🧔♂️ A friendly human may check it before it goes live. More news here
Taiwan’s Pegatron targets 10-fold growth in AI server revenue
Pegatron said it aims to grow AI server revenue more than tenfold this year from last year, driven primarily by orders for Nvidia’s GB300 server platform and by orders from neocloud solutions clients, and large cloud service providers.
Pegatron is a Taiwan-based electronics manufacturing services provider.
The company said server shipments will rise each quarter with stronger performance expected in the second half.
Gary Cheng, co-CEO, told an earnings conference that the AI server segment would face gross margin pressure in the short term from infrastructure investment but would improve beyond industry averages over the long term.
Johnson Teng, co-CEO, said the computer, consumer electronics, and communications segments will face seasonal revenue drops this quarter and that PC sales are exposed to geopolitics and memory shortages.
Pegatron reported Q4 net profit of NT$5.3 billion (US$168.5 million), up 15.9% quarter-on-quarter and 40.9% year-on-year, while fourth-quarter revenue fell 2.3% year-on-year to NT$319.5 billion (US$10.16 billion).
Communications products made up 62% of Q4 sales, computing goods 15%, and consumer electronics 7%.
🔗 Source: Taipei Times
🧠 Food for thought
Implications, context, and why it matters.
Nvidia’s GB300 platform shift is creating an opening for Pegatron
- Pegatron said its AI server growth depends on demand for Nvidia’s GB300 platform plus orders from neocloud solutions clients (cloud providers that rent out GPU computing capacity) and large cloud service providers.
- With GB300, Nvidia is expected to sell several core parts separately. GB200 differed because Nvidia shipped the full Bianca board. That change brings more original equipment manufacturers (OEMs) and original design manufacturers (ODMs) into the compute tray market 1.
- More builders, including Pegatron, can take on final assembly and system integration for large cloud providers 1.
- Pegatron is expanding AI server output in Taiwan, Mexico, and the US. Its Mexico AI server line finished trial production late last year. The first Texas factory is slated to finish by end of March, with trial production planned for April 2.
The transition to next-gen AI hardware is reshuffling AI server suppliers
- As new AI platforms roll out across the industry, some suppliers gain share while others lose ground rather than everyone benefiting equally 3.
- Pegatron said its AI server business continues to grow at triple-digit rates. Super Micro cut fiscal 2026 revenue guidance to US$33 billion from US$40 billion, citing slower sales, margin pressure, and inventory tied to older products 4.
- Super Micro has now come in below revenue expectations for four straight quarters. Foxconn and Wistron reported faster AI server shipment growth over the same stretch 3.
- Fast execution on complex platform changes is becoming a clear divider among hardware suppliers. That includes handling wider use of liquid cooling in systems such as Nvidia’s GB300 NVL72 (a high-end, rack-scale server configuration that uses Nvidia’s latest GPUs) 5.
Stay updated on the go with our mobile app.
Get latest insights with smoother, more personalized experience through TIA mobile app.