The future of productivity growth: what awaits us in the next few years

It was already mentioned earlier what problems the further decrease in technical process faced, that the productivity growth slowed down, and neither multicore nor manycore solutions are a long-term salvation from the problems that have arisen. But the semiconductor industry does not stand still, but tries to solve these problems, both in the short term (~ 5 years) and in the long term. This time the story will be about the nearest plans.

ITRS (International Technology Roadmap for Semiconductors) short-term development plans focus around an approach called “More-than-Moore”. Its goal is to apply the same principles to analog circuits, which for decades have allowed us to reduce the size of digital devices. And then integrate analog devices as part of SoC (System on Chip) or SiP (System in Package). This is done to increase the energy efficiency and capabilities of the system as a whole, increase the complexity of the devices (and therefore answer the question of where to put the available transistors :)) and designate development plans for the technologies associated with this process. Heterogeneous integration is expected to be a major development factor in many areas, from data transfer to entertainment.



Instead of focusing on the CPU alone, as a component that enhances the capabilities of the system and gives it new areas for application, More-than-Moore stresses the importance of the integration and effectiveness of each component. If More-Moore (the traditional approach to creating a more powerful processor, using the achievements of the process technology) refers to the ability to perform computations, then More-than-Moore is associated with the possibilities of interaction with the environment and users.
')


The original More-than-Moore approach says nothing about converting analog circuits to digital ones. But the article presented by Intel as part of ISSCC 2012 demonstrates otherwise: Existing wireless modules for mobile devices perform digital data transfer, but use a lot of analog components. Analog transistors are more difficult to miniaturize, and therefore it makes sense to create a purely digital radio transmitter. It is also argued that it is possible to effectively protect the CPU and radio from the mutual influence that previously prevented the integration of these devices.


NTV-processor, which runs Windows and Linux from a small solar battery.

One of the most remarkable trends in modern semiconductor research is how the search for energy efficient solutions completely replaced the race for the clock frequency. Now success is measured in how long the device has been in a state of minimum power consumption and how quickly it transitions from it to operating mode.


Intel research in digital radio and NTV (Near Threshold Voltage) electronics, 3D integration , etc. - examples of how the focus of development has shifted. However, this applies not only to mobile devices. The supercomputer K (in reality - “Kei” - 京 - “10 quadrillion” in Japanese), created by Fujitsu and Riken was also designed with energy efficiency as one of the main factors. In particular, the supply voltage of each individual processor is selected according to current requirements, which allows saving 1 MW of power and ~ $ 1 million per year.

Again, perhaps, this approach Fujitsu will never be directly used in PC or mobile devices, but it shows how manufacturers are sometimes looking for the most exotic ways to reduce energy consumption.



Such developments and initiatives are partly visible now and will dominate the next few years. According to the forecasts of the DARPA Exascale Challenge, the performance of supercomputers by 2015 would have to overcome the level of 1 exaflop with power consumption up to 20MW. But instead, due to a number of reasons, development slowed down. And the current performance figures are 3 orders of magnitude behind.

On the long-term plans associated with productivity growth, mainly related to the search for alternatives to silicon, next time.