• “Intel was frankly falling behind… we saw this as a huge problem and began to sound the alarm.” – Secretary of Commerce (Obama Admin) Penny Pritzker (July 2023)
• “Intel’s Stumble is Very Bad for America.” – Sen. Marco Rubio (Aug 2020)

[Note to self: I should have written this column a year ago, when the story became more or less obvious – it would have made me look a lot smarter now…😏]

[Note to the reader: This is a lengthy analysis – anyone who is impatient can skip to the last few paragraphs for the summary.]

Intel is one of the most important companies in America. They invented the microprocessor. They laid the foundation for the digital era, reaching now into practically every corner of the economy. They enabled Microsoft, Google, Facebook and other digital giants to become what they are today. The Internet and the “Cloud” still run mainly on Intel hardware.

Yet, sometime around 2018 Intel is said to have “stumbled.”

• “Intel’s continuing problems manufacturing its next generation of chips are opening the door for rivals to threaten its near-monopoly in the personal-computer and server markets.” – The Wall Street Journal (2018)

Soon, it was seen as more than a stumble. Much more.

• “Intel ‘Stunning Failure’ Heralds End of Era for U.S. Chip Sector. World’s largest chipmaker considers outsourcing manufacturing. Crucial technology expertise is shifting overseas…. [this] endangers one of the last bastions of U.S. technology leadership.” – Bloomberg (2020)

Revenue did continue to grow for a while, and the financial market was complacent. The company plowed through a couple of down quarters, but its market capitalization reached an all-time high of almost $300 Bn in January 2020. It even weathered the pandemic and reached nearly the same level again in April 2021.

Then, last year, on July 28 2022, Intel released its 2^nd quarter results.

• “This is an ugly, ugly print…much worse than we thought…nothing to like here… worst I’ve ever seen in 15 years…maybe this was the kitchen sink, but it also feels like it could be going down the drain…” – Bernstein analyst Stacy Rasgon, on CNBC

The ugly fundamentals all pointed in one direction.

• a 22% decline in top-line revenue –
• a $454 million bottom-line loss for the quarter (compared to $5 Bn profit for the same quarter in 2021)
• A collapsing gross margin – from 50.4% to 36.5% (compared to Intel’s traditional margins of 55-60%)

Analysts were… disappointed. Deeply disappointed. Worse – they were embarrassed. Intel’s results were so far below expectations that it made them look bad, as though they didn’t have a good grasp on the situation.

• Revenue came in 15% “below expectations”
• Adjusted earnings were 60% “below expectations”
• Forward guidance for the next quarter was also cut by 60%

Now, the market was aggrieved. The stock fell 10% the next day – and continued to slide. In the next two months, shares were down 40% from July, and 65% from its April 2021 high.

It was value destruction on a massive scale – in 18 months the company’s market capitalization lost $176 Bn overall.

Things got much worse. The 1^st quarter of 2023 was a disaster. Quarterly revenue fell below $12 Bn, the lowest since 2010. The company lost almost $3 Bn in that quarter and posted the first 12-month loss in decades. There were large lay-offs and executive pay-cuts. The dividend was slashed by 2/3^rds.

The world’s strongest and most valuable semiconductor company had been shattered. Nvidia, with less than half the revenue, and half the profit, was worth seven times more – becoming the first trillion dollar semiconductor company. Intel now lagged even its perennial ex-also-ran “remora” – AMD.

Analysts’ sentiment turned ferociously negative: “dismal results” - “dark financial times” – “atrocious” – “bleak” – “the company’s plans proved to be a fantasy” – “astonishingly bad even vs. low expectations” – “we’re questioning Intel’s ability to deliver” –

• “We have written the phrase ‘Worst earnings report in our history of covering this company’ on more than one occasion over the last couple of years. But this time we REALLY mean it…” – Market Watch (Jan 28 2023)

And as often, the denunciations in the Twitter-verse were most colorful:

Terrible fundamentals, against a backdrop of high expectations; blisteringly negative sentiment; outraged shareholders; business obituaries from “the experts”; lots of detailed dissections of the semiconductor market and the company’s product offerings, supporting generally pessimistic conclusions…

Oddly, however… the share price seemed to stiffen. The stock began to creep sideways, and even to climb back up from the very bottom.

