StatementDog | AI & Tech Supply Chains (@statementdog) header
StatementDog | AI & Tech Supply Chains (@statementdog) profile picture

Tracking the companies and technologies behind the AI buildout. Research informed by company interviews, site visits, analysts, and industry experts.

Taipei City, Taiwan
Joined June 2024
I think Street forecasts for 2028 AI chip spending, aka the numbers behind $NVDA and $AVGO estimates, are too high for the capex that’s supposed to pay for them. The market has GPU and ASIC spending growing about 46% in 2028, to roughly $1.0T. Consensus capex for the big five cloud providers and neoclouds grows only about 14%, to roughly $1.78T. Put those side by side and chips rise from about 35% of that capex in 2026 to 44% in 2027 and 57% in 2028. That implies spending on everything else, from racks and power gear to networking and buildings, falls about 11% in 2028. I expect that spending to grow with the chips. New GPUs and ASICs need power, cooling, networking and floor space before they can earn anything. To support the chip forecasts with chips holding 2027’s 44% share, 2028 capex would have to come in about 28% above consensus. I think it’s more likely to come in 5–15% below. When the two numbers have to meet, I think the chip forecasts are the ones more likely to come down. Some chips go to buyers outside these companies, so the ratio is approximate. A 13-point jump in a single year is still hard to explain that way. The chip forecasts also leave less room for growth than the headline suggests. On the market’s own numbers, the 2028 increase in chip spending, about $318B, is already about 6% smaller than 2027’s $339B. At a 44% share, it would be about 70% smaller. Fab and advanced-packaging expansion follows that yearly increase. I think a smaller increase shows up first in new equipment orders for 2028–29 capacity. 2027 tool shipments are still backed by orders already placed, with $LRCX guiding 2027 WFE above $200B.
I think 2028 AI capex is more likely to come in 5–15% below consensus. OpenAI and Anthropic are growing revenue too slowly to cover the rent on everything being built for them. By my model, the two labs need about $15B of net new ARR every month through 2028 to pay their compute bills. This year they’ve averaged about $10.5B. I started from the landlord’s side. NVIDIA puts a gigawatt of Rubin capacity at about $60B. Depreciate the IT gear over 5 years and the building over 14, charge a 9% cost of capital, add power and operations, and the owner needs roughly $14B a year in rent per GW to break even. That’s on the low side. CoreWeave says its new 4–5 year contracts price at $20–25M per MW. Consensus has the big five cloud providers plus neoclouds spending about $1.0T, $1.6T and $1.8T in 2026–28. After stripping out non-AI spend and capacity they use themselves, I estimate OpenAI and Anthropic will be paying about 70% of the rent on leased-out AI capacity by 2028. Covering that would take about $441B of combined revenue in 2028. At this year’s pace, I estimate they reach about $347B, 21% short. Monthly net new ARR is the number that matters, and it has been choppy even as the base grew. Anthropic’s annualized revenue rose about ninefold this year, yet it added about $16.7B in April, around $3B a month in August and September, and about $10B in October. Token prices roughly halve every year, so usage has to grow about 4x a year just to keep revenue on track. GPU life is the one variable that closes the gap by itself. CoreWeave said Ampere and Hopper pricing rose 30–50% quarter over quarter in Q3. If older chips keep renting at strong prices and servers last 7 years instead of 5, the revenue needed in 2028 falls to about $342B, just under this year’s pace. That has only been tested while compute is scarce, so I keep 5 years as my base case. Most 2027 spending is already under contract, so I expect the adjustment to land on 2028 expansion. Oracle and the neoclouds look most exposed. Analysts estimate OpenAI accounts for roughly half of Oracle’s backlog, and by my math OpenAI alone needs about $90B of outside money to cover 2028 compute, while Anthropic has about $19B to spare. Neoclouds also borrow at higher rates. Financing drives the size of the cut. My base case assumes OpenAI raises about $27B more. With no new money, the 2028 cut widens to 16–26%. With about $65B, consensus holds. A 5–15% cut would slow 2028 GPU and ASIC spending growth from the roughly 46% the market expects to 24–39%. Chipmakers would need less new equipment for 2028–29 capacity, so tool orders could soften first, as early as 2027. Even after that cut, 2028 capex would still run about 50–70% above 2026. I still see AI demand expanding over the long run. Capex expectations got ahead of revenue. Anthropic’s Q4 roadshow guidance implies the two labs’ combined ARR barely grows from October to year-end, which lifts the monthly pace needed in 2027 to about $16.5B. Two straight quarters at that level would ease the pressure on 2028 capex and change my view. $NVDA $CRWV $ORCL #OpenAI #Anthropic
