Technology industry – Broken rhythm

Reply to this post

The breaking of Moore’s Law will make software even more important.

• Intel’s significant change in the rhythm of its die shrinks is the biggest indication yet that Moore’s Law is finally beginning to break down.
• Since before almost any of us can remember, the technology industry has been governed by the concept of Moore’s Law.
• This states that the number of transistors inside an integrated circuit will double every two years or so.
• This has had massive repercussions because in practice it means that computing power has doubled every 2 years while the cost has stayed the same.
• It is this law that has underpinned the breakneck pace of innovation in the hardware industry and has allowed many companies to compete by producing cutting edge hardware.
• In its latest 10K filing, Intel has stated that its Tick (process) – Tock (architecture) rhythm before moving to another geometry will now have an extra stage (optimisation) added making three stages rather than two.
• In practice this means that Intel will spend longer on the 14nm and below geometries with three product generations per geometry rather than two.
• I suspect that it has done this for reasons:
  • First. Each geometry migration is meaningfully more difficult than the last, requiring more time to get yields to the point where they are commercially viable.
  • Second. The cost of each migration is also increasing substantially necessitating greater revenues in order to make a return on the investment made.
  • Third. The industry is rapidly approaching a time where smartphones and tablets have enough computing power meaning that technical specifications will no longer carry a high price premium.
• For Intel, this represents a huge risk.
• This is because it has always competed on its technical prowess in being able to make transistors so small that its devices could always outperform those of the competition.
• By lengthening the amount of time it spends on a particular geometry, Intel will give the competition more time to catch up and offer an equivalent product at a much lower price.
• However, these days how the silicon is implemented can be more important when it comes to performance than the exact specifications of the silicon itself.
• Consequently, it looks like Intel is intending to compete along these lines more in the future rather than marching as quickly as it can to smaller and smaller geometries.
• For the rest of the technology industry it is an indication that the speed of commoditisation as going to accelerate even more than it has done already.
• As smartphones and tablets become powerful enough to do almost anything that the user is likely to expect, the value attributed to adding even more power on top will vanish.
• This will mean that differentiation will move even more into the functionality and into the software that sits on top of the hardware.
• This will be true for all devices not just smartphones and tablets but TVs, home appliances, consoles, wearables and so on as they become relevant to the Digital Life and Digital Work ecosystems.
• The ecosystem is the glue that holds all of a user’s Digital Life together in a seamless, easy to use and fun way.
• It is here where differentiation will occur and where the winners of the technology industry will earn their profits.
• It comes as no surprise to me that the ecosystem companies: Apple, Microsoft, Google, Baidu, Alibaba, Tencent, Amazon and Facebook have a greater market capitalisation than almost all of the others combined.
• There are of course exceptions to this which tend to be companies that have developed a sustainable differentiation in hardware.
• Examples of this are ARM which owns the core processor technology for mobile devices, Qualcomm which has an edge in cutting edge cellular radio and MediaTek through its low cost production and media expertise.
• For everyone else stuck in the middle, the choices are to either eke out a commodity existence (like Samsung does very well) or try and differentiate in the ecosystem.
• It is the companies that make this bold choice that are likely to outperform the most but at the same time they also carry the greatest risk.