And then exactly one year after the dismal/disastrous/ugly earnings release of July 2022, came the July 2023 earnings release – and the surprises were on the upside. Analysts had expected another losing quarter, but the company announced a $1.5 Bn profit, “exceeding the high end of the company’s guidance,” and decisively beating Wall Street’s earnings-per-share estimates. The shares soared 8% the next day.

In short, it began to look like Intel might be gelling into a “value play.”

Some Thoughts on the Nature of “Value”

Before reviewing the case for Intel as a “value” stock, it is important to understand what “value” means.

It’s a Human Enterprise, Not a Statistical Fluke

Value investing focuses on finding and investing in companies that are felt to be currently underpriced, under-valued, by the stock market. Buying under-valued shares then involves waiting for the market to “correct” to the proper value. At least that is how it is generally conceived.

There are various ways to think about “value” investing. Many of them are rather mechanical. They focus on statistical phenomena (“regression to the mean” or “mean reversion” – the tendency of trends to reach a turning point and change direction), or on investor psychology (e.g., “behavioral finance” explanations, citing quirks in human decision-making, like over-reaction to bad news). But the “value” scenario is really about what the company does when it stumbles or finds itself in trouble, and/or what it can do in light of constraints and opportunities in its environment.

In short, I don’t see “value” as a bet on some external force, or some statistical mechanism, that operates on the share price in some impersonal way.

Instead, “value” is what happens when the human beings that are paid to steer the company and fix its problems succeed to some degree in doing so. The possibility/probability of that “success” is what the value investor has to assess.

This implies that a true “value stock” is not really under-valued. It is probably more or less correctly valued, in most cases. (Certainly, looking at the collapse of Intel’s fundamentals in 2022 and early 2023, it would be hard to argue that the true value of the company was not impaired or that the shares were substantially mispriced.)

A “value” company has had a set-back. It needs to respond, with a change of strategy, or focus, or operational correction. It may have been wounded, and will need to heal and recover. When (if) it does reset and recover, its value will increase. The issue is not whether its true value was somehow obscured, or that investors were grossly and collectively mistaken in their assessment of that value. It is about whether the company can in fact correct the problem, whether a wounded company can heal, or whether it will be permanently impaired. And whether the broader environment is favorable. Strategy, and environment – the focus on those two questions is what distinguishes a value investor.

Timing and the Nature of Due Diligence for “Value”

A second comment about “value” is that it involves a disconnect between the company’s short-term challenges and its long-term prospects.

The stock market reacts more to the present situation, the breaking news, and the next few months or quarters, especially when the situation is adverse. Most investors discount or ignore the future beyond that horizon.

The classic “value” investment is different: it involves a bet on the long-term.

This has two implications. The first is well-known: a “value” investor must be prepared to wait for the “long-term” to unfold. Benjamin Graham’s original formulation of the “value” thesis cited this delay factor:

• “The interval required for a substantial under-evaluation to correct itself averages approximately 1.5 to 2.5 years.” – Benjamin Graham (1949)

More accurate research has confirmed that “value” stocks typically undergo the looked-for reversal and recovery after 18-36 months. Some studies have pushed that horizon out even further:

• “Over a prolonged period of time (at least five years), stock prices revert to the mean, with low-P/E stocks outperforming and the high-P/E stocks underperforming the market.”

This uncertain delay, waiting for the business correction to develop and for the financial markets to recognize it, is what makes value investing difficult to sustain, both psychologically (fighting the herd instinct to go with “winners” rather than taking a chance on companies in trouble), and pragmatically (for investment managers who need to show performance on a shorter cycle – quarterly or annually).

The second implication of the timing disconnect is more subtle. It bears on the type of “due diligence” that a value investor ought to undertake.

Most investors and many analysts, when they confront a company in trouble, put a lot of effort into getting a clear picture of things as they stand right now. They dive deep into the mass of detail describing the current situation in the company’s operations, its markets, its products, its customers, its technology strengths and weaknesses, its competition. They dissect the company’s earnings releases and study the executives’ phraseology during the earnings calls. They talk to customers, employees and ex-employees, and consult with industry “experts.” They consume vast quantities of “data” from a wide range of sources.

From a “value” perspective, I believe, all this is largely misguided.

