Foxconn Secures Exclusive NVIDIA Supply Deal For Groq 3 LPX Cabinets And Dominates The Vera Rubin Platform With Tenfold AI Hardware Growth
Foxconn maintains its position as the leading supplier for NVIDIA's supply chain while Groq 3 LPX accelerates its development. NVIDIA has delayed its AI inference hardware release schedule by moving the Groq 3 LPX cabinet launch to the third quarter of this year. Foxconn which operates under the name Hon Hai Precision Industry gains an essential role in semiconductor operations through strategic acceleration that establishes Foxconn as a crucial player in semiconductor operations throughout the world. UDN's most recent news update confirms that Foxconn will exclusively supply all computing trays needed for this product line. Foxconn has obtained the main agreement which allows them to build all LPX cabinets for NVIDIA's inference system expansion in their market share.
The partnership extends to a size which results in a tenfold increase of the forecasted hardware delivery volume. UDN supply chain information indicates that the LP30 and LP35 chips used in these cabinets will be shipped at respective amounts of one point five million and two point five million units during the period between 2026 and 2027. The first batch which will be released during the third quarter will include about 6000 cabinets. The company expects to reach a total of 10000 units by 2027. Foxconn becomes an essential part of high performance computing through its role as a core manufacturer for systems which use LP40 silicon technology.
Financial analysts are watching Foxconn's increasing presence in the NVIDIA Vera Rubin platform. The company will achieve a market share range between 55 percent and 60 percent for its VR200 NVL72 cabinets according to projections. The cloud service providers Google Microsoft and Amazon AWS have made large procurement orders which drive this growth. Foxconn will maintain almost total control of Google orders while securing more than half of the business from the other two cloud companies. Production for VR200 NVL72 units will start in mass quantities during the third quarter.
Foxconn leadership has directed a quick manufacturing capacity growth to handle the extreme demand for AI infrastructure which requires new capacity at their manufacturing sites. Chairman Young Liu previously indicated that the company had reached a production rate of one thousand data center cabinets per week. The manufacturing industry has set goals that require the company to produce 2000 cabinets by the end of this year. Foxconn maintains its operational capacity to produce manufacturing products for NVIDIA while serving as an essential production partner for Google TPU server systems. Foxconn will keep 15 percent of the Google TPU server market until 2026 because AI agent services have become more popular which helps him maintain a steady income stream from the complete artificial intelligence industry.
