SEMICON Day 1: IoT Everywhere…and China

The first day of SEMICON West for me is usually imec day. They have a full day of presentations. But I saw the most interesting ones in Brussels recently so I decided to go to the SEMI/Gartner day instead. They run in parallel ballrooms in the Marriot on 4th street.

To me there were two big themes. One was the unavoidable internet of things (IoT) although mostly wearables and automotive. Ian Ferguson of ARM touched on a zillion things in his whirlwind tour of the future (including his personal favorites to watch which are computer vision, medical and agriculture, all of which he thinks technology will change a lot).

Michele Reitz of Gartner took a stab at wearables to answer two big questions:

• What is the semiconductor revenue potential in wearables through 2019?
• Which semiconductor vendors are gaining share?

They had access to lots of teardowns and analyzed different market segments, so it was a rather more informed guess than just throwing out some number like 50 billion units.

So let’s start with the good news. The market should grow to 500M units by 2019, 2.5 times what it was in 2014.

But unfortunately there is bad news too: wearables are a high volume low profit opportunity. By 2019 the semiconductor revenue will grow to $4.2B by 2019 (but remember that is for 500 units so each one costs less than $10 on average). But $4.2 billion sounds a lot but is actually less than 1% of semiconductor revenue.

And which companies are benefiting? Freescale (soon to be acquired by NXP), TI, STMicroelectronics, Atmel, Intel. Intel is giving the keynote tomorrow morning so we might find out more then. But then, remember this is Intel, so we might not find out much.

The second big theme was China. Allen Lu had a great overview. China is an enormous semiconductor market. Because they manufacture so much electronics but don’t build that many chips, they import a lot. In fact, I was surprised to know, semiconductor imports are larger than oil imports. Of course they re-export quite a bit but they keep a lot too since they are also the largest market for mobile, for example, in units at least).

China has strategically decided that they want to build their own semiconductor industry and are making a lot of money available and putting a lot of pressure on making partnerships with global experts such as Intel and Qualcomm to make that happen.

There are lots of challenges and lots of opportunities. First, Allen’s challenges:

• Global industry consolidation creates higher entry-barrier for new-comers (such as Chinese companies)
• China semiconductor industry started late, lack of scale & core competencies
• Advanced IC manufacturing requires huge funding
• Technology, talent, market, supply-chain all global for Semiconductor industry
• Chinese companies’ capability (or not) to go global (technology, marketing, business practice, decision process, …)
• Reliance on government initiatives and funding hinders competitiveness
• Need to establish solid business model & core competencies, avoid competing on price only

And the opportunities:

• Government resolve and resources to scale up China’s IC manufacturing
• Semiconductor industry entering mature phase with narrower margin. Electronic device
  manufacturing continue to transfer to Asia from US, Japan and Europe

• Semiconductor no longer favored by investors in US. Almost NO US VC funding to equipment and
  materials companies in the last 10 years

• China’s economy scaled up: capital and talent no longer the “show-stoppers”.
• Electronic product eco-system established in China: products increasingly being defined and
  developed in China (especially in mobile and consumer sectors)

• Culture, language, eco-systems, of the application market increasingly replace cost and become
  new advantages

• Abundant engineer & science graduates coming to industry
• Cost advantage still exists

When I was in Brussels recently they announced a joint venture with SMIC (majority owner), imec, Huawei, Qualcomm to collaborate on developing a 14nm process by 2020. There is other joint stuff going on.

And then, before we’d even go to the end of Allen’s Q&A, an audience member announced that Tsinghua Unigroup made a bid for Micron. If Allen had wanted a better way to convince any skeptics that the Chinese market is big, can’t be ignored, and the Chinese government is serious, then it would be hard to think of one.