4
1
55
10,301
Elon Musk said GPU power swings in AI training can break the shafts on gas turbines. SpaceX pairs its AI data centers with Tesla Megapacks partly for that reason. Without power smoothing “at a wide range of frequencies,” he said, training runs are very difficult. We see power smoothing becoming a fixed cost of every large training cluster, and Taiwan’s power-supply makers are built into the layers closest to the GPUs. NVIDIA says a rack can swing from about 30% to 100% utilization and back in milliseconds, while power plants can take one minute to 90 minutes to respond to a sudden ramp. It also warns that repeated swings can cause resonance and stress equipment, the same problem Musk described. No single device covers that range, so storage gets layered by speed. Capacitors inside the rack’s power supplies absorb millisecond spikes. Supercapacitor units cover sub-second to several-second swings. Battery backup units (BBUs) bridge about a minute, and site batteries like Megapack handle the slower swings at the turbines. Taiwan’s suppliers sit in the rack layers. • Lite-On (2301 TT) co-designed the GB300 NVL72 power shelf with NVIDIA, which built in electrolytic capacitors to absorb millisecond spikes. NVIDIA says the design cut peak grid draw by 30% versus GB200 in a Megatron LLM training test. • Delta Electronics (2308 TT) covers three layers: AI rack power supplies, complete BBU systems, and standalone supercapacitor cabinets that sell for more than $200,000 each, according to our research. • Dynapack (3211 TT) makes the battery packs inside Delta’s BBUs. That rack-level content grows when Rubin-era 800V power racks arrive in 2027 and 2028 and BBUs move into them at 20 to 25kW per module. We estimate the BBU market inside IT racks and power racks rises from $3.3 billion in 2026 to $11.2 billion in 2028. $SPCX $TSLA $NVDA
2
1
20
5,607
Two U.S. ASIC customers could need up to 2 million sheets of CCL a month as they ramp purchases from Q1 2027. The initial demand is expected to come from boards for AI servers and network switches. That could push M4 and higher grades above half of Nan Ya Plastics’ (1303 TT) CCL revenue, up from 30% in Q3 2026. The share is expected to approach 40% in Q4, ahead of the ASIC ramp. Nan Ya’s CCL lines are running at over 90% utilization. Total capacity is expected to reach 6.48 million sheets a month by year-end and stay at that level through 2028. That puts the focus on what Nan Ya makes with those lines. Higher-grade laminates could account for a much larger share of sales even with overall capacity flat.
2
39
2,060
I think 2028 AI capex is more likely to come in 5–15% below consensus. OpenAI and Anthropic are growing revenue too slowly to cover the rent on everything being built for them. By my model, the two labs need about $15B of net new ARR every month through 2028 to pay their compute bills. This year they’ve averaged about $10.5B. I started from the landlord’s side. NVIDIA puts a gigawatt of Rubin capacity at about $60B. Depreciate the IT gear over 5 years and the building over 14, charge a 9% cost of capital, add power and operations, and the owner needs roughly $14B a year in rent per GW to break even. That’s on the low side. CoreWeave says its new 4–5 year contracts price at $20–25M per MW. Consensus has the big five cloud providers plus neoclouds spending about $1.0T, $1.6T and $1.8T in 2026–28. After stripping out non-AI spend and capacity they use themselves, I estimate OpenAI and Anthropic will be paying about 70% of the rent on leased-out AI capacity by 2028. Covering that would take about $441B of combined revenue in 2028. At this year’s pace, I estimate they reach about $347B, 21% short. Monthly net new ARR is the number that matters, and it has been choppy even as the base grew. Anthropic’s annualized revenue rose about ninefold this year, yet it added about $16.7B in April, around $3B a month in August and September, and about $10B in October. Token prices roughly halve every year, so usage has to grow about 4x a year just to keep revenue on track. GPU life is the one variable that closes the gap by itself. CoreWeave said Ampere and Hopper pricing rose 30–50% quarter over quarter in Q3. If older chips keep renting at strong prices and servers last 7 years instead of 5, the revenue needed in 2028 falls to about $342B, just under this year’s pace. That has only been tested while compute is scarce, so I keep 5 years as my base case. Most 2027 spending is already under contract, so I expect the adjustment to land on 2028 expansion. Oracle and the neoclouds look most exposed. Analysts estimate OpenAI accounts for roughly half of Oracle’s backlog, and by my math OpenAI alone needs about $90B of outside money to cover 2028 compute, while Anthropic