Archive for February, 2010

Harry’s SEO Homework

Wednesday, February 17th, 2010

William ShakespeareAs I’ve mentioned before, I live in California, the state with the 46th best elementary school system in the country. Thank you California Lottery! So keep that in mind as you read the rest of this post.

One of the more challenging homework assignments my 3rd grade daughter receives regularly is to write a short story using a list of the week’s dozen or so spelling words. For instance, this is one that she received not so long ago:

Write about a time when you worked very hard to learn something. Tell what the experience was like. Use spelling words from the list.

And the list was:

coach    blow    float    hold    sew    though

sold    soap    row    own    both    most

She wrote about the time she learned to play the piano at summer camp. I won’t embarrass her by posting the story here, but suffice it to say that it was pretty forced. Don’t even think about asking how she got the word “soap” into the story!

So, this evening, whilst walking the dog, I was listening to this week’s episode of Leo Laporte’s This Week in Tech podcast (aka TWiT). On the podcast, someone mentioned a site called Wordstream. On this site, you can enter a keyword and it will tell you the most common search terms that includes that keyword. The idea is that, if you want to increase your SEO (search engine optimization), you should use the words that are most common in searches and the search engines will send people to you.

I immediately thought of my daughter’s homework assignment. The users of this site must feel like her, trying to weave the words generated by this site into their prose. I wondered how odd that would be. So, I decided to try it, just so I could get a taste of what my daughter went through. And also, because I thought it would be kinda fun.

Being “the ASIC guy”, what word other than “ASIC” could I have entered. After entering my keyword and my email address, I received an email with the 10,000 most common search terms that include “ASIC”. I decided to focus on the top 50 search terms, separating them out into individual words and listed them on a sheet of paper.

Now, without further ado, is Harry’s SEO Homework:

 __________

The alarm rang.

I lurched up out of bed, already in a panic, staring at the clock to see what time it was.

11:00am. Damn!

I took care of the basic biological necessities, then threw on my jeans, a T-shirt, and my brand new ASIC Gel-Kayano running chaussures. At least the company I worked for didn’t have a dress code and they didn’t care what shoes I wore. Designing ASICs and FPGAs is much easier when I’m comfortable.

I had been assigned to the verification team. My job was to search for bugs and to wrestle them down. Thankfully, I was able to use Verilog and System-Verilog for this project. Not like those VLSI design days, when I, and so many of my fellow engineers, had to wear a tie to impress the boss and had to use VHDL because they made us . A language by any other name is better than VHDL. Sure, VHDL is more structured. But, Verilog is a whole lot easier to use.

I’d been searching in some DCT4 code for one particular bug that had eluded me for 15 days. It should have been implemented in analog, but some Einstein decided digital logic was easier to design, so here I was.  It was me vs the bug. And the bug was winning!

Then it hit me. I was looking at the wrong register!

I felt a surge of power as I unlocked and modified my testbench. The combination of sleeplessness and Mountain Dew made me delirious. For a moment, I thought I was wearing a women’s dress and Onitsuka ASICs while playing volleyball in a prison cell. Gotta stop hanging out with those guys from the UK who watch Monty Python all the time.

I acted quickly, changing an “lt” to a “gt“, invoking the recompile flow on the new code, and kicking off the regression sim.

The simulation worked and I breathed a sigh of relief. My boss had threatened to bring in some hotshot design services company that he’d found on a website if I couldn’t find this bug. The nimbus that had been floating over my head for weeks was gone.

Now I could keep my job.

And now it was time for the layout guys to sweat!

__________

Phew! That was a lot harder than I thought. (Especially since those ASICS running shoes get a lot more hits apparently than the ASICs I usually write about). But now that I wrote and published that story, I expect I’ll be #1 on Google Search in the morning:-)

To be fair, I think there is certainly some value in understanding how people find this blog through various search terms. It helps me to understand what kind of information they are looking for and that helps me choose better topics to write about. But, taken to the extreme, if I write content for the search engines instead of all of you (my readers), then I’m in trouble. You may find me, but you won’t like what you find. And that would be much worse.

If anyone else wants to give this a try just for grins, just go to Wordstream and try it out. Just let me know where to find your “masterpiece”.

harry the ASIC guy

So, you want to start an EDA company?

Tuesday, February 9th, 2010

www.flickr.com/photos/cayusa/ CC BY-NC 2.0Lightbulb

In the almost 2 years since I started this blog, I’ve been paying pretty close attention to the EDA industry. And one of the themes I keep hearing goes something like this:

“There’s no more innovation in EDA”
I hear it on blogs and on Twitter. I hear it from design engineers, from consultants, from old media, from new media, and even from EDA people.

One person I know, someone who has been an executive at an EDA company and a venture capitalist, says that EDA is persona non-grata for VC folks. Maybe you can start a “lifestyle company” doing EDA, but don’t expect any more companies like Synopsys to come along.

And then, about a month ago, I get an email from someone out of the blue. He’s got an idea for a new EDA tool that would transform the industry. He’s been in the semiconductor business. He’s developed EDA tools. He knows everybody there is to know. And he’s not able to get anyone’s attention. As he puts it, nobody is working on anything “disruptive”. They are all doing “incremental improvements” that are “woefully inadequate”.

I spent about an hour talking to him on the phone. As I got off the phone, I was not sure what to make of the conversation. He was either insane or a visionary. He was either deluded or optimistic. He was either obsessed or determined. I’m still not sure which.

And that is what makes this industry so much frickin’ fun! You never know. That crazy idea of turning VHDL into gate-level schematics … who figured that would be the biggest innovation in design in decades?

Then, last week, I heard about this event/gathering/workshop happening during DVCon at the San Jose Doubletree. Presented by EDA veterans Jim Hogan and Paul McLellan. It’s called “So, you want to start an EDA Company. Here’s how …” And I immediately thought of my new friend with the idea about a new EDA company. This is exactly what he was looking for … an audience of people with open minds who were asking “why not” instead of “why”.

Maybe you also have a crazy idea. Maybe it really is crazy. Or maybe not.

I invited him and I hope I can get there myself. If so, I think you might want to come too.  You might just meet the founder of the next Synopsys. Here’s the skinny: San Jose Doubletree on Feb 23 at 6:30-7:30 in the Oak Ballroom.

I’ve also written a little prediction of what I expect to hear on the Xuropa Blog. Who knows? Maybe the naysayers are right and EDA is Dead. Then again, maybe not. I, for one, am dying to find out which.

harry the ASIC guy