Image

Taiwan’s surging vitality costs are a warning for power-hungry American AI firms—and a boon for the most important chip producer

As American AI firms and knowledge facilities proceed to proliferate and pressure the facility grid, TSMC is an instance of the vitality crunch that might lie of their not-too-distant future. TSMC’s financials are hurting from elevated vitality prices within the quick time period. Nevertheless, its near-monopoly on essentially the most energy-efficient chip know-how means it’s poised to learn from extra widespread energy shortages in the long run, as its know-how turns into important for firms trying to save on energy prices.

“There’s no denying that energy in AI data centers is becoming a big choke point, probably the next choke point,” Michelle Brophy, director of analysis for tech, media, and telecom at analysis platform AlphaSense, instructed Fortune. “In actuality, the [need for] rising energy efficiency in computing enhances [TSMC’s] position.”

TSMC posted stable earnings Thursday morning, driving “insatiable” demand for its highest-end AI chips to a 17% yearly improve in internet income, according to CEO C.C. Wei. 

However vitality prices had been a headwind for TSMC, which manufactures more than 90% of the world’s most advanced AI chips at its factories in Taiwan. TSMC is by far Taiwan’s largest vitality client, and was hit exhausting by a 17% value hike in its energy prices final April adopted by one other 25% improve earlier this month, executives stated in an earnings name with analysts.

TSMC CFO Wendell Huang estimated that elevated vitality bills had diminished the corporate’s margins by as much as almost a full proportion level. Huang additionally adjusted the corporate’s second-quarter margin steering down by barely greater than a proportion level, citing vitality prices.

How rising vitality costs might assist TSMC in the long term

Vitality payments is perhaps hurting TSMC now, however in the long term, it stands to realize from international vitality shortages associated to AI. That’s as a result of TSMC has an enormous head begin in manufacturing essentially the most superior, energy-efficient chips. It’s presently the one provider able to producing so-called three-nanometer chips, essentially the most highly effective and energy-efficient fashions in the marketplace, and it expects to begin mass producing its next-generation two-nanometer chips next year, which might be even sooner and extra environment friendly.

“If we’re thinking about the U.S. [manufacturing facilities] that are going up, it’s more likely that [capital] is going to be spent on a more efficient [two-nanometer chip],” Brophy stated. “That really bodes well for TSMC.”

AI firms have already been scrambling to seek out alternate options to getting their energy from an more and more strained grid. OpenAI CEO Sam Altman has backed a startup that guarantees to make use of small nuclear reactors to energy AI knowledge facilities. Non-public fairness agency Blackstone is reportedly investing $25 billion in knowledge facilities in the course of the desert, positioned proper subsequent to photo voltaic arrays that might energy them immediately. TSMC took matters into its own hands in 2020 when it leased energy from an offshore wind farm, anticipating shortages in Taiwan’s energy grid.

Elevated vitality prices squeezing consumers is one issue that provides TSMC leverage to lift its costs. The chipmaker elevated costs by about 10% in 2022 and one other 5% final 12 months, in accordance with Morgan Stanley analyst Charlie Chan. Consumers demanding a restricted provide of TSMC’s most energy-efficient AI chips might bid up costs even increased.

CEO Wei additionally instructed on a name with analysts that after TSMC’s chip factories in Arizona—which it’s building with support from the CHIPS Act—start manufacturing, excessive labor and gear prices within the U.S. might push costs up additional nonetheless.

“If a customer wants to have manufacturing done within the U.S., obviously, it’s going to cost more—not only from an energy perspective, it’s going to cost more from a labor perspective,” Brophy stated. “Those costs are going to be passed on to the customer.” 

TSMC founder Morris Chang previously estimated that manufacturing costs in the U.S. could be double the price of producing them at TSMC’s Taiwan facility.

“If you give up the competitive advantages of Taiwan and move to the U.S. … the costs are going up,” Chang said.

Though TSMC beat analysts’ estimates, its income was down 3.8% quarter over quarter, which CFO Huang attributed to low seasonal demand for TSMC’s lower-end chips utilized in smartphones. Its inventory fell virtually 7% on Thursday, however Brophy stated that the market’s response was extra a product of its astronomical expectations than any actual failures at TSMC.

“It’s elevated expectations of AI-related semiconductor companies … They’ve been on quite a tear in the last few months. So I’m not totally surprised. They were priced for perfection,” Brophy stated. “The consensus opinion is that the strength of AI is going to outweigh potential weakness in those [lower-performing smartphone] segments.”

Subscribe to the Eye on AI e-newsletter to remain abreast of how AI is shaping the way forward for enterprise. Sign up without cost.

SHARE THIS POST