The AI Boom is Now a $12.4T Hardware Story

The analysis shows that the AI hardware boom extends far beyond the handful of high-profile chipmakers dominating the headlines. The 28 chip & hardware suppliers in the S&P 500 index are collectively worth $12.39 trillion, with AI compute alone accounting for $6.17 trillion.

Anthropic is reportedly eyeing a valuation of more than $2 trillion in what could become the biggest IPO in history. At the same time, the Claude maker is arranging a pre-IPO credit facility of more than $10 billion, showing how much capital is now flowing into the AI infrastructure race. Many companies beyond the AI makers seem to be benefiting, and for more details, I am reaching out with a recent analysis examining the S&P 500 constituents, directly linked to the booming AI economy.

The team at BestBrokers analysed all 503 S&P 500 constituents and classified companies according to their role in the AI value chain – from chipmakers and cloud providers to power utilities, infrastructure suppliers and enterprise software developers – to measure how much of America’s flagship stock index is now directly linked to artificial intelligence. The full data we used for the analysis can be accessed on Google Drive via this link.

The analysis shows that the AI hardware boom extends far beyond the handful of high-profile chipmakers dominating the headlines. The 28 chip & hardware suppliers in the S&P 500 index are collectively worth $12.39 trillion, with AI compute alone accounting for $6.17 trillion.

But the rest of the supply chain is also substantial, spanning networking, memory and storage, chip manufacturing equipment, power and analogue components, as well as the testing and photonics needed to turn AI designs into working hardware.

Here’s how that $12.39 trillion breaks down:

• AI Compute (NVIDIA, AMD, Intel) – $6.17 trillion
• Analogue, power, RF & embedded chips (Qualcomm, Texas Instruments, Analog Devices, Microchip Technology, NXP Semiconductors, ON Semiconductor, Skyworks Solutions) – $673.54 billion
• Chip design & EDA (Cadence Design Systems, Synopsys) – $186.63 billion
• Chip manufacturing & equipment (Applied Materials, Lam Research, KLA, Qnity Electronics) – $1.26 trillion
• Custom silicon & networking (Broadcom, Marvell Technology, Monolithic Power Systems) – $2.12 trillion
• Memory & storage (Micron Technology, SanDisk, Western Digital, Seagate Technology) – $1.73 trillion
• Photonics, optical & semiconductor testing (Coherent, Lumentum, Keysight Technologies, Teledyne Technologies, Teradyne) – $263.51 billion

Here are other highlights from the analysis:

• The analysis identified 28 S&P 500 companies directly involved in making the hardware behind the AI boom, with a combined market capitalisation of $12.4 trillion. The group covers everything from the processors that power AI models to the memory, networking, manufacturing equipment, optical components and testing technology needed to build and run AI infrastructure.
• AI Compute is by far the biggest part of the group in terms of actual valuation on the stock market. The three companies within this segment are collectively worth $6.17 trillion. This includes NVIDIA, AMD and Intel – companies producing the processors and accelerators that provide the raw computing power needed to train and run AI models.
• Custom silicon and networking companies are worth another $2.12 trillion. Broadcom, Marvell Technology, and Monolithic Power Systems develop specialised chips and connectivity technology that help AI systems move data quickly between processors and across massive data centres.
• Memory and storage companies account for $1.73 trillion. Micron, SanDisk, Western Digital, and Seagate provide the high-speed memory and data-storage hardware needed to feed AI systems with enormous amounts of information.
• The companies building the equipment used to manufacture advanced chips are worth $1.26 trillion. Applied Materials, Lam Research, KLA, and Qnity Electronics sit further up the supply chain, providing the machinery and technology semiconductor manufacturers need to produce increasingly sophisticated chips.
• Another $673.5 billion comes from companies making analogue, power, RF, and embedded chips. Qualcomm, Texas Instruments, Analog Devices, Microchip Technology, NXP, ON Semiconductor, and Skyworks Solutions produce the supporting components that manage power, signals, connectivity and other essential functions around AI hardware.
• Photonics, optical technology, and semiconductor testing add $263.5 billion to the overall capitalisation of the S&P 500 constituents. Coherent, Lumentum, Keysight Technologies, Teledyne Technologies, and Teradyne provide technologies used to transmit data using light, connect high-speed systems and test whether increasingly complex chips and hardware actually work as intended.
• Finally, chip-design and electronic design automation companies – namely Cadence Design Systems and Synopsys – account for $186.6 billion. These provide the specialised software engineers use to design, simulate and verify the chips before they are physically manufactured.