In the words of Bjork’ 90s indie hit “Oh So Quiet” –It's, oh, so quiet Shhhh, Shhhh, It's, oh, so still Shhhh, Shhhh, You're all alone Shhh, Shhh And so peaceful until…Until… that is the question, and considering it is ‘peaceful’, it's probably best to review the minutes from the Fed as it is signalling that the quiet time is not far from ending soon.FOMC: The Pressure BuildsThe May 6th to 7th Federal Open Market Committee (FOMC) minutes reaffirmed the Fed’s cautious stance, with Chair Powell keeping to the “wait and see” script. But under the surface, the outlook has become more complicated as event risk is getting louder.Clearly, Trump’s Tariffs have created new complications for the Fed’s dual mandate.As the minutes note:“With uncertainty higher due to ‘larger and broader’ than expected tariffs, the Committee may ultimately face a more difficult trade-off between its price stability and full employment mandates.”And this was well before the Trade Court’s decision that the Liberation Day tariffs are illegal under the Economic Emergency Act of 1977, and then it was subsequently overturned 24 hours later by the appeals court.The Fed has flagged increased downside risk to real activity and now sees the probability of recession as nearly equal to its baseline forecast. At the same time, inflation risks for 2025 have been revised upward, though longer-term projections remain skewed to the upside, particularly as inflation expectations creep higher.Seen in these quotes from the minutes:“The staff continued to view the risks around the inflation forecast as skewed to the upside, with recent increases in some measures of inflation expectations raising the possibility that inflation would prove to be more persistent than the baseline projection assumed.”“Many participants reported that firms planned to partially or fully pass on tariff-related cost increases.”To paraphrase Milton Friedman, “Tariffs are not a tax on the sovereign, they are a tax on the consumer.” And this is what is being missed by government officials and the President himself.A counterargument to higher cost is that Fed officials suggested there is a chance of weakening demand, lower immigration driven housing inflation, and competitive pricing tactics. Which would feed back into the risk of recession as mentioned above, and signal that the US is entering a new stagflation era.Seen here:“Several argued that there might be less inflationary pressure for reasons such as reductions of tariff increases from ongoing trade negotiations, less tolerance for price increases by households, a weakening of the economy, reduced housing inflation pressures from lower immigration, or a desire by some firms to increase market share rather than raise prices.”On employment, the labour market remains tight but is potentially vulnerable to hiring pauses as policy and trade risks weigh.“The labour market was seen as ‘broadly in balance’ and the unemployment rate as ‘low.’”“Participants were concerned that tariff uncertainty could lead to a pause in hiring and the labour market to soften in the coming months.”Financial market signals were mixed. Several participants noted an unusual pattern: long-term Treasury yields rose even as the dollar weakened and equities sold off, raising concerns about shifting correlations and safe-haven perceptions.“Some participants commented on a change from the typical pattern... with longer-term Treasury yields rising and the dollar depreciating despite the decline in the prices of equities and other risky assets... [noting] that a durable shift... could have long-lasting implications for the economy.”Monetary framework discussions continue as well. The Fed appears to be reconsidering its post-COVID commitment to flexible average inflation targeting (FAIT). The minutes state:“Participants indicated that they thought it would be appropriate to reconsider the average inflation-targeting language in the Statement on Longer-Run Goals and Monetary Policy Strategy.”An interesting development is putting more rigidity into the mandate currently, suggesting the Fed is looking to ‘safeguard’ policy changes from external political forces.Where does this leave the US and the Fed in the short term? Don’t expect any near-term policy change, but the longer the Fed delays, the steeper the eventual rate cuts may need to be as the risks of a tariff-induced recession lead to the monetary brake being released.The consensus is that by January 2026, a possible 125 basis point will come out of the Federal funds rate, some even are forecasting 175 due to the need to stimulate the economy rather than restrict it. The consensus figure would see the Federal Funds rate landing on the terminal rate of 3.00% to 3.25%, the unknown is when, the size and velocity of reaching this point will be.It is oh so quiet, but it won’t be for long if the Fed is anything to go by.
Jensen Huang stood on stage at GTC 2026 and projected US$1 trillion in cumulative AI hardware revenue through 2027, spanning the current Blackwell generation and the newly announced Vera Rubin architecture. That is not just a corporate forecast. It is a gravitational pull reshaping parts of the global technology sector.
