The end of Moore’s Law (which famously posits that the number of transistors in a dense IC doubles about every two years, assuming an optimal price/performance ratio) has repeatedly been predicted, yet somehow it has endured for more than half a century. In fact, even Gordon Moore himself only projected the ‘law’ over a 10-year timespan when he penned his aptly titled article ‘Cramming more components onto integrated circuits’ in 1965.
As originally formulated in that article, Moore observed that “the complexity for minimum component costs has increased at a rate of roughly a factor of two per year. Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years.”
Fast forward to 2021 and the law still holds for the most part, but we’re on the cusp of reaching the physical limitations of how small a transistor can be as semiconductor manufacturing features, or process shrinks, approach just a few atoms in length. This has compelled researchers to explore new paths of innovation such as alternative materials to silicon, while the nascent field of quantum computing promises to completely upend the very essence of how computing power is measured.
That hasn’t stopped others from trying to improve further on silicon-based processes, however, and IBM is once again at the forefront of this drive. As you can read about at http://www.dataweek.co.za/13322r IBM researchers have broken their own silicon process benchmark of 5 nm set just four years ago, by developing a groundbreaking 2 nm ‘nanosheet’ technology. This could result in 45% better performance, or 75% lower energy usage, than today’s most advanced 7 nm node chips.
Don’t expect to buy a new smartphone sporting 2 nm chips anytime soon though, as there is a lesser known ‘Moore’s Second Law’ (also called Rock’s Law) which states that the capital cost of a semiconductor fabrication plant also increases exponentially over time. That means 2 nm processes won’t be commercialised until a few years from now at the soonest.
Looking beyond the headline performance figures, the engineering benefits of having a wider margin to play with in the trade-off between power consumption and processing speed cannot be overstated. For a long time, increasing processing power was the main focus of attention, but over the last few years things have shifted more towards lowering power consumption, driven in large part by the advent of the Internet of Things (IoT). Because the whole concept of the IoT is predicated on having connected devices all over the place, constantly travelling around replacing their batteries would be prohibitively time consuming, not to mention expensive.
Another key enabler of IoT evolution is 5G technology, which will enable further energy savings for deployed devices. The Texas Instruments article at http://www.dataweek.co.za/13313r serves as a valuable technical guide for preparing for the 5G world. As if the technical challenges weren’t enough to deal with already, one of the biggest issues designers have to wrap their heads around is the security aspect. ABI Research has taken a deep dive into this topic with a new market report exploring the vulnerabilities and opportunities of IoT security. You can read an article covering that report by going to www.dataweek.co.za/13291r
| Tel: | +27 11 543 5800 |
| Fax: | +27 11 787 8052 |
| Email: | malckey@technews.co.za |
| www: | www.technews.co.za |
| Articles: | More information and articles about Technews Publishing |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version