Okay, so I've got one from my own past. So ride with me here for a few minutes. So from 1990 to 1995, six years, maybe just before 1990, might have been 89. I worked for a company, a little company in Minneapolis called Artist Graphics. And in 1990, competition grew from what we call two guys in a garage, I'm missing something already. Artist Graphics made add-in display cards that accelerated AutoCAD. So if you just took a dumb EGA card and put it in your PC and you ran AutoCAD, and it's a 3D drawing program, okay? And you changed your view, AutoCAD would go draw, in software, pixel at a time. So it had a built-in driver and it knew had to draw lines and fill and various things. But they had an API in place that enabled third parties, like Artists Graphics, to build an add-in display card and dramatically increase the redraw speed significantly. And so we built a board based on the TI34020 with our software and our printed circuit board design. And we started getting competition from these outfits that we called 2 guys in a garage, because these were just popping up overnight. And they introduced these add-in display controller cards for AutoCAD. Which were basically not doing anything at all, other than building TI's reference design and shipping TI software driver with the product. But the point was that they were selling them from 300 to $500, right around in that range. We were selling ours for $4,000. What do you think happened to our sales? Start going down. So my colleague Tom Becklund and I, old friend of mine. We worked prior to this together at mainframes. We convinced the owner that we could and should build our own graphics controller chip that would outperform the TI34020. And to his credit, he believed in the two of us, we could build a chip from scratch. A blank piece of paper, and he funded the project. We designed the first graphic chip, it was known as the GPX, but the manuals in for these things. Well, you know is really real, I really did this years ago. From a blank piece of paper, we did it in 18 months, we were working 12 hour days, to get this chip done. It had a whole bunch of bugs in it, because we did our own verification and design engineer's a big, or it could be started on verification. [LAUGH] The design engineer that designs a piece of hardware should not be the person that does the verification that's the worst possible person to design it as yourself, or to verify it as yourself. because you know too much about it and you will not test certain cases that need to be tested. So I have a reputation team that can do that for you, and you want them to break your design, so you'll find all your bugs. But it had many bugs and we were able to make this product be the flagship product of the company. None of the bugs were catastrophic. In firmware, figure out a way in the driver to work around all the hardware bugs in this first chip. And there was this big drawing called Shoot88F. I went and looked for it, I couldn't find it. I even have a picture, that's too old, but it was a subdivision drawing. And it took the TI34020, 24 seconds to render this image. So it does lots and lots and lots of little lines, and all this information. And when we first brought up a TI34020, it was the fastest. This drawing had never been drawn in 24 seconds before, but it had been drawn in 50 or 100 seconds or 150 seconds or something. It's a big drawing. So the 34020 can do it in 24 seconds. So our first graphics chip rendered it in just over 4 seconds, that was a 6 times performance increase. So they were really excited at this point, it's like wow, okay, we're going to design the next one now. And we did, so then we designed a second chip called the 3GA. And that one rendered this drawing in seven-tenths of a second. From measure from the 34020, 34 times reduction in speed, so this is productivity for draftsman and engineers. Using AutoCAD, and they're working on architecture, a building or whatever it is a bridge maybe perhaps. And they want to change their perspective and once they change their perspective, they want that drawing to be redrawn very very quickly, so they would enable this. And in the beginning of 1995, we began working on the third graphics chip at this point in time. So while work was ongoing on the third chip, I stood back and I looked at the Intel PC space, and where graphics were likely to migrate over time. And my conclusions were that they could stay as add-in cards. They could migrate to the motherboard as a solder-down component, possibly sharing the CPUs DRAM. And the thrid place, it could end up is it would migrate into the Intel AMD chip sets, possibly again, sharing the CPU's main DRAM for the buffer to hold the image to be displayed. And that did come to pass. [LAUGH] So the owner ran the company as an S-corporation. It's much like a sole proprietorship where the owner gets personal tax benefits. He had this vision, and this was the downfall, I think. He had this vision of running a family owned chip business, okay? He wanted to sell graphics chips to other companies that then would build the display controllers. And he was never ever able to let go of this notion of running this family owned business. He wanted to leave this legacy for his kids, and he could just absolutely not let go of that yet. He wanted to be a big graphics chip provider at the same time. So I had a talk with the owner and I explained where I saw graphics migrating, and I encouraged him to think about a couple of things. Artists Graphics at that time was a very widely recognized brand name. Everybody in the graphics business at that time knew the name Artist Graphics and our big competitor was Matrox located up in, I think it was Toronto, Canada. So Artist Graphics and Matrox, I mean, we were always going head to head in competition. But my idea was to use our widely recognized brand name, at that time, to develop business relationships with Intel and AMD and other leading motherboard manufacturers. Get our graphics technology into these chips and onto the motherboard. And convert to a C-corporation with a board of directors. So like a real corporation, okay? And he would be chairman of the board and he should hire, my recommendation to him was to hire a Silicon Valley chip savvy CEO. Our revenue at that time was $10 million annual, and he foresaw the revenue growth to $100 million. And I said, Bob, you don't know enough about this business, there are people out in the Bay Area that are experts in this, so think about this, all right? I think it's really important. If you want to get to that $100 million revenue mark, this is what needs to happen. And this model, we have for expensive adding graphic cards is going to go away, okay? It's going to change dramatically. So he thought about it for some time, and then he got back to me. And he walks up to me and he says, I thought about that conversation the other day. And I remember, he put his hand on my shoulder and he patted me and said, glad to see you're always thinking. Truth, he just turned and he just walked away. And he never took any action, it never ever came up again. [LAUGH] I was a bit baffled by that. So in 1995, it was a new round of start up companies, I remember one of them was called S3. I don't recall what happened to them, remember, there were others as well. The window of opportunity was diminishing quickly. To create a strategic business alliance with AMD and Intel and motherboard manufactures as was windows is closing up. So I told him that, Bob, I guess I kind of ticked off. I said, if you don't heed my advice, you're going to end up with owning a 100% or nothing [LAUGH] and he didn't care for that very much. I said it nice, I said, I think you're making a mistake here. And anyways, I left the company at the end of 1995 and moved my family from Minnesota out here. And a year or two later, Artist Graphics closed its doors. It was failure, well [INAUDIBLE] What lessons can be learned here? This is the part where we talk. Anything else? >> The importance of a good verification plan. >> [LAUGH] Yes. [LAUGH] We did much better on the second chip. We had more engineers working on it. I still look back on that and think, my gosh, two guys with a blank piece of paper in 18 months and we pulled that off. That was a really stressful period in my life. I just was married at the time and I had one small child and another one on the way. And the wife was not happy that I was down at the LSI Logic Office doing chip design for 12 hours and an hour commute in each direction. So I was gone from the house every day for 18 months for for 14 hours and she didn't care for that. But the pay was good and we were successful. And these products became the flagship products of the company for some period of time. Yeah, verification. I look at it and we'll see here at the end of this slide deck as it was a failure to pivot and I'll talk more about pivoting when we get there.