For one thing, all of this detail is more or less cast in the present tense (to include the very accurate past, a year or so at most). This may be helpful for thinking about the company’s short-term performance over the next quarter or two (which is all that many investors are concerned about). Beyond that, however, it has been shown to be largely useless – especially for troubled companies that are the potential “value plays.” Troubled companies are the ones that are not just cruising along. They are actively responding to their difficulties, adapting and implementing corrections, which will lead to deviations from the prior trend. Even for companies that may not be under such pressure, the value of diving into the complexities of the Here-And-Now is evidently quite limited, at least for purposes of forecasting a company’s future performance. One study of analysts’ estimates of corporate earnings-per-share (the simplest investment metric there is) found that the accuracy of EPS forecasts deteriorates very rapidly. After 12 months they are no more accurate than projections based on a “random walk” time series – a proxy for an un-informed extrapolation of the past trend.

The question for a “value” investor is: What is the situation for this company going to look like in 2 to 5 years? Will the future evolve to favor (or not) the “success” scenario? It is well beyond the horizon of traditional financial models.

The answer won’t emerge from “the details.” It will appear, if at all, in big broad strokes. And it won’t be simple to detect clearly. Details are actually much easier to deal with, because they are concrete and increasingly available and abundant (think Bloomberg). Projecting the Big Picture and its interaction with new company strategies several years out is harder, or at least very different, because it requires a different kind of knowledge, and a very different kind of due diligence. Instead of diving into the briar patch of the company’s present day difficulties, “value” oriented research should focus on the large-scale strategic factors that could create the opportunity for the company to recover from its current difficulties. This has little to do with quarterly operating margins and inventory levels. It is to be found instead in broad industry trends, the evolution of the macro-economy, the study of demographics, social movements, politics, geopolitics, technology trends, a knowledge of government fiscal, monetary and industrial policies… not to speak of history, psychology, art, literature, culture…

The Case for Intel

Intel offers a good illustration of how this thinking works.

First of all, a basic screen for a low price/earnings ratio flashes “value” – Intel is the most depressed stock in the sector.

[Note: The P/E for Intel in this chart is the long-term average, since at the moment the ratio is technically unavailable because of the losses in the previous quarters.]

At the same time, its 2022 earnings were strong, and, as noted, the latest quarter also showed a profit. In fact, quite a bit more of it than the company’s highflying peers. This is the second facet of a potential “value” signal.

The “mainstream” approach to investing and the “value” approach both focus on a company’s fundamentals. Both aim to assess the true earning potential of the company’s business by looking at the substantive characteristics of its operations, products, markets, competitors, etc. But they have different views of what is really fundamental, which lead to different approaches to due diligence.

The mainstream model calls for the analyst to get down deep into the weeds. For Intel, analysts study the specifics of “technology nodes,” debating the merits of 7 nanometers vs 5 nanometers etc in chip fabrication, pondering whether and why Intel has “fallen behind” its competitors. They apply esoteric technical benchmarks to compare Intel’s product offerings against AMD’s or Nvidia’s. They agonize over the company’s market share in various segments: servers for data centers, personal computers, mobile devices, artificial intelligence engines… etc. They wrestle with the significance of quarterly shifts in gross margin or operating earnings, “seasonality” and “cyclicality.” They consult all the reports on the latest trend in the semiconductor industry.

For a “value” investor, all this urgent detail is largely irrelevant – because it will all change and evolve, especially in an industry as technologically dynamic and disruption-prone as the chip business. “Value” searches beyond the seasons and cycles and “nodes” to understand the strategic fundamentals over a multi-year horizon.

In the case of Intel, these long-term fundamental trends fall into four categories:

• Strategic market trends
• The company’s strategic position in the industry
• Politics, and Geopolitics
• Cultural factors

The Market

Raw Growth

The semiconductor industry is expected to double – or even triple – in revenue in the next decade or so, up to as much as $1.8 Trillion, with a compound annual growth rate (CAGR) of over 12% annually – significantly higher than many other tech-heavy industries.

This is a major acceleration in the long term trend, driven by the intensifying digitalization of just about everything.

• ‘It took 50 years to become a half-a-trillion-dollar industry. It’ll take just eight to reach a trillion dollars.’ — an industry CEO

First Conclusion: Intel is in the right business.

Diversification of the Customer Base

Semiconductor demand is also broadening significantly. For many years, the chip business was driven by customers in the computer industry. Then smartphones. Now, the applications set is expanding into almost every product category, from kitchen appliances to automobiles. PC’s and smartphones still account for more than half of the demand, but other segments are growing faster – e.g., automotive applications, which are forecast to increase at anywhere from 9-14% CAGR over the next decade (compared to 5-8% CAGR for personal computers and 3-7% CAGR for the smartphone segment).

