Etching coolant "into the chip"! What does Microsoft's new technology mean?

Etching coolant "into the chip"! What does Microsoft's new technology mean?

```

As the AI computing power race heats up, chip power consumption and cooling have become the physical bottlenecks restricting industry development. Recently, a Microsoft technology that "engraves coolant channels into chips" directly targets this industry pain point, causing a sensation.

On September 23 local time, Microsoft CEO Satya Nadella announced on X: "We are reimagining how we cool chips, making the AI infrastructure of the future more efficient and sustainable."

Microsoft's "in-chip microfluidics" cooling technology etches tiny channels inside chips, allowing coolant to directly remove heat. Laboratory tests show its cooling efficiency can be up to three times that of traditional cold plates.

However, Wall Street has taken a cautious stance toward this technology. According to ZF Trading Desk, Morgan Stanley stated bluntly in its latest research report that, despite the impressive performance, which may negatively affect the sentiment around traditional cooling solution suppliers, commercialization remains challenging: the path from technical demonstration to industry standard is still long.

Coolant "engraved into chips": Cooling efficiency up to three times higher

AI chip power consumption has soared from hundreds to thousands of watts, and traditional air and cold-plate cooling technologies are increasingly overwhelmed. Citing media reports, Microsoft developed a new "in-chip microfluidics" cooling system that channels coolant directly into the minuscule etched channels inside the chip, cooling it at the heat source.

Microsoft's experimental data revealed that this technology's performance is striking:

  • Its heat dissipation efficiency can be three times higher than that of current mainstream cold plate technology.
  • It can reduce internal GPU temperature rise by 65%.

In theory, this means data centers can safely "overclock" without adding hardware to cope with surges in computing demand, thus significantly improving key metrics such as cost, reliability, and sustainability.

Morgan Stanley: Three major practical challenges cannot be ignored; widespread adoption will take time

While laboratory data is exciting, Morgan Stanley promptly identified three core obstacles for the technology's transition from lab to large-scale commercial use in its research report. For data center operators, these practical concerns outweigh pure performance metrics.

  1. Reliability: The report emphasizes that liquid leakage has always been a "pain point" for liquid cooling solutions. In actual server rack operations, “reliability is far more important than performance.” Any leak could lead to catastrophic outcomes.
  2. Design Integration: Microfluidic design requires connecting the chip's internal liquid circuit to the server tray manifold, and then to the rack manifold. This demands deep integration with server architecture, and is extremely complex.
  3. Serviceability: For hyperscalers seeking ultimate efficiency and lowest operational costs, whether the solution is easy to maintain and replace is key for large-scale adoption.

Because of these challenges, Morgan Stanley expects that it will take time for in-chip microfluidics technology to be widely adopted.

For capital markets, Microsoft's announcement is undoubtedly explosive. The report points out that the news "may have a negative emotional impact on the market sentiment of the cooling supply chain."

However, the report’s conclusion is not pessimistic. Rather, it believes that since broad adoption of the new technology will take time, any pullback in supply chain companies’ share prices is actually a buying opportunity.

 

~~~~~~~~~~~~~~~~~~~~~~~~

The above excellent content comes from ZF Trading Desk.

For more detailed interpretation, including instant analysis and frontline research, please join [ZF Trading Desk · Annual Membership]

Risk Warning and DisclaimerThe market carries risks, and investments should be made cautiously. This article does not constitute personal investment advice and does not take into account the particular investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article are appropriate to their particular situation. You are responsible for any investment decisions made based on this information. ```