We live in exciting times. In the past few years, we have seen the acceleration and emergence of a wide slew of technologies that are revolutionizing nearly all industries. From deep learning and IoT (the Internet of Things) to additive manufacturing (3D printing) and blockchain, these emerging technologies are starting to appear in products and services throughout the marketplace, creating immense value for both individuals and organizations.
In this article, we will explore how some of these technologies will improve the manufacturing process in semiconductor processing and lead to the next generation of chips that will power the future with artificially intelligent, augmented reality devices.
IoT:
The benefits of IoT for the semiconductor industry will help both boost revenues and decrease costs for innovative players. On the revenue side, as McKinsey highlighted in October 2015, the semiconductor industry is one that will benefit immensely from the advent of IoT. According to the executives McKinsey interviewed:
“The Internet of Things would significantly boost semiconductor revenues by stimulating demand for microcontrollers, sensors, connectivity, and memory. They also noted that the Internet of Things represented a growth opportunity for networks and servers, since all the new devices and services will require additional cloud infrastructure.”
On the cost side, IoT can be used in the manufacturing environment to reduce costs by enhancing automation, permitting predictive maintenance, and improving logistics. Thus, semiconductor chip manufacturers will benefit not only economically from making parts for IoT devices, but also operationally by using them to enhance production and efficiency in the plants.
Deep learning:
Deep learning, or artificial intelligence – has applications in every industry. For companies operating in the semiconductor industry, the potential benefits (as for IoT) are two-fold, both:
(i) increasing revenues and growth through the increasing demand for parts and the chips that will run neural networks
(ii) decreasing costs associated with manufacturing by using deep learning to improve the manufacturing process, automating the QA process, or even designing recipes for next generation chips.
As Dave Lamars of Semiconductor Manufacturing and Design points out:
“The design and manufacturing of advanced ICs [integrated circuits] can become more efficient by deploying neural networks trained to analyze data, though labeling and classifying that data remains a major challenge. Also, demand will be spurred by the inference engines used in smartphones, autos, drones, robots and other systems, while the processors needed to train neural networks will re-energize demand for high-performance systems.”
Mentor Graphics is using deep learning to improve models of the lithography process steps, a complex issue that senior director of engineering Rey said “is an area where deep neural networks and machine learning seem to be able to help.”
EUVL:
Extended ultraviolet lithography, or EUVL, is one of the leading technologies set to enable manufacturing of the next generation of high-end computing chips. The focus will be on two fronts, the first being advances made in the short-term that will determine how fast and effectively EUVL is used in fabs. The second front, as Vivek Bashi highlights in Solid State Technology:
“…relates to longer term progress, as EUVL is a multi-node patterning technology expected to take us to the end of Moore’s Law. Top topics are higher NA [numerical aperture] (0.55) EUVL scanner, pathways to increase power closer to 500 W and even higher, actinic patterned mask inspection (APMI) and resist performance (stochastics, LER, micro-bridging, etc.)”.
Switching to EUVL will impact the entire industry. Four of the biggest names in the space that have already announced plans to engage in this route are: GlobalFoundries, Intel, Samsung Electronics, and TSMC
According to Robert Castellano of Seeking Alpha, the EUV timelines at these 4 companies are as follows:
“Samsung Electronics has been the most aggressive in adopting EUV. In 2018, the company plans to implement EUV at the 7nm node in its foundry process.
TSMC is ramping its 7nm foundry process in 2017 using optical lithography, but will begin testing EUV in Q2 2017. TSMC will move to 5nm in 2019 utilizing EUV.
Intel will move to 7nm in 2020 and utilize EUV.
GlobalFoundries will be producing 7nm chips in 2018 and move some layers to EUV in 2020.”
The take away:
With these three transformational technologies set to enhance products and services while increasing profitability in the semiconductor industry, it is imperative for organizations competing in this marketplace to (1) understand the science behind these technologies, (2) assess the opportunities these technologies offer, and (3) develop an innovation strategy that will allow them to compete using these technologies.
