Semiconductors: the power behind AI

The advent of complex AI-driven chatbots has captured the world's attention, however, it’s not just software that’s powering the AI revolution, it’s powerful semiconductor chips.

In order to meet the booming demand for AI-capable hardware and software, industry leaders require hardware setups with tonnes of processing power. This is where silicon chips and semiconductors come into play. It’s fair to say that the future of AI depends on the speed with which chips can be developed, advanced and manufactured.

Semiconductor companies design, manufacture, and market microchips and microprocessors.?Supply chain constraints, geopolitical sales restrictions and rising interest rates have caused some turbulent times for semiconductor stocks in recent years, however, over the past few months it’s been a very different story, mostly due to AI-driven demand. During the month of May alone, the US Semiconductor Index saw a jump of 21.6%.

Nvidia (NVDA)

The company with, by far, the most significant gains has been Nvidia, which has seen its stock price soar around 175% this year. Nvidia is the global leader in visual computing technologies and the inventor of the graphic processing unit (GPU). Over the years, Nvidia’s focus has evolved from PC graphics to AI-based solutions that now support high-performance computing, gaming and virtual reality platforms. The company’s GPU success can be attributed to its parallel processing capabilities supported by thousands of computing cores, necessary to run deep-learning algorithms.

Nvidia’s rising stock price adds credence to the organisation’s claim to be the only company whose technology is capable of meeting demand from across the industry to build generative AI, systems capable of creating humanlike content. Shares in Nvidia rose 24% after its $11bn sales forecast for the three months ending in July came in more than 50% ahead of Wall Street’s previous estimates.

Nvidia added $184bn to its market capitalisation following the release of its quarterly report in May, more than the entire value of Intel or Qualcomm and the biggest one-day gain ever for a US stock, according to figures from Bloomberg.?

With a market capitalisation now over $1tn, Nvidia joins the elite group of companies; Apple, Microsoft, Alphabet, Amazon and Saudi Aramco. This is due to the growth in demand for computing power from cloud and internet companies as well as the automotive, financial services, healthcare and telecoms industries.

Described by some analysts as the ‘King of AI’, Nvidia undertakes in-depth research in nearly all applications of high-performance computing (HPC) as well as hyper-scale data mining and modelling that create the foundations for autonomous driving, protein and molecule discovery, factory design and automation, and the future of scientific discoveries in cancer, energy, agriculture, longevity, and climate.

The company is able to do this as in addition to building the hardware GPU semiconductors that make hyper-scale possible, they also build the software engines that integrate with them in a full stack of developer tools known as CUDA, (Compute Unified Device Architecture). Along with work at Alphabet (GOOGL) and Microsoft, these are the most powerful platforms in the world that can use neural networks to create machine learning (ML) reinforcement and deep learning (DL) inference.

According to Goldman Sachs analyst Toshiya Hari, although competition within the semiconductor industry is booming, they envision Nvidia maintaining its status as the industry standard for the foreseeable future given the urgency with which customers are developing and deploying increasingly complex AI models.

Advanced Micro Devices (AMD)

Directly challenging the market leader Nvidia is Advanced Micro Devices (AMD) a leading chip manufacturer providing x86 microprocessors and graphics processing units to millions of users worldwide.

The company operates across four segments - Data Center, Client, Gaming, and Embedded sectors. In addition to offering accelerated processing units (APUs) and professional-grade GPUs for data centres, AMD also provides semi-custom system-on-chip (SoC) products for embedded processors.

AMD recently unveiled its Instinct MI300 series chips for generative AI. The MI300 chip is set to power the world's fastest exascale supercomputer, expected to come online later this year.

Exascale supercomputers are the next generation of supercomputers, allowing scientists to better simulate the complex processes involved in medicine, biotechnology, advanced manufacturing, energy, material design and the physics of the universe, more quickly and with higher definition.