In market circles, this effect is often linked to Huang's ability to move sentiment across AI-related stocks.
Here is the part that many retail investors can miss: NVIDIA is a fabless chip designer. It conceives the architecture and writes the code, but manufactures none of the actual silicon. Every dollar of that US$1 trillion projection would need to flow through a highly concentrated manufacturing pathway, and that route runs directly through Asia.
For APAC traders, the headline rally in New York is only half the story. The broader opportunity sits inside the Asian technology giants linked to the hardware supercycle: the companies making the parts, infrastructure and capacity without which none of this works.
Why the hardware stack matters
The largest passive exchange traded funds (ETFs) in the world are moving through a highly concentrated market structure. According to Morningstar Direct and Trivariate Research data, approximately 31.3% of the S&P 500 is now concentrated in just seven stocks. When too many dollars chase too few names, diversification can become less reliable and valuation multiples are more exposed.
The APAC enablers tell a different story. They are less crowded than the US mega-cap AI trade, central to the buildout and driven more by volume capture than multiple expansion.
The thesis is direct: identify the companies supplying the raw materials, components and infrastructure, regardless of which AI model ultimately wins the commercial software race.
Five stocks across the AI infrastructure chain
Value Chain Stack Architecture // Individual OperatorsTaiwan Semiconductor Manufacturing Company is the foundry that makes the most advanced processors used across NVIDIA's AI accelerator roadmap. There is no credible alternative at scale for the cutting-edge chips the industry currently requires. That gives TSMC significant strategic relevance in this cycle.
For Q1 2026, the company posted revenue of US$35.9 billion, up more than 40% year-on-year, with a gross margin of 66.2%. High-performance computing (HPC), including AI-related revenue, accounted for about 61% of Q1 revenue.
Samsung sits one layer above the processing core in the AI chip stack, supplying the high-bandwidth memory (HBM) that helps advanced processors operate at the speeds artificial intelligence workloads demand.
Samsung says its sixth-generation HBM4 is now in mass production and designed for the Vera Rubin platform. That places Samsung inside the next phase of AI infrastructure demand, alongside other HBM suppliers competing for allocation across advanced systems.
SK Hynix pioneered earlier generations of HBM architecture and remains deeply integrated into the NVIDIA value chain. That relationship is visible in upstream data: FormFactor reported SK Hynix accounted for 29.5% of its Q1 2026 revenue, with NVIDIA accounting for another 10.2%.
SK Hynix is also reportedly evaluating whether its memory products can work with Intel's packaging technology. That move reads as a potential hedge against TSMC's constrained CoWoS capacity.
While the semiconductor companies capture the manufacturing layer, Alibaba represents the enterprise adoption layer. China's 15th Five-Year Plan for 2026 to 2030 places significant emphasis on an "AI plus" initiative and technology self-reliance.
Alibaba gives investors exposure to China's domestic AI infrastructure push, including customised computing clusters using locally designed application-specific integrated circuits (ASICs) as an alternative to Western-restricted hardware.
Hitachi is not a chip company. It is an industrial conglomerate with deep expertise in factory automation and power grid infrastructure. AI data centres consume enormous amounts of electricity, which can place serious pressure on power networks.
Hitachi recently announced a major collaboration with Intel covering factory automation, energy infrastructure and custom chip design. Hitachi links the digital AI story with the infrastructure layer in Japan, where grid investment, automation and industrial efficiency are becoming part of the same conversation.
This is the main macro date APAC tech traders need to watch.
A hawkish hold is expected as policymakers weigh energy-driven inflation. The RBA's posture is likely to remain important for the yield floor in Australian dollar carry trades.
Markets are pricing a 66% probability of a move to 1.00% as policymakers weigh yen weakness and the risk of a disorderly breach of the 160.00 level.
Do not just watch the green candles in New York. The broader AI infrastructure story runs through memory in Seoul, foundries in Hsinchu and power grids in Tokyo. For traders, the task is to understand which parts of the hardware stack are most exposed before the next macro catalyst arrives. On 16 June, central bank decisions in Australia and Japan could shift the backdrop for APAC technology names.








