An Intel executive recently told me that my Intel articles on SemiWiki are used to motivate employees to work hard and prove me wrong. The converse is also true. The senseless Intel fabless ecosystem bashing motivates me to continue to write so it is a win-win scenario, absolutely. In fact, I should credit Intel’s Mark Bohr for motivating me to write a book chronicling the great and powerful fabless semiconductor ecosystem, but I won’t.
At the SEMI ISS conference this week there were some excellent presentations filled with important market data. Most of which I already knew but seeing it all in one place with a historical perspective was well worth the price of admission. And networking with industry executives from around the world is priceless. I had lunch with David K. Lam, founder of famed semiconductor equipment manufacturer LAM Research. How cool is that!?!?!
Unfortunately, I do not have permission to post the slides so I will summarize the best I can:
The Concorde reference is from the keynote by Rick Wallace, President and CEO of KLA-Tencor. The point being that the Concorde failed not because of technology, it failed because of economics and lack of competition, adding that Moore’s law is much more likely to die in the board room than the manufacturing floor. This is absolutely true.
In my biased opinion, that is Intel’s problem exactly. The laws of physics will not defeat Intel, the natural laws of economics and inflated egos from lack of competition will. BK (Intel’s CEO) said it all with his latest quote, “We will not get into a price war with TSMC”. News flash: TSMC is not your competitor, Samsung is and they are going to eat your economic lunch, absolutely. Just ask Apple.
Bill McClean, President of IC Insights, had the most informative slides which I will use to support my anti Intel bias:
World Wide Semiconductor Sales in 2013:
It is interesting to note that TSMC revenue surpasses Intel if you do an apple to apple comparison using the final market value of the chips TSMC manufactures. That would be the price TSMC customers sell their chips for.
Top 10 IC Foundries in 2013:
TSMC again posted double digit gains as did Samsung. SMIC is the real winner in 2013 with a 28% gain. I would attribute this to home court advantage in the China semiconductor market. Notice the 4.6x gap between #1 and #2 and the 39X gap between #1 and #10. TSMC’s lead will continue to grow in 2014, definitely.
Top 10 CAPEX Spenders in 2013:
Yes, Samsung and TSMC both outspent Intel. Just wait until you see the capacity numbers and you will know why.
Top 10 IC Wafer Capacity Leaders in 2013:
The majority of Samsung’s capacity is memory. Intel also has memory in there so if you only look at logic capacity (SoCs) TSMC is the clear leader by a very large margin.
The question I most commonly get asked (other than people asking for my autograph): Is Intel serious about the foundry business? My answer unfortunately has not changed in the last year. No, Intel is not serious about the foundry business. It is a head fake to appease investors and make waves in the industry, just my opinion of course. Let’s hope the Intel Foundry team proves me wrong.