NVIDIA launches Vera CPU purpose-built for agentic AI

NVIDIA introduced the Vera Central Processing Unit (CPU), a processor designed to enhance data processing and support for agentic Artificial Intelligence (AI) and reinforcement learning, aiming to improve efficiency and speed in computational tasks.

The deployment of Vera CPUs targets applications that require large-scale AI service responsiveness, such as coding assistants and agentic tools. The processor's development aligns with increasing demands on infrastructure to manage complex AI models controlling task planning and execution.

Vera incorporates 88 custom-designed cores with NVIDIA Spatial Multithreading, providing capabilities to handle simultaneous tasks within AI environments. It offers a memory subsystem based on LPDDR5X technology, doubling bandwidth and halving power consumption compared to general-purpose CPUs. Connectivity is enhanced by NVIDIA NVLink-C2C delivering 1.8 TB/s coherent bandwidth, supplementing system performance for GPU-accelerated workloads.

Collaborations involving cloud service providers and system manufacturers have been established to deploy Vera CPUs. Organizations such as Alibaba, Meta, Oracle Cloud Infrastructure, Dell Technologies, HPE, Lenovo, and Supermicro are involved in integrating Vera into server configurations suitable for reinforcement learning, agentic inference, and High performance computing (HPC).

Michael Truell, cofounder and CEO of Cursor, said, “We're excited to use NVIDIA Vera CPUs to improve overall throughput and efficiency so we can deliver faster, more responsive coding agent experiences for our customers.” Alex Gallego, founder and CEO of Redpanda, stated, “Redpanda recently tested NVIDIA Vera running Apache Kafka-compatible workloads and saw dramatically better performance than other systems we've benchmarked, delivering up to 5.5x lower latency.” John Cazes, director of HPC at the Texas Advanced Computing Center, said, “Vera’s per-core performance and memory bandwidth represent a giant step forward for scientific computing, and we look forward to bringing Vera-based nodes to our CPU users on Horizon later this year.”

The Vera CPU will be available through partners in the second half of the year, with configurations designed to address various AI, data processing, and HPC workloads.