AI chip testing firm Aehr Test Systems forecasts $60M to $80M in bookings for the second half of fiscal 2026, driven by wafer-level and packaged-part burn-in systems for AI processors. The California company received $5.5M in orders for Sonoma systems in Q3, exceeding the entire prior quarter and reflecting accelerating demand across global semiconductor manufacturing.
Aehr's lead production customer provided a large AI ASIC forecast with shipments starting May 2026. The company's Silicon Valley test lab processes multiple orders for high-power Sonoma configurations reaching 2,000 watts per device, with production capacity now exceeding 20 systems monthly for both wafer and package-level testing.
The supply chain buildup extends beyond testing. US-based Credo Technology guides to 63.8%-65.8% GAAP gross margin for Q3 fiscal 2026 on strong demand for datacenter interconnects linking GPUs and memory in training clusters. Advanced packaging provider Amkor is expanding capacity globally to meet AI chip assembly demand, while HBM3e high-bandwidth memory ramps across multiple suppliers as hyperscalers in North America, Europe and Asia build out training infrastructure.
Market focus centers on Nvidia's February 25 earnings as the primary indicator of sustained AI hardware spending worldwide. Component suppliers are securing multi-quarter visibility, suggesting continued buildout across major datacenter markets.
Aehr Test's Q2 revenue fell 27% year-over-year to $9.9M on lower WaferPak volumes, but effective backlog reached $18.3M including post-quarter bookings. The company raised $10M through an at-the-market offering to fund operations during the revenue trough. It partnered with ISE Labs and Taiwan's ASE to expand wafer-level and packaged-part testing services for HPC and AI customers.
Beyond datacenter infrastructure, Norwegian firm Ensurge Micropower is positioning solid-state microbatteries for edge AI devices, while Apple's reported AI wearables development signals potential consumer hardware expansion. Datacenter infrastructure remains the primary growth driver globally.
Component suppliers face execution risk if AI capital spending slows, but current forecasts and capacity expansions across the US, Taiwan, and Southeast Asia indicate confidence through mid-2026. Nvidia's guidance will clarify demand trajectory for the international supply chain.