Second Conclusion: The market is becoming more robust, which should dampen the cyclicality risk over time.

Decoupling & Derisking the Supply Chain (The China Factor)

The shift away from Chinese markets and suppliers is beginning to take hold. For several reasons, many semiconductor customers are moving to logistically “de-risk” their supply chains. Companies like Apple, Dell, and Hewlett Packard have all announced programs to diversify their supply chains and reduce dependence on China. For many customers, lower risk also means “closer to home.”

Third Conclusion: The trajectory of this process is uncertain, but in general it will tend to favor North-American-centric companies like Intel.

In sum, the market for semiconductors is undergoing a giant secular shift as the world economy digitizes furiously. It is expanding rapidly, it is broadening beyond traditional applications and markets, and it is reorienting away from China in particular and in favor of less risky (and often domestic) sources of supply. All these trends will be favorable for Intel’s opportunity set.

2. The Company

Intel should be well-positioned to capitalize on this great boom, for the following reasons.

• Scale: This is an industry where scale is critical. Intel is the largest company in the U.S. semiconductor sector by revenue, the strongest manufacturing base (by far), and the largest number of employees (the most human capital).

• Staying Power: Intel has been the largest U.S semiconductor company by revenue since the 1990’s, and the largest in the world until 2021 (when TSMC overtook them, though with a much narrower business model). As the required level of capital expenditure increased, many companies have capitulated in the race for leadership. This includes former big players like IBM and Motorola. It includes AMD (which sold off its chip fabrication business). Intel is probably the only American company that can afford to play in the major leagues of the global semiconductor manufacturing sector.

• Technology Leadership: In an industry where design and fabrication are both incredibly complex, incredibly expensive, and tightly coupled, the ability to cross-fertilize know-how from both sides of the business is a competitive necessity – and Intel may be just about the only company left in the world that can do that within its own house. They are the only American Integrated Device Manufacturer capable of designing and manufacturing integrated circuits at the high-performance end of the spectrum at the global-level. They have a history of innovation. They have been the industry’s acknowledged technology leader for most of the last several decades. Intel has stepped up to be the first to deploy the next generation Extreme Ultraviolet lithography equipment coming next year from ASML (Netherlands) to support the company’s “five nodes in four years” plan – said to be “on track” – which explicitly aims to regain full technology leadership by 2025. In any case, true technology leadership is a long game, and a cumulative one. Intel has invested far more than its competitors in R&D, last year and over the last 10 years.

• Global Reach: As the industry triples in size, the ability to operate with a global footprint also becomes more important, for a number of reasons including access to technology and supply chain risk reduction. Intel is arguably the only semiconductor company that can truly claim this capability today. The CEO of Intel put it thus:
• “We operate in a globally distributed way. Both TSMC and Samsung essentially have one hub. They’ve never operated in a distributed way, they don’t know how to move people around…”

Rising tides are powerful – and digitalization is a huge economic tide. The simplest summary of Intel’s positioning is this: The semiconductor market will double (or more) in ten years. Intel is the largest American chip company. As the tide rises, Intel will rise with it, likely doubling or tripling its own business. That is the baseline. If Intel is able to rise faster than others, or faster than it has in the accurate period, that is an incremental bonus.

Some semiconductor companies currently enjoy “success” that is based on rather narrow foundations. TSMC is a pure foundry, dependent on a few large customers (Apple is almost one quarter of their business). They have no end-user business of their own. Qualcomm generates almost two-thirds of its revenue from China, and almost all of it from wireless applications. It is probably the most vulnerable to geopolitical tensions now developing (see below). Nvidia has grown by leaping from gaming to bitcoin mining and now to generative AI – all a bit opportunistically. The industry is prone to technological disruptions, and shifts in customer needs. Intel has the breadth and scale to play in almost any corner of the exploding semiconductor market, which should lead to less risk and less volatility in the long run.

