Key Highlights
- At GTC 2026, Nvidia introduced the Groq 3 LPU, a specialized processor designed exclusively for AI inference operations
- The company’s new LPX server configuration features 128 Groq 3 chips and delivers up to 35x improved throughput per megawatt when combined with Vera Rubin NVL72
- A standalone Vera CPU rack system was announced, marking Nvidia’s direct entry into competition with Intel and AMD’s data center processors
- The Vera processor targets agentic AI applications, including web navigation and file data retrieval operations
- Fiscal 2026 data center revenue reached $193.5 billion for Nvidia, representing significant growth from the previous year’s $116.2 billion
During Monday’s GTC 2026 conference in San Jose, Nvidia demonstrated its strategic expansion beyond traditional GPU manufacturing with an array of innovative chip designs and server architectures.
The star announcement centered on the Groq 3 language processing unit, commonly known as an LPU. Following a substantial $20 billion transaction finalized last December, Nvidia acquired licensing rights to Groq’s technology and welcomed key personnel including founder Jonathan Ross and president Sunny Madra to the team.
Groq 3’s primary function addresses inference workloads — the operational phase following AI model training. Each interaction where users submit queries to AI chatbots and receive responses represents inference in action. This rapidly expanding AI market segment benefits significantly from purpose-built processors that outperform general-purpose GPU solutions.
According to Ian Buck, Nvidia‘s VP of hyperscale and HPC, the Groq 3 features superior memory speed compared to the company’s GPU offerings, despite having less total memory capacity. The strategy involves leveraging the complementary advantages of both technologies.
This hybrid approach materializes in the LPX server rack — a configuration containing 128 Groq 3 LPU processors. When deployed alongside the Vera Rubin NVL72 rack system, customers can achieve 35x enhanced throughput per megawatt alongside 10x increased revenue potential. The architecture specifically targets trillion-parameter models and million-token context capabilities.
Vera CPU Enters Intel and AMD Territory
The second major announcement featured the Vera CPU rack system. While previous discussions positioned Vera as a component within the Vera Rubin superchip — pairing one Vera CPU with dual Rubin GPUs — Nvidia now offers Vera as an independent processor solution.
This rack configuration integrates 256 liquid-cooled Vera processors into a unified system. Nvidia markets it as the optimal CPU for agentic AI applications — autonomous systems capable of web browsing, file data extraction, and executing complex multi-step operations independently.
“We’ve designed a new kind of CPU, the Olympus core, engineered by NVIDIA for AI execution,” Buck stated. Beyond AI execution, Vera handles data mining, personalization functions, and context analysis that supports AI model operations.
This strategic move positions Nvidia as a direct competitor to Intel and AMD within the data center CPU marketplace — a sector these two manufacturers have controlled for decades.
Just last month, Nvidia secured an agreement with Meta for large-scale deployment of its previous-generation Grace CPU architecture — representing the largest deployment of its kind. The Vera introduction amplifies this competitive positioning.
Additional Hardware Announcements From GTC
Nvidia’s presentation also featured the Bluefield-4 STX storage rack system and Spectrum-6 SPX networking rack, completing a comprehensive data center hardware portfolio.
Major hyperscale providers including Amazon, Google, Meta, and Microsoft plan combined AI infrastructure investments totaling $650 billion throughout this year.
Nvidia’s data center segment generated $193.5 billion in fiscal 2026 revenue, marking substantial growth from fiscal 2025’s $116.2 billion figure.
Wall Street analysts maintain a consensus Strong Buy rating for NVDA, with 38 Buy ratings and one Hold recommendation issued over the past three months. The average price target stands at $273.61.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
