The Silicon Oligarchy: A Modest Investment

by

AI Infrastructure

The current enthusiasm for ‘growth stocks’ – a term which, one suspects, disguises a frantic search for anything that might postpone the inevitable – has been sustained for some years. The latest iteration centers, naturally, on Artificial Intelligence, a field promising either utopia or, more likely, a more efficient means of being exploited. Assuming a degree of continued prosperity – a reckless assumption, admittedly – a modest investment in those companies poised to benefit might prove… diverting. One proposes, therefore, a tripartite allocation of three thousand dollars, directed towards those entities best positioned to profit from this digital delirium.

Nvidia

The spending on infrastructure required to support this AI fever is, frankly, vulgar. Five of the largest ‘hyperscalers’ – a term redolent of bloated ambition – have pledged a collective seven hundred billion dollars this year alone to expand their ‘cloud capacity.’ This sum, one notes, exceeds the gross domestic product of all but twenty-four nations. The beneficiaries will, predictably, be few. Nvidia (NVDA +1.74%) appears, at present, to be amongst them. They command approximately ninety percent of the market for graphics processing units – the very engines of this new digital age – a dominance achieved not merely through ingenuity, but through the shrewd construction of barriers to entry. Their CUDA software platform, one gathers, is where most of the foundational code is written, effectively holding the industry hostage. They have even begun to offer ‘end-to-end AI server solutions,’ which sounds suspiciously like selling shovels during a gold rush. Despite this success, the stock trades at a forward price-to-earnings ratio of just over 23.5, which, in the current climate, is almost… reasonable.

Alphabet

Alphabet (GOOGL +0.47%) (GOOG +0.37%) possesses, rather comprehensively, the entire AI ‘stack’ – a term which, one suspects, is designed to intimidate the less technologically inclined. They boast both a leading large language model – ‘Gemini,’ naturally – and their own custom AI chips, ‘Tensor Processing Units.’ Designing one’s own chips, one observes, provides a structural cost advantage, allowing them to train and run their models more cheaply than the competition. They are now, rather generously, allowing customers to lease these TPUs, opening up yet another revenue stream. Anthropic, one learns, has placed a substantial order with Broadcom, Alphabet’s chip partner, while Google is to become Apple’s preferred cloud provider – a development which, one suspects, will not be greeted with enthusiasm in Cupertino. Alphabet remains, of course, the dominant player in search, a position secured through its control of Chrome, Android, and a revenue-sharing agreement with Apple. They have embedded ‘Gemini’ throughout their solutions, contributing to an increase in search queries and, consequently, revenue. The stock trades at a forward P/E of 26.5, which, while not exactly a bargain, is preferable to many of the alternatives.

Taiwan Semiconductor Manufacturing

Taiwan Semiconductor Manufacturing (TSM 0.53%) is, quite simply, the leading chip manufacturer in the world. While its competitors struggle to produce functional silicon, TSMC has established a virtual monopoly on the production of advanced chips, including GPUs and TPUs. This dominance stems from both technological expertise and sheer scale. With the market for AI infrastructure booming, TSMC is working closely with chip designers to increase manufacturing capacity and has greatly ramped up its own spending to build new fabrication plants. They are, unsurprisingly, enjoying strong pricing power and, according to reports, have already informed customers of planned price increases over the next four years. The stock trades at a forward P/E of 25.5, making it another attractively valued growth stock. Between its growing production capacity and its ability to boost prices, it possesses, one suspects, considerable long-term growth prospects.

Read More

2026-02-19 16:23