3. Politics and Geopolitics

Political trends will amplify the favorable strategic position of Intel in the coming years. These include

• The CHIPS Act: The Federal government has launched the largest program of industrial policy since World War II to stimulate investment in semiconductor research ad manufacturing. Intel has said they will be submitting four proposals for funding support for plants in Arizona, Ohio, New Mexico, and Oregon. The Commerce Department will devote about $28 billion (of the total of $39 Bn) to “investments in leading-edge manufacturing” – where Intel is the main domestic player. [See links to previous columns on CHIPS.]
• The “National Champion” phenomenon: “Intel is the US standard-bearer in chip manufacturing ‘on a de facto basis’… ‘America’s champion.” This implies an enhanced receptiveness in government circles to help Intel play this role successfully – recognizing that there really is no other 100% domestic option. It has created a keen interest in the company’s fortunes, akin to stewardship, on the part of the federal government. And the states lavish Intel with subsidies to attract the company to invest.
• European Support: The EU has developed its own €43 Bn semiconductor subsidy program. Intel is going to build a €30 Bn plant in Magdeburg, Germany, with €10 Bn in subsidy from the German government – and the project may grow to nearly €70 Bn in the next 10-15 years – said to be the largest single foreign direct investment in Europe ever.
• Mitigating the Taiwan/China Risk: The industry is discreetly panicking over the geopolitical risk associated with a potential Chinese move against Taiwan. Currently, TSMC and Taiwan are a “choke-point” in the global semiconductor eco-system, for the most advanced fabrication capabilities, vital for many U.S. companies today, from Apple and Nvidia, to AMD and Qualcomm, who need TSMC to fabricate their high-performance IC’s. The industry needs a more secure and diversified foundry supply chain. Intel’s new foundry operation may prove to be the best answer for many of these customers.

4. The Culture Factor

Culture is a powerful intangible factor, for good or ill. It encompasses the psychological and social framework that conditions how a company operates, how it innovates, how decisions are made, how people follow through, individually and collectively.

In the Intel situation, culture comes into the picture at two levels, both of which were touched on in a accurate panel discussion with Intel CEO Pat Gelsinger.

The Company Culture

The style and experience of the executive leadership is critical in a technology company. The current CEO, Pat Gelsinger, represents a return to the hard-core hardware engineering mindset – “I know how to manufacture silicon at scale” – of the company’s founders. After a string of CEO’s who perhaps didn’t fit that mold, many observers see Gelsinger as the right person to recoup the corporate culture of success.

• Q: Are you shifting the culture back to an Andy Grove culture?
• Gelsinger: I mentored with Andy Grove for 35 years… Tough, aggressive, paranoid, data-driven — we are rebuilding the Grovian culture at Intel.

The National Culture

In the same interview, Gelsinger also expressed his confidence in the national culture of innovation in the U.S. which has formed the context for Intel’s growth from the beginning.

• “I believe that the American, chaotic, open, innovative cycle has destroyed [i.e., defeated] any structured approaches [referring to the top-down models in Japan and China]. Our system is superior because it fosters open, innovative collaboration across the world. Our system won this race…You can have really good engineers but it doesn’t replicate the American system of innovation.”

Feel free to take issue with this, if you wish. Still, in my personal experience, culture counts – a lot. Especially where it involves creative technological innovation. I would place it near the top of the list of the strategic factors that will condition Intel’s prospects for making the “value” turnaround.

Summary

Intel is the largest company in the industry, dominant over decades through many technological transitions, with by far the heaviest and most consistent tradition of investment across the broadest spectrum of key technology categories. The company is already in position to ride the enormous wave of digitalization in the economy which will double or triple the entire semiconductor market in the next decade or so. Intel now has (I think) the leadership, and the strategy to reclaim the pre-eminent position it enjoyed for several decades. It invented the industry and it can reinvent itself.

The company also stands to benefit from favorable political trends and economic subsidies in the U.S. and Europe. There is a new bipartisan commitment to “industrial policy” in the U.S. which will channel naturally to support Intel’s position as the “national champion” in the strategic semiconductor industry. Geopolitical tensions will force semiconductor companies and their customers to diversify away from both China and Taiwan. Intel’s new initiative in the foundry business should draw business from fabless IC companies concerned about that risk.

Timing the turn in a “value play” is always difficult, and as if to underscore the need for patience, Intel’s shares quickly gave back the gains from last month’s surprise good news.

Still, as of yesterday (August 8, 2023), Intel had gained 35% since last October – and that occurred even in the face of a strong cyclical downturn throughout the entire semiconductor industry. And since the beginning of June, Intel is up 13.0% - edging out Nvidia (up 12.3%), doubling the S&P 500 (up 6.3%), and soundly beating Qualcomm (up 0.7%), Texas Instruments (down 4.2%), TSMC (down 4.7%), AMD (down 5.5%), and ASML (down 7.1%).

