Tuesday, July 22, 2008

My Business Plan Is Simple


Investment firms from around the world (including big companies like JP Morgan and others) frequently visit my blog. Look folks, my plan is simple. I want to create a line of kick-ass, energy efficient, single-core and multicore processors with a suite of free and easy-to-use, graphical dev tools for easy parallel programming. My stuff will blow everybody out of the water. It will cost you about $US 10 million and take about two years. That is all.

Related Articles:

How to Solve the Parallel Programming Crisis
Transforming the TILE64 into a Kick-Ass Parallel Machine


Anonymous said...

A "suite of free and easy-to-use, graphical dev tools for parallel programming" doesn't require any inherent costs.

Investors hear pitches all the time, and often from people that already have something to show. Nobody is going to fork out $10M to someone that promises to make "a line of kick-ass, energy efficient, single-core and multicore processors with a suite of free and easy-to-use, graphical dev tools for parallel programming" in "about two years" when there is nothing concrete except a quicksort (that seems to have taken years to write) and a chess-ai (that seems to be about as good as a randomizer). Investors are protective about their money, and are therefore pretty much immune to hot air.

I'm not trying to put you down, I'm just trying to say that you won't get any funding for as long as you don't have lots and lots of good stuff (preferably concrete simulators or even prototypes) to show the potential investors. I know it's a kinda catch-22, and that sucks, but there's just nothing one can do about that.

Louis Savain said...


I am not deluded to the point of expecting venture capitalists to invest any money in Project COSA. I am not that dumb. VCs never invest their money in a business unless they get lots of recommendations from people they can trust, i.e., their expert advisors. Even if I had a working prototype COSA processor, they still would rely on their advisors to tell them whether or not it's a safe investment. Besides, if I had enough money to develop a working CPU and the dev tools, I would simply go it alone. As an aside, this may indeed turn out to be the case. One never knows.

The VCs and investment firms who visit my blog are just on the lookout for trends in the industry. They can sense that something huge is on the horizon and they want to be as informed and prepared as possible. I have made a lot of noise and they just want to find out what's going on.

I really posted the article above for the record and for the fun of it. I like leaving a web paper trail, so to speak. My current strategy is based on three things as follows:

a) The computer industry is in a world of hurt because they have a very nasty problem on their hands with no solution in sight. Tens of billions of dollars are at stake and time is their worst enemy. They can't go on much longer.

b) Whether or not you believe it, I am 100% convinced that COSA is the solution. Time is my friend. I can wait.

c) Soon, the industry will realize that their various research efforts are not bringing any worthwhile result. Sooner or later, their desperation will be such that they'll come knocking on COSA's door. I'll be waiting.

Zach Saw said...

Have you actually patented any of these stuff?

Louis Savain said...

Have you actually patented any of these stuff?

The answer is no. Everything you read on my blog or the Project COSA pages is public domain. However, there is plenty of room for IP on the hardware side. There are several important aspects of the hardware design that I have not publicized. Still, this does not prevent anybody from using my ideas to design and implement their own proprietary COSA-compatible processors if they so desire.

Vincent SONG said...

Let's say someone give you the opportunity to work on COSA, full time. Let's say you can release a prototype of a COSA processor and a COSA tool/execution environment.
Lets say I study and practice COSA and develop a video game engine (like unreal engine 1 for instance).

How can I be sure that my engine will show linear performance scaling on future COSA processor ?
How can I be sure that my engine will be as performant as the original engine ?

Henrik14 said...

Yes, problem with investors are same as Hen and Egg problem, they require >>Proof of concept<< and one needs money to produce it....
But, Luis, I would say You overlook obvious, and this is that all can be simulated in software!
It cannot (and therefore, would not) work at full speed as it would work in hardware, but first You must prove that CONCEPT work and that it has no flaws.......
For making simulation of COSA, You just have to be able to sustain Yourself. At beginning, Your >>Graphic Programing Language<< can have just few instructions, like BASIC had at the beginning, long time ago.
But, Graphic programming can wait actually as this is superstructure and just a tool for programing.
Therefore, it could be developed separately. First prove that Your paralelization work, and has no flaws, and this You can test on any multicore multicpu computer.
Just keep in mind that it is easy to extend system on multicomputer (aka swarm) system as separate computers could be threated just like more cores which to use in paralel, that is all......
Therefore once COSA provenly work on one multicpu and multicore computer, it would work on any number of cpus and cores available.
There is no need to produce new CPUs either, and in fact, You MUST NOT do it, because You cannot expect that everybody would scrap their computers and buy Your type of computer of future.
I say it is only compiler problem to effectively implement paralel processing, but I see You have some prejudices that would prevent Your idea to be practical, as before making any program, there should be >>Requirements Analyse<< and I see that You would benefit if You analyse system You are proposing first of all, as showeling code on pile never worked and never could work.....