Full Breakdown
Groq Expands AI Chip Production with Samsung Electronics Amid Market Demand Surge
3/10/2026, 8:25:38 PM
Increased Production Orders from Groq
Groq, an AI chip startup recently acquired indirectly by NVIDIA, has requested Samsung Electronics Foundry to increase its wafer production for inference AI chips from approximately 9,000 to 15,000 units. This increase reflects a growing demand for AI chips as Groq prepares to transition from producing sample chips to large-scale commercial production in 2026. The collaboration is seen as a strategic move for Samsung to establish a foothold in the inference AI chip market, which is gaining traction as industry interest escalates.
Context of the Semiconductor Market
The semiconductor industry is currently experiencing a surge in demand for AI chips, particularly those optimized for inference applications. Groq's decision to boost production aligns with broader market trends, as NVIDIA is expected to unveil an inference-optimized chip based on Groq's designs at the GTC 2026 event. This new chip may utilize static random-access memory (SRAM) instead of the high-bandwidth memory (HBM) typically used in traditional AI chips, potentially enhancing performance and reducing costs.
Implications for Samsung's Foundry Business
Samsung's foundry division is actively pursuing contracts to enhance its profitability and competitive position against rivals like TSMC. The production of Groq's AI chips will utilize Samsung's advanced 4nm process technology, which has been improved to boost performance. This strategic partnership is critical for Samsung as it seeks to capitalize on the growing demand for 4-5nm processes in the semiconductor market.
Delays Impacting Other Projects
While Samsung Foundry is set to benefit from Groq's increased orders, delays in multi-project wafer (MPW) production linked to Tesla may disrupt other projects. Reports indicate that Tesla has postponed its MPW production, which could delay the mass production of DeepX's next-generation neural processing unit (NPU). This NPU, designed for Samsung's 2nm process, is crucial for various applications, including autonomous vehicles and humanoid robots. The specific reasons for Tesla's delay have not been disclosed, but industry observers suggest that it may be related to broader production timelines and investment schedules.
Official Statements & Responses
Industry analysts emphasize the importance of Groq's increased orders for Samsung, noting that the collaboration could significantly impact Samsung's foundry business. Additionally, Tesla has recently engaged with Samsung to discuss expanding production of its 2nm AI6 chip, which could increase wafer orders from 16,000 to 40,000 units per month if finalized.
Criticism & Opposition
Despite the positive outlook for Groq and Samsung, some industry experts express concerns about the potential impact of Tesla's delays on the semiconductor supply chain. The postponement of MPW production could hinder the timely rollout of critical technologies in the AI and automotive sectors.
Verbatim Quotes
- “The combined momentum from Nvidia's AI chip initiatives and Groq's expanded production signals robust growth ahead for the inference AI chip market.” — Industry Analyst
- “Given the higher cost of this node and strong industry demand for 4-5nm processes, securing such orders is strategically important for Samsung as it competes with TSMC.” — Semiconductor Expert
- “The specific reason for Tesla’s MPW delay has not been disclosed, the report notes.” — Industry Report
This evolving landscape in the semiconductor industry highlights the interplay between emerging AI technologies and the production capabilities of leading foundries like Samsung Electronics.