Perhaps the “value” genie is finally stirring?

For further reading on the CHIPS Act:

Simplify your complex manufacturing processes implementing Dynamics 365 ERP solutions with Dynamics Square.

London, UK - July 25, 2023 —

Dynamics Square UK, a leading provider of business technology solutions, is delighted to announce its groundbreaking initiative to fuel the growth and efficiency of the manufacturing industry through the implementation of Microsoft Dynamics 365 for Manufacturing.

In an ever-evolving global landscape, the manufacturing industry faces a myriad of challenges, including increasing competition, supply chain complexities, and the constant need to optimize operational processes. Recognizing these critical pain points, Dynamics Square UK has committed to empowering manufacturing enterprises with the cutting-edge capabilities of Microsoft Dynamics 365 CRM and ERP Solution.

With a proven track record of delivering innovative technology solutions across industries, Dynamics Square UK's expertise in implementing Microsoft Dynamics 365 Business Central and other ERP solutions. We are set to revolutionize the way manufacturing businesses operate. Such comprehensive ERP solution seamlessly integrates vital business functions, including finance, procurement, inventory management, sales, and production, into a unified platform. By streamlining workflows and enabling real-time data insights, manufacturers can make more informed decisions, enhance productivity, and respond swiftly to market demands.

Key Benefits of Microsoft Dynamics 365 ERP Solution for the Manufacturing Industry:

Enhanced Operational Efficiency: Dynamics Square UK's Microsoft Dynamics 365 ERP Solution optimizes business processes, reduces manual tasks, and minimizes downtime, resulting in improved overall efficiency and cost savings.

Data-Driven Insights: Leveraging the power of advanced analytics and reporting tools, manufacturing enterprises can gain valuable data-driven insights to identify opportunities, anticipate trends, and mitigate risks effectively.

Seamless Collaboration: The ERP solution fosters seamless collaboration among different departments, enabling better communication, coordination, and cooperation to achieve shared objectives.

Scalability and Flexibility: Designed to accommodate the evolving needs of manufacturing businesses, Microsoft Dynamics 365 ERP Solution ensures scalability and flexibility, adapting effortlessly to growth and market changes.

Regulatory Compliance: With compliance being a top priority in the manufacturing sector, the ERP solution helps businesses stay aligned with industry regulations and standards. 

"With Microsoft Dynamics 365 ERP Solution, we empower manufacturing businesses to elevate operations, embrace digital transformation, and achieve sustainable growth."

Manish Goyal, CEO of Dynamics Square UK

Dynamics Square UK invites manufacturers to explore the endless possibilities of Microsoft Dynamics 365 ERP Solution, like Microsoft Dynamics 365 Finance and Operations, and invites interested parties to schedule a consultation with their team of experts.

About Dynamics Square UK:

Dynamics Square UK is a leading provider of business technology solutions, catering to various industries, including manufacturing, retail, finance, and more. With a customer-centric approach and a focus on cutting-edge technology, Dynamics Square UK empowers businesses to thrive in the digital era.

Contact Info:
Name: Nitesh Sharma
Email: Send Email
Organization: Dynamics Square UK
Address: Kemp House, 124 City Road, London, EC1V 2NX, U.K.
Phone: +44-207-193-2502
Website: https://www.dynamicssquare.co.uk/

Release ID: 89103116

If there are any errors, inconsistencies, or queries arising from the content contained within this press release that require attention or if you need assistance with a press release takedown, we kindly request that you inform us immediately by contacting error@releasecontact.com. Our reliable team will be available to promptly respond within 8 hours, taking proactive measures to rectify any identified issues or providing guidance on the removal process. Ensuring accurate and dependable information is our top priority.

The world of artificial intelligence (AI) is currently facing an unprecedented challenge, and it's not about algorithms or data privacy. The AI industry is grappling with a severe shortage of Nvidia H100 GPUs, a crucial component for training large language models (LLMs). This shortage has far-reaching implications, from stalling research projects to affecting the bottom line of major tech corporations.

As of August 2023, the demand for Nvidia H100 GPUs has reached an all-time high. Major tech giants like OpenAI, Azure, and Microsoft are in a fierce race to procure these GPUs, highlighting their significance in the AI ecosystem. Elon Musk's startling remark that "GPUs are harder to get than drugs" and Sam Altman's revelation about OpenAI's delayed projects due to GPU limitations paint a vivid picture of the crisis.

