access Software, and I even asked about their thoughts on VMware and Blast 2.0.
Alta zero clients
Teradici has been supplying the PCoIP protocol technology for thin clients for years, and a few months ago they announced a new hardware platform for zero clients called Alta. Alta has built in WiFi, can drive dual 4K displays, and supports USB 3.0. They recently demonstrated Alta for the first time at AWS Re:Invent, and you should start seeing the availability of Alta-based thin clients in Q1 2017.
What’s interesting is that Teradici is no longer making the chips themselves for these zero clients. Instead, they’ve leased the IP Block to chipmaker Socionext, which is using it to build out their own System-on-Chip that will then be sold to partners like Dell, HP, Samsung, and others.
The virtues of zero clients are well-known, so we don’t need to get into all that, but you can expect to see this latest generation of PCoIP-based zero clients in a number of new form factors, like all-in-one monitors and laptops. Previous PCoIP-based clients were pretty much limited to your typical tiny desktop form factor.
Even though partners are making the SOC and thin clients, the hardware will universally support Teradici’s management console, which will let you update firmware, set policies, and manage user permissions for devices based on the Alta chip, no matter which vendor you use. This is important because, according to Teradici, about half of all PCoIP users are on thin clients.
Teradici’s software platform
If you’ve paid attention to Teradici for as long as we have, you’ll no doubt remember their RDSH-based product Arch. Arch wasn’t really a standalone product because it required you to have the Horizon View connection broker, but it added functionality that was still a few years out for VMware. It was more or less put on pause while Teradici helped AWS create Workspaces (which uses PCoIP as its protocol).
While Teradici was helping AWS, they were also building a standalone cloud platform called the Pervasive Computing Platform that had open APIs and SDKs that people could use to integrate into whatever they happened to be doing with remote desktops. This was a good plan, but the product was more of a framework, and implementing it meant that you had to do a lot of work on your own, not to mention that you had to have some other platform that you could integrate the Pervasive Computing Platform into. Customers wanted something they could use easily by setting it up at a cloud vendor rather than building their own cloud.
This year, the next evolution of Teradici’s software platform was released. Called Cloud Access Software and Cloud Access Platform, the combination allows you to create and access remote desktops on the cloud. It’s not really about traditional desktop virtualization, though. Instead, it’s intended for high-end use cases that need to leverage the compute power of a desktop that resides on essentially unlimited resources in the cloud, like engineering or graphics rendering. It’s also not limited to GPU-intensive workloads (in fact, it works with both GPU and non-GPU cloud instances).
More information on Cloud Access Software can be found on their website.
PCoIP vs Blast
I simply had to ask what they thought about VMware releasing Blast in spite of the longtime relationship between VMware and Teradici. The answer was as polite as you might expect, something along the lines of “VMware is welcome to do whatever they feel is necessary to support their customers,” but it’s not hard to read between the lines. Teradici believes that PCoIP is better in the vast majority of situations compared to Blast 2.0.
To be clear, that’s not my interpretation of the conversation. All you have to do is look at Teradici’s blog and you’ll see that there is a difference of opinions. There are posts extoling the virtues of lossless graphics alongside others showing how PCoIP outperforms H.264 both visually and from a security perspective.
I’ve yet to see independent testing done between the two (probably because one of the main guys that did the independent testing works for VMware now…thanks, Shawn), but I’m certain there are situations that call for H.264-based protocols and others that require PCoIP. From what I can gather from people that have used both, typical office workloads still perform better with PCoIP, though for many users Blast provides an experience that is good enough in most use cases.
Either way, we know which side of the argument Teradici is on.