Market — AI Servers & Rack-Scale Systems

Verified claims and product-axis read for AI Servers & Rack-Scale Systems. Every fact below is sourced; every product judgment traces back to underlying signals.

Verified facts

• Celestica's ATS (Advanced Technology Solutions) segment is being de-emphasized as HPS (AI rack/datacenter) is now >50% of revenue mix as of late 2024. ↗ (financial)
• Jabil completed divestiture of its Mobility business to BYD Electronic for ~$2.2B in 2023, redirecting proceeds toward AI/cloud capacity. ↗ (financial)
• Flex announced a $2B share buyback in 2024-25 timeframe while increasing AI capex, signaling confidence in cash-generation through AI cycle. ↗ (financial)
• HPE projected FY25 AI server orders to exceed $4B, vs prior-year ~$2.9B. ↗ (financial)
• Celestica's FY25 EPS guidance was raised to $4.75-5.00 (from $4.42 prior), citing AI compute and networking mix. ↗ (financial)
• Dell's CFO disclosed that AI server revenue in FY25 was ~$10B, on track to exceed $15B in FY26 per management guidance. ↗ (financial)
• Supermicro pre-announced FY26 revenue target of ~$40B (vs ~$22B FY25E), implying 80%+ YoY growth heavily dependent on GB200/GB300 rack ramps. ↗ (financial)
• Jabil announced an expanded partnership with Hewlett Packard Enterprise to manufacture HPE Cray XD systems following the 2024 'Mexicali' EMS deal valued at ~$2B. ↗ (financial)
• HPE acquired Morpheus Data in 2024 for hybrid-cloud orchestration that now layers over rack-scale AI deployments. ↗ _(historical_event)_
• Celestica's CEO publicly stated >50% of capex in 2025 is dedicated to expanding rack integration footprint, primarily in Texas and Thailand. ↗ _(historical_event)_

Cross-cutting opportunities (industry read)

• PCIe Gen5/Gen6 & CXL Retimer / Switch IC — Hyperscale AI racks need PCIe Gen5→Gen6 + CXL retimers/switches just to keep host-to-accelerator links alive at scale; this is Astera Labs' core ALAB-listed business.
• 800G / 1.6T Co-Packaged Optics & Silicon Photonics module — Scale-out beyond ~100K GPU clusters (verified Colossus class) is bandwidth-bound; pluggable optics are hitting reach/power limits, so the industry is racing to ship CPO at 800G→1.6T while still selling AEC for short reach.
• UALink AI Scale-Up Fabric (GPU-to-GPU) — NVLink lock-in is the single biggest non-NVIDIA accelerator pain point; the UALink consortium is the answer and Astera is the first merchant silicon making it ship-worthy.

See the Products and Strategy modules for the full product list and forward-looking judgment.