Also read: Acer Nitro 16 Gaming launches in India with Nvidia RTX 4060 GPU

So, what makes the Nvidia H100 GPU so sought after? The answer lies in its technical prowess. The Nvidia H100 is designed for high-performance computing, making it ideal for training LLMs. The H100 boasts a significant number of Tensor cores, specialized hardware for accelerating deep learning tasks. These cores enable faster matrix multiplications, a fundamental operation in neural network training.

Furthermore, with higher memory bandwidth, the H100 can handle vast datasets efficiently, reducing the time required for data transfer between the GPU and the main memory. The architecture of the H100 also allows for seamless scalability. Researchers can link multiple GPUs together to handle larger models and more complex computations.

Despite its power, the H100 is designed for energy efficiency, ensuring that while it delivers top-tier performance, it doesn't excessively contribute to energy costs. Meeting the soaring demand for the Nvidia H100 is no easy feat. The production process is intricate, involving multiple stages: The GPUs require specific rare metals and semiconductors. With the current demand, sourcing these materials in the required quantities is a challenge. The production of GPUs demands high precision to ensure each unit functions optimally. Any compromise here can lead to faulty GPUs, which can be detrimental to AI projects.

With the pressure to increase production, there's a looming risk of quality control oversight. Ensuring that each GPU meets stringent quality standards is paramount.

Also read: Nvidia is probably the most important company in tech right now: Here's why

The Nvidia H100 GPU shortage is a testament to the rapid advancements in AI and the increasing reliance on high-performance computing. As the industry navigates this challenge, it's crucial to strike a balance between meeting demand and maintaining the quality and integrity of the GPUs. The coming months will be pivotal in determining how tech giants, startups, and Nvidia itself address this bottleneck, shaping the future trajectory of AI innovations.

Australia may be running in both lanes of the clean energy race as a trading buddy of the United States and long-standing source of critical commodities for China.

But it is barely out of the starting blocks on building the sovereign industrial base needed for an advanced, low-carbon, job-rich economy, a manufacturing summit has learned.

The nation does not need to be merely "a farm, a quarry and a nice place to live", Assistant Minister for Trade and Manufacturing Tim Ayres told the summit on Thursday.

Advanced economies led by the US want to end the dominance of China in every category of clean energy - from electric cars and factory-ready battery minerals to wind turbines and their components.

Under a framework of tax credits, loans and subsidies, the Biden administration is splashing more than $US1 trillion on clean energy and bold investment plans have been echoed in Europe, Canada, India, Japan and South Korea.

"We want to be beneficiaries, not victims, of the massive investments Americans are making in clean energy manufacturing," Mr Ayres said.

As a free-trade partner of the US, Australia stands to benefit from preferential treatment as a source of metals, minerals and know-how.

"But the IRA (US Inflation Reduction Act), of course, distorts global investment as well as growing global investment," Mr Ayres said.

He said Australia's response to the competitive challenge of the IRA would identify actions in response by the end of this year.

But steel and energy industry leaders urged more haste.

In Australia, nine-fold increases to wind and solar electricity generation capacity are required by 2050 to meet net-zero demands.

Energy equipment giant Vestas expects international demand for wind turbines to outstrip supply by 2025, when Australia's first offshore wind farms will still be in the planning queue.

"Even if you want a turbine, you might not be able to get it," Vestas executive Fredrik Andren-Sandburg said.

Story continues

A research report by the independent Centre for Work released at the summit rejects the claim the federal government cannot afford to match the scale of the US subsidies and tax credits.

Adjusted for the scale of the Australian economy, up to $138 billion in fiscal supports for renewable energy-related manufacturing is needed - and doable - over the next decade, according to co-authors Jim Stanford and Charlie Joyce.

Australia's footprint is mostly limited to the bottom of the value chain - extraction and shipping rocks - they said.

The government has plans to value add to get more out of the country's rich endowment of in-demand minerals.

"Australia's critical minerals strategy? Lovely words but $500 million doesn't cut it," energy finance expert Tim Buckley said.

American economist Adam Hersh said the future of manufacturing in Australia extended well beyond batteries for electric cars as part of the Biden administration's "like-minded club".

"What brings me here is the climate crisis and the imperative for renewable energy," Dr Hersh told AAP.

"It's not all about the US sucking up the resources in manufacturing from the rest of the world."