At COMPUTEX 2026, Hewlett Packard Enterprise announced a major expansion of its enterprise server lineup with the debut of the HPE ProLiant Compute DL394 Gen12. Engineered in collaboration with NVIDIA and data platform provider Redpanda, the new server is powered by the highly anticipated NVIDIA Vera CPU. The system is custom-built to support the enterprise transition from traditional generative AI models toward autonomous agentic AI, complex financial modeling, and scalable reinforcement learning within modern AI factories.
As organizations scale their AI operations, sequential logic and data-heavy processing require unprecedented single-core performance and ultra-high memory bandwidth. The HPE ProLiant Compute DL394 Gen12 addresses these infrastructure challenges by combining hardware acceleration with advanced security features and cloud-native server management. The ecosystem is currently being explored by major financial institutions, including the New York Stock Exchange (NYSE), to future-proof core transaction infrastructures.
“With our new HPE ProLiant Compute server, we are delivering a new class of infrastructure to help customers accelerate insights and operate with confidence in the most demanding environments,” said Antonio Neri, president and CEO of HPE.
Engineering a Shift to Agentic AI and Real-Time Reasoning
The technological transition from passive generative AI models to interactive, multi-agent systems demands an architecture capable of processing massive data pipelines with deterministic execution. The HPE ProLiant Compute DL394 Gen12 provides a unified framework tailored for real-time reasoning, advanced data storage, active monitoring, and hardened security.
“The shift from generative models to agentic systems is redefining the role of compute across the enterprise,” said Antonio Neri, president and CEO of HPE. “These workloads require high-performance servers with exceptional CPU performance to enable real-time reasoning across agentic AI and financial services applications. With our new HPE ProLiant Compute server, we are delivering a new class of infrastructure to help customers accelerate insights and operate with confidence in the most demanding environments.”
For enterprise networks managing immense data volumes, the integration of specialized processing cores is critical to maintaining low latency and extreme reliability.
Also Read: SK hynix Introduces iHBM Thermal Architecture to Maximize AI Computing Performance
“At the NYSE, our focus is to optimize the latency, throughput, and reliability of the systems underpinning our unrivaled infrastructure,” said Lynn Martin, President of NYSE Group. “NYSE processes more than 1.1 trillion messages per day, and in collaboration with Redpanda and HPE, using NVIDIA Vera CPUs, we will be scaling our capacity while further optimizing latency to power a high-performance, resilient, and AI-ready market infrastructure.”
The architectural departure from legacy x86 designs allows enterprises to orchestrate workflows more efficiently, offering a sustainable path for high-density computing clusters.
“Agentic AI has arrived, and it needs a new CPU,” said Jensen Huang, founder and CEO, NVIDIA. “Vera was built to orchestrate AI factories delivering 2x the efficiency and faster task completion than x86. With HPE ProLiant Compute DL394 Gen12, enterprises can put Vera to work, and NYSE shows what purpose-built AI infrastructure can do in the world’s most demanding environments.”
Overcoming Multi-Processor Memory Bottlenecks
Traditional high-core-count processors often rely on complex chiplet designs that cause non-uniform memory access (NUMA) issues, which can induce variable latencies and inconsistent performance across multi-processor systems. The HPE ProLiant Compute DL394 Gen12 circumvents this hurdle using a 2U monolithic design.
Thanks to the use of LPDDR5X memory technology, which allows data processing rates up to 14 GB/sec for each separate core, the combined memory bandwidth of the chip is 1.2 terabytes per second. The use of such memory makes it possible for the chip to consume high-speed data feeds without difficulty, which allows the Vera CPU to be considered an efficient orchestrator of enterprise computing operations.
Silicon-Level Security and AI-Driven Server Management
To safeguard sensitive workloads within heavily regulated environments, HPE embeds hardware-level protection directly into the server framework via its Silicon Root of Trust firmware. Operating alongside HPE Integrated Lights-Out (iLO 7) technology, the platform utilizes a secure enclave to protect hardware assets across every phase of the equipment lifecycle. Notably, these next-generation servers are engineered to align with NIST’s quantum-computing resistant security standards, providing businesses with a future-proof defense against emerging cryptographic threats.
The operational efficiency of the servers is supported by HPE Compute Ops Management, which is an intelligent dashboard that makes use of AI for visualizing and automating the topology of distributed servers. The software can help IT teams save on overhead expenses through the automation of server updates and minimize revenue losses due to unexpected system downtime.