In 2022 the world’s first declared exascale computer, Frontier, came online at Oak Ridge National Laboratory. It will soon have competition, however, from El Capitan, housed at the Lawrence Livermore National Laboratory (LLNL), widely expected to be the world’s faster supercomputer. What both Frontier and El Capitan supercomputers have in common is that they are both powered by AMD’s silicon.?In fact, AMD is behind 121 of the top 500 world's fastest supercomputers.

Shares of AMD recently lifted after Reuters reported that Amazon’s AWS cloud-computing business was considering using the company’s new AI chips.8? If the company is able to sign a few high-profile customers such as Amazon, it could also see a boost in its financial results.?While AMD stock has doubled this year, it's still down 18% from its all-time high, in stark contrast to Nvidia stock.?However If AMD manages to gain further market share in the AI data centre space over the next few quarters, this would likely boost its stock.

Taiwan Semiconductor Manufacturing Company (TSMC) and ASML Holding (ASML)

Dutch firm ASML and Taiwan’s TSMC, two of the world’s most important semiconductor firms, have also received a share price boost on the back of Nvidia’s earnings. This can be explained by the fact that Nvidia does not manufacture its own chips - it instead relies on TSMC to manufacture its semiconductors, and TSMC relies on ASML for some of the most advanced chipmaking equipment in the world.

TSMC accounts for an estimated 90% of the world’s super-advanced computer chips, supplying tech giants including Apple and Qualcomm. TSMC operates as a contract manufacturer that produces chips designed by other companies such as Nvidia and AMD, who seemingly find it easier and more cost-efficient to outsource production to TSMC.

The main competitors for TSMC, Samsung (SSNLF) and Intel (INTC), are both integrated device manufacturers (IDMs), involved in both chip design and manufacturing, and presently do not come close to TSMC's dominance. The company has stayed ahead of Intel and Samsung over the past few years, due to its early adoption of ASML's high-end extreme ultraviolet (EUV) lithography systems, used to produce the world's smallest and densest chips. ASML is the only company in the world that can make and sell its $200 million extreme ultraviolet (EUV) lithography machine.?

Although Intel is investing heavily in new plants in both the United States and Europe, it is still lagging far behind in 5nm chip production. TSMC is underway to deliver 3nm chips in 2023 and is well-positioned on manufacturing 2nm ones. Samsung has a 16.50% share in the global?market and is also aiming to produce 3nm chips, however, they do not have the same demand as TSMC.

The realm of semiconductors extends far beyond personal computers, reaching an ever-expanding number of devices and applications. Below I've highlighted an ETF that provides exposure to the companies mentioned in this article, as well as others, set to benefit from the drive towards next-generation technology.

Please do not hesitate to contact me regarding your investment strategy or if you're interested in investing in the semiconductor industry.

Global X Semiconductor ETF

ASX: SEMI

The world's next generation of innovative technology will require semiconductors to power it.?The Global X Semiconductor ETF (SEMI) seeks to invest in companies that stand to potentially benefit from the broader adoption of tech-enabled devices that require semiconductors. This includes the development and manufacturing of semiconductors.

General Advice Warning: Morgans Financial Limited (Morgans) and its associates may hold securities in the companies/trusts mentioned herein. Unless otherwise stated any advice contained in this article is of a general nature only and has been prepared without taking into account your relevant personal circumstances. Those acting upon information contained in this article without first consulting one of Morgans investment advisers do so entirely at their own risk.

To the extent permitted by law we exclude (and where the law does not permit exclusion, limit to the extent permitted by law) all liability for any direct, indirect and consequential losses, damages and expenses incurred in any way (including but not limited to that arising from negligence), connected with any use or access to or any reliance on information contained in this email or any attachments.

Morgans Financial Limited (ABN 49 010 669 726 AFSL 235410) A Participant of ASX Group A Professional Partner of the Financial Planning Association of Australia

Please visit www.morgans.com.au to understand our products and services