NT: Re: I don't believe we'll be looking at a R/S, and Brad has recently addressed this issue. Any significant financing will most likely be long-term debt rather than an equity transaction. And, movie revenue (other than the "modest" upfront money) will be coming in this Summer.
Certainly the movie will help the situation. Without knowing how much the deal is for, and knowing that Cono takes his cut first, and then Top Secret takes their half, and let's not forget Ray's cut, it's hard to say how much real revenue will be realized. I hope it is a lot, though. As for long-term debt, I would like to see that, but it hasn't happened so far. I am not sure what would change in the future that would initiate the ability to obtain long-term financing when they have not been able to so far. I hope it happens, but I am not optimistic.
I believe that the ASIC truly can be produced in 2-3 months when utilizing one of the overseas fab companies. (Tom Cooper recently did some preliminary work on NV's behalf while in Asia.)
I think that it is not likely. Semiconductor design and manufacture is a *very* drawn out process. It takes a minimum of 6 weeks to go from wafer start to packaged product. So, let's say that they can map the FPGA code to an ASIC design in 2 weeks. That would be unbelievably fast. Then from the design, the Fab would have to make the masks. Let's say another phenomenally fast 2 weeks for that. I will even assume that tooling takes place while the masks are being put together, so there will be no time allotted for that. Throw in 6 weeks for a flawless first spin, and after a total of 2-1/2 months, you have A0 silicon. That does not mean that you have a product. In semiconductor manufacture, there are *always* problems with the first spin. There are speedpath errors, hot spots, logic flaws, and a myriad of other possible problems. Shoot, even just developing a testing protocol could take several weeks. So, you take your A0 silicon, test the crap out of it, find the flaws, redo the masks, and do another spin. Remember, each spin takes a minimum of 6 weeks from wafer start to packaged product. Oh, and 6 weeks is world-class. I doubt that the fabs-for-rent in Singapore can do it that fast. If you are *really* lucky, you get everything right in the second spin. So, assuming that you spend a month testing and retooling (another phenomenal achievement), and another 6 weeks in production, you now have invested 5 months total, and you still haven't tested the second spin.
Now, here's the kicker. You test spin 2, and everything is great. Guess what you have to do now? That's right. You have to start a production spin. That means 6 more weeks before *any* packaged product comes out the other end. Before you can release the product, you must have enough volume for release. At the end of the 6 weeks, there will be some number of thousands of chips coming off of the process every week. It will take some number of additional weeks, therefore, to accumulate enough volume for release.
Now, you could always assume that you got everything right in the second stepping, and begin volume production then, while hoping against hope that you really *did* get it right. But, remember that you will not be able to test for 6 weeks. And during that 6 weeks you start new wafers every day. So, if you still have problems, you will have to discard the first few tens of thousands of chips coming off of the line, while you retool for another stepping. That won't be cheap. Sure, Intel does it, but they have literally billions to spend on R&D. NVEI does not. It would be foolish to blow the whole wad in a bad stepping when they could wait two more months and make sure they got it right.
I sincerely hope that nobody in the company is telling you that they can get an ASIC into production in 2-3 months. This is something that I know more than a little about. If the company is telling you this, then they are blowing smoke and would have lost every bit of credibility in my eyes.
In addition, it is almost a certainty that we will have a contract in hand at the time of beginning the ASIC process.
That would be nice, and it would be very possible if it were a licensing contract. It would be less possible if it were a product contract. Semiconductor manufacturing is so frought with pitfalls and delays that a company would likely be very reluctant to give money up front for a future product. Even Intel can only sell product once it is on the shelf.
A more probable outcome is that a company like - hmmm let's see, I used Alcatel in the last one, let's use Intel this time - Intel contracts to license Embarq. They pay megabucks for the license, and contract to pay a royalty on each chip they sell. Then they design and fab their version and call it "Embartium". All NVEI has to do is rake in the money and quickly develop Embarq II.
I'm optimistic that this will occur by Fall of this year.
Your optimism is admirable. I just don't see any possible way that they could have an ASIC product in volume production by fall. I frankly don't think that it is physically possible. According to the RAQ, work on the ASIC will not begin until the FPGA is in the hands of potential customers, and they have provided some feedback. We have no idea when that will be. To have even the slightest outside chance of a product this fall, they would need to begin work on the ASIC *now*, and they would have to get it perfect on the first spin. That just doesn't happen.
Once again, don't take my comments as negative. I am just trying to be realistic. I am still very hopeful about my NVEI investment, but I am working hard not to endulge in false hope.
AI
