Cadence Wins Texas Instruments' Supplier Excellence Award

I attended the online ceremony recently in which Texas Instruments (TI) formally awarded Cadence their Supplier Excellence Award. It is the first time ever that this award has gone to an EDA/IP company. The award itself is a heavy piece of glass (see the picture) which has already been shipped to San Jose. I suspect it is sitting in shipping and receiving waiting to be put somewhere that we can all see it when we eventually get to go to the office again.

I assume you know who Texas Instruments are. You probably don't know that they were founded in 1930, and took the name Texas Instruments in 1951. So obviously they predated even the invention of the transistor, let alone the invention of the integrated circuit. In fact, they had to invent the integrated circuit themselves! Jack Kilby invented the integrated circuit in 1958 at TI's Central Research Labs. Fairchild's Jean Hoerni and Robert Noyce are usually regarded as the other two of the trio that came up with the main features to enable the manufacture of planar integrated circuits, with multiple die on a wafer.

The above image summarizes Texas Instruments. Their main focus is analog, making them one of the most profitable semiconductor companies since that has high differentiation. They do over $10B in analog, another $3B in embedded, and over a billion of "other" (not sure what's in there — do they still make "Speak & Spell"s and calculators?).

TI has a set of supplier expectations that goes under the name C E T R A Q. That stands for "cost, environmental responsibility, technology, responsiveness, assurance of supply, and quality". Those are the dimensions on which they measure their suppliers and which fed into the award to Cadence.

TI described the importance of Cadence to them over the last four or five years. Of course, Cadence has been a supplier to TI for much longer than that. When I was last at Cadence twenty years ago, TI was the main driving company behind the "superchip" program that I managed, to merge the analog and digital Cadence environments. You can argue plausibly either that it took us a decade longer than originally planned, or that it is still work in progress.

• In 2016, TI needed an IP partner that could supply the full range of IP for the next-generation processor platform in 16nm technology targeting automotive quality and industrial reliability requirements.
• Cadence was chosen not just because they had mature IP and favorable business terms but because they:
  • were willing to customize deliveries to significantly minimize TI's integration effort, silicon bringup, and qualification
  • were willing to secure the IP that were missing from their portfolio to complete their offering
  • demonstrated early indications of superior program management and customer support
• Cadence teams worked seamlessly with TI's team on every aspect of design to meet requirements
• Consistently delivered to an aggressive schedule and went beyond the typical IP handoff criteria
• Demonstrated consistent responsiveness and collaboration to address the issue
• Throughout this multi-year program, Cadence has demonstrated a commitment to TI success
• Cadence is no only a strategic IP supplier but also a strategic EDA and emulation hardware supplier
  • Cadence EDA tools touch every TI silicon design, both analog and digital
  • Cadence continues to work with teams across TI to ensure our products are the best in the industry

Tom Beckley

Tom accepted the award on behalf of Cadence. He gave us a glimpse into the Tom of 1960 and the hi-tech toys he had to play with: etch-a-sketch (I had one, too, since my Dad went to the US around then and brought one back — they didn't show up in Britain until years later).

But there was a more serious point. He compared technology (beyond the robotic penguin) from 1960 to today. "You've come a long way baby". In fact, so far that phrase, which was used to advertise cigarettes to women, now seems like it is in a different language, like Chaucerian English.

Tom compared things today to back then. In between, he grew up, graduated, and went to work for Harris on rad-hard semiconductors. I'm going to interpret his 1960 as "the sixties". Chips were not tiny in 1960, they didn't exist. Gordon Moore hadn't written his famous article in Electronics yet. But for sure, our concerns have changed. If you ask anyone in the semiconductor industry what are the most important markets, they will have marginally different ideas. But we all know where future semiconductor volume will come from, at least in the short term. Longer term, who knows? Maybe soon, a Steve Jobs-type will get on a stage somewhere, produce something out of his pocket, and change the world again.

But in the meantime, we have the almost unlimited future of 5G, specialized SoCs, AI, data centers, and parallel algorithms, increased integration into full flow, and, of course, deep learning. Deep learning has transformed everything in less than a decade, and is only getting started.

Sign up for Sunday Brunch, the weekly Breakfast Bytes email.