Asked and Answered: A Blog is a Two-Way Street
Gene Frantz
TI Principal Fellow and Business Development Manager, DSP
You know a blog is working when you receive well-thought-out responses that move the discussion forward. It turns out that my ruminations on architecture versus application stimulated just such a reaction. I love it when people agree with me, but sometimes it’s more fun when people disagree – certainly it makes for more lively conversation. Let’s take a look at what a few people had to say:
Alan Cook wrote: “Technology perhaps doesn’t drive innovation, it is more my experience that innovation often is delayed until technology catches up. As you rightly point out large companies cannot possibly cater to the needs of small companies, and often these small companies are the originators of innovation, but are stymied because the custom low cost hardware they need isn’t available. Remember also that innovation isn’t always a large scale thing, the MSP430, for example has been responsible for some innovative designs. In fact I had to sit on a design for several years until the right combination of processing grunt, low power consumption, size, cost and peripheral structures finally arrived in the general marketplace with the MSP430. The next problem, of course, becomes keeping up with the pace of innovation generated by these new parts, which naturally leads to the requirement for even more of something, be it clock speed, peripherals, less power, or something else that is novel, such as hybridized or asynchronous cores.”
I believe you’ve got it, Allan. I agree that in many cases innovation leads technology. Not too long ago, I contacted one of our business teams in Japan about trying something new. Their response was that it was impossible. So, we traveled to Japan to talk to them. At the meeting they put on a whole presentation explaining why it was impossible. I stopped them on the first slide and said, “I know it is impossible. If it were possible it would have already been done.” I now use two definitions of impossible: 1) It can’t be done 2) It hasn’t been done yet.
The second definition happens when innovation leads technology. The experts can always tell you why it is impossible. The innovator tells you why it hasn’t been done yet. So, keep innovating and keep pushing us until we catch the vision. Once convinced, we do know how to move fast (spoken on the part of all large corporations).
He Yun Xi wrote: “I think application and architecture are mutual effected. Sometimes the application drives the architecture. Sometimes the architecture promote application, but for me the most is how easier to approach the usage of DSP, more guide to develop, edit, debug, I hope a new way to emerge.”
I agree. I call the first (applications driving architecture) the “technology pull effect” and the latter the “technology push effect.” But, I am on the fence about making it easier. On the one hand, the individual who seems to make the huge breakthroughs is the one who understands the technology in great depth and takes advantage of the unexpected. On the other hand, making the mundane hard stuff go away allows the innovator to go to the next level of innovation. Just last week we had a brainstorming session that demonstrates this. There is a new sophisticated technology emerging. But, the purpose of the session was not to understand the new technology or to improve on it. The purpose was to talk about how to use it, given that it exists. This was an example of technology push. I could certainly give examples in the other direction, as you could.
Jonas Wrote: “Which is easier and faster to deal with, for example in term of DSP processor, I do not really care whether is it new DSP processor or old DSP processor, but definetely I prefer newer DSP processor if the migration path is not difficult and the cost to own it is reasonable. It also because the newer version of DSP processor has more processing power and speed and most important fully supported by TI. So at the end I agree with that it has mutual effect and technology does not driven innovation, because there are lots of ways to come out with innovative products you still can use transistors to develope innovative products but perhaps you can’t develop high end product based only on transistor. innovative product with latest technology is always the best as long as the product has a price that acceptable to end user.”
You’re right in the fact that what you want is a solution that is cost effective and timely. You are also right that you don’t need to depend on the latest architectural innovation from us, or any other company. I guess where I would stray from your thought process is in the fact that the lonely transistor is a significant part of technology (I have a very broad definition of technology). So, I would state it slightly differently: “If no one has an implementation that can be useful for me, I can use technology and create my own.” Does this make sense? Have I understood your comment?
Please keep your questions and comments coming. We try to answer them all in a timely manner and if I don’t receive them, I have the sensation that I am talking to myself – which, at times, is the best conversation I have all day, but defeats the purpose of a blog.