Collins Labs Live

MPC8641D Product Summary Page

Oh.  Yes.  Must. Find. Eval. Board.

Finally, they have updated their website, and user registration and login now works.

The T18 Compiler/Simulator has been announced as being available on the night of Friday, 22 September.

Keep your fingers crossed.

There’s a company named Intellasys who has a very interesting-sounding product coming up soon. It’s called the SEAforth24 processor.

The idea is that it’s a highly distributed microprocessor architecture; 24 18-bit processor cores on the same die, communicating via shared registers. It sounds intriguing, particularly for certain embedded applications (that’s the target application).

However, on the Intellasys website is an announcement of the availability of tools (i.e., a compiler, simulator, and probably more complete documentation of the chip itself), targeted for June 30, 2006.

Well, June 30 has come and gone. And there are still no tools or other further information on the Intellasys website. I’m bummed. I wanted to play with their simulator, because I’ve got an application in mind that their processor might be well-suited for.

Tech Workers of the World Unite?

The following post (link above) pretty well articulates my thoughts on unions.

If there ever was an organization dedicated to mediocrity, impeding productivity and forcing people to be on strike and not earning money when they want to, a union would be it.

I’ve been a programmer for over ten years now. Keep your skills fresh, work hard, be a team player, and you’ll tend to get further work (if one job dries up, someone you know and have impressed will pipeline you into another) and plan for contingencies like being out of work for a while. It’s a nice indoor job with good benefits (I’m a contractor but I have been an employee enough times to know that) and good pay rates. Sure, you might get outsourced - that just means what you were doing is something somebody else *should* be doing since they can do it cheaper. Get into some part of the industry that is new and not likely to flow to parts of the world with poor infrastructure, language barriers, or non-existent IP laws. Or get into Defense or Security work, those won’t likely offshore anytime soon.

In short, stop crying and start working towards the future you want. High-tech is still one of the best ways to get there for the middle class guy. Sure, the rich get richer, but if anyone can tell me when this wasn’t the case, I’d be glad to cut the legs out from under them. Yes, you work hard. But if you enjoy the job, that’s actually not a bad thing.

And if you don’t like the field, get out. If you don’t like your employer, move on. If you don’t like the work, retrain, expend some of your resources readying yourself for something you do like.

It seems to me the article’s poster expects the world owes him/her something. Get over yourself, I say. The world owes you nothing, isn’t fair, and a Union won’t do anything but take your money, impose restrictions that hamper the hard workers and the competent, and drive the work away faster. Oh, and add to that sometimes pull you out of work when you don’t want to go. And consume your union dues along the way (like all bureaucracies).

Unions… no thanks. I’m doing just fine without them. The only people who need unions are lazy folks, people without foresight, or people without initiative. Do yourself a favour and go out and take the world on and beat it into the shape you want, don’t wait for someone to fix it for you.

DRC Computing has created a really cool product. The idea is to place a re-configurable element (essentially just a big FPGA) directly onto a hypertransport bus. What they’ve done is take a Virtex FPGA, solder it onto a custom board, and fit it to an open Opteron socket. The result? A sweet way to connect an FPGA to an opteron system and have the FPGA live directly on the Hypertransport bus, able to fully utilize the 12.8 Gb backplane.

I have decided recently that trying to innovate and come up with good products and actually implement them is more than twice as difficult than doing it with the help of one other person.

I’ve had two or three conversations with people recently that presented opportunities to find solutions to problems (that’s what Collins Labs Research is about, btw).  As one person with limited resources, there’s no way I could, for example, viably compete for a bid to create an electronic records system for a university.  It’s not because of a lack of reputation or whatever; it’s because I simply wouldn’t be able to do it alone.  It would take me years to do it, and the end result would be nowhere near as good as it could be if I had a team of 2-3 people to work on it with me.  Not to mention that such a team could finish such a project in months, rather than years (I’m talking ballpark scale here).
This argument can be made ad infinitum with regards to almost any project of interest.

Until I find a “few good men”, as it were, I suppose I’ll have to press on solo.

I have had recent exposure to a company, called Arbonne, that uses a network marketing strategy to sell their products (skin care, makeup, etc).  The idea is that they will sell you their products.  If you then sell, via referral, their products to someone else, you get a percentage of the retail value of that sale.  If they do the same, then they get a percentage and you get a percentage.  This continues ad nauseum.

The result is that you have a ton of non-professional salespeople running around selling your product like crazy, but not requiring massive commissions or management overhead.  You don’t have to worry about competitiveness with people you originally sold to because what they sell benefits you almost the same as if you’d sold it yourself.
What if semi-shrink wrapped software could be sold this way?  It would have it’s own twists, obviously.  Perhaps you would come up with a solution for a particular problem for a customer, but try to make it sufficiently general that you could make modifications and apply the same software to other customers’ problems, and then share some royalty revenue with the customer who originally funded the work; thus, shrink-wrapped custom software solutions.  The trick is getting funding for the very first go-around.

I think that some of my research is going to head down the path of DSP and acoustics.  There are some neat applications in that area that could use a lot of work.  I’m not an audiophile, and this will not be an exercise in my becoming one, but there are some embedded applications just waiting to be developed that could take advantage of audio signals.

Just think for a minute about how your brain processes information.  You have your 5 senses.  I suspect for most people, as it is with me, the most important of those senses is sight.  I also suspect that for even more people, the sense of hearing is second or higher in importance.  Audio signals help us keep our balance.  It helps us know our surroundings.  It provides the medium for our primary means of communication (voice), albeit with all the electronic forms of communication, that may be changing (or at least the gap may be shrinking).  Much capability comes from our brain’s ability to process audio signals in a sohpisticated way.  One mechanism I’d like to explore is how people can tell that they are being addressed when someone else speaks.  There is still room for research in the area of voice recognition, but there is lots of existing research and a number of pretty effective implementations.

I’ve decided that it would be appropriate to separate my personal blogging from blogging of a more technical and professional nature. Of course, there will still be some overlap. I also thought it would be neat to have a sort of journal for Collins Labs starting from before there really was a Collins Labs (because there isn’t, really).