has about $19B to spare. Neoclouds also borrow at higher rates. Financing drives the size of the cut. My base case assumes OpenAI raises about $27B more. With no new money, the 2028 cut widens to 16–26%. With about $65B, consensus holds. A 5–15% cut would slow 2028 GPU and ASIC spending growth from the roughly 46% the market expects to 24–39%. Chipmakers would need less new equipment for 2028–29 capacity, so tool orders could soften first, as early as 2027. Even after that cut, 2028 capex would still run about 50–70% above 2026. I still see AI demand expanding over the long run. Capex expectations got ahead of revenue. Anthropic’s Q4 roadshow guidance implies the two labs’ combined ARR barely grows from October to year-end, which lifts the monthly pace needed in 2027 to about $16.5B. Two straight quarters at that level would ease the pressure on 2028 capex and change my view. $NVDA $CRWV $ORCL #OpenAI #Anthropic
1
6
2
48
10,756
NVIDIA is locking in more capacity in Taiwan for two key steps in making its AI chips: ABF substrates and advanced packaging. Kinsus (3189 TT) plans to more than triple the advanced ABF substrate capacity it dedicates to NVIDIA at one of its sites, targeting over 3.5× current capacity. Production is expected to start in H2 2027, with additional capacity targeted for Q3 2028. SPIL, part of ASE, is getting billions of dollars in funding from NVIDIA to expand its contracted advanced packaging capacity for NVIDIA by mid-2029. $NVDA $ASX
1
1
29
1,808
AI demand is giving suppliers more pricing power in everyday circuit-board materials. At Ventec (6672 TT), general-material prices have nearly doubled this year, while that business’s gross margin rose from roughly 16–17% in Q1 to about 40% in Q3, according to a Nov. 16 CTBC Securities report. Ventec makes copper-clad laminates and prepregs, the materials used to build and bond the layers of printed circuit boards. The report describes pressure at two levels. Glass-fabric capacity is shifting toward higher-end grades, squeezing the supply of conventional E-glass. Major laminate makers are also prioritizing AI orders, leaving less capacity for lower-end products. Ventec has been taking spillover orders while keeping lead times at about four weeks. Higher prices helped lift margins, alongside a growing share of higher-margin prepregs and more heat-resistant, halogen-free materials. We see this as a way AI spending can raise costs beyond the AI hardware itself. Buyers of conventional board materials can face tighter supply even without stronger demand for their own products. Suppliers that can still deliver gain room to raise prices. That pricing power depends on how long the shortages last. CTBC already sees price increases slowing for thick glass fabric, which can be made on Chinese weaving equipment. The report says Ventec expects supply and demand for thick fabric to improve after Q2 2027.
17
1,358
Taipei 101 is lit up tonight with “Taiwan ♡ Jensen.” Less than a week after Jensen Huang resigned as Nvidia CEO, a group of Taiwan tech industry associations pooled their money to pay for the light show. The tower is rotating through “Thrive Together,” and a line about building the future of AI together. For a lot of people here in Taiwan, Jensen never felt like a foreign CEO. He was born in Tainan, he’s at Computex every year, and he’s famous for night market runs and big dinners with his Taiwanese suppliers. Nvidia orders keep a huge part of the island’s supply chain busy, from TSMC to the server makers. The organizers say tonight is about years of working together and hoping that partnership continues.
3
16
1
284
7,163
CMP pad supplier Praise Victor (7768 TT) told us it is short of capacity and expanding, as chip packaging takes on polishing steps that used to belong mainly to wafer fabs. Its subsidiary IV Technologies makes CMP pads, and its customers include $TSM and $MU. Semiconductor products were 66% of revenue in the first half of 2026, up from 48% in 2023. At this month’s NCKU Materials Industry Forum, we are pleased to talk to IV Technologies President Yang Wei Wen. He told us CoWoS lines in Taiwan already use a lot of CMP, and that CoPoS will bring in new materials, including glass substrates in the future. Supply is the constraint here. At its August earnings call, the company said existing lines were close to full and that a major advanced-packaging customer had asked it to hold 90 days of inventory. New plants in Hefei and Taichung are expected to contribute from 2027, and Yang said a new R&D center will be completed next year. Our read: pads are consumables, so demand follows the number of polishing steps as well as the number of wafers. If packaging keeps adding planarization, pad demand rises with packaging capacity too. How quickly Praise Victor benefits depends on when its new plants pass customer qualification. Our full interview with Yang is in the video.
1
1
1
18
4,456