As with many a good idea, the idea for Nimbus really came through self preservation ....

The original two founders have been kicking around the computer field for more years than we like to admit — and frequently it was the computers kicking us around more than vice-versa.

The Situation

Anyway, we had just sold the last startup (one of those independent phone companies that started in the late '90s) and were trying to figure out what to do next.

Our wives had some sage advice — Get the hell out of the house! — and despite the best of intentions our golf games were clearly never going to rise above the level of complete embarrassment.

Since we were somewhat (completely?) idle, and knew computers, we were now fair game to become the computer expert to help out everyone in the family, friends, friends we didn't even know we had, former colleagues, former enemies, and it seemed like anyone we passed in the street!

The Problem

After a while looking at various flavors of PC disasters, a couple of things became obvious:

Most people can't manage a PC.

The modern PC is an absolute marvel.  It can browse the web, capture video, print documents, play music, work in every language known to man, store vast quantities of information, do all these things at the same time, maybe even do open heart surgery for all we know (you can probably download an app from the Web).

Good stuff, but this brings a very large downside — managing a modern PC can require a vast amount of technical expertise and a large quantity of time.  The owner of a PC should be concerned about virus scans, firewalls, backups, defragmenting, trojan horses, botnets, ghost images, security updates, application updates .... and a whole lot more.

Most PC owners don't even know what all the terms above even mean, much less have the time or inclination to deal with them.

Hey, we're terrified ourselves every time we need to install a new application on our own machines.

Many people just want the basics.

This usually consists of browsing the Web, getting and sending e-mail, looking at some pictures sent by friends, and a few other things.

Of course, a lot of people really do need to do a lot of things on their PC.  They should go buy one, and we wish them all the best.


So this leaves us with many people who just want to do a few simple things, but have to buy the whole PC with all its problems.

And it's not even a cost issue, but a complexity issue (although no one would complain about a lower price).

The Concept

The basic concept was to apply an old idea (thin clients), to a new environment (PCs and the Internet), for a new market (Consumers).

  • Deliver the screen, keyboard, and mouse to the user,
  • Run their cloud computer in a professionally managed datacenter,
  • Use the Internet to connect the user to their cloud computer,
  • Make the user equipment dead simple,
  • Make the user equipment extremely low cost,
  • Make it instant on, low power, minimal size,
  • Offer the cloud computer as a subscription to eliminate big up front costs,
  • Handle all administration (backup, virus scans, etc, etc) in the datacenter

The Journey

The first thought was that this wasn't possible — the engineering calculations of screen size, number of pixels, rate of change, Internet bandwidth, cost for the client equipment (the Nimbus unit), and a bunch of other things, just didn't seem to work.

So we spent the first year beating the problem into submission — too much data, compress it differently; unit cost too high, get rid of standard processors and use an FPGA; performance not good enough, write hardware accelerators for the hot spots.  Many pots of coffee later, we convinced ourselves it would work, and actually work pretty well.

Then we raised a bit of money and hired some engineers to help design the production Nimbus units and setup the initial datacenter.

We also realized that the basic technology was going to be useful in a lot of areas beyond the initial thought of a simple device for consumers.  This site talks about several of them.