We are the Frontier
U.S.

HP Expanded Local AI Computing Portfolio

The company announced new workstation and laptop configurations designed to offload AI inference from the cloud.

Updated on Oct. 7, 2026 in Artificial Intelligence

Close-up of a matte-black professional workstation chassis showing intricate metal cooling fins on a concrete workspace.
HP has launched the ZGX Fury workstation, a new hardware configuration designed to process generative AI workloads locally, reducing dependency on cloud infrastructure. AI Illustration. Upload story photo >

Live Poll

Would you prefer using local AI-enabled PCs over cloud-based AI services for your daily tasks?

HP has announced an expansion of its AI computing hardware, featuring the high-performance ZGX Fury workstation powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip. This move marks an effort to shift AI workload development and deployment to local infrastructure.

Why it matters

By providing local AI hardware, HP aims to reduce enterprise and consumer reliance on cloud computing for generative AI tasks. This strategy targets businesses and developers seeking greater control over data processing for creative and analytical workflows.

The HP ZGX Fury integrates the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, while the new RTX Spark platform combines NVIDIA AI, RTX graphics, and unified memory. These systems are designed specifically for local inference tasks.

The players

HP

A global hardware manufacturer specializing in enterprise workstations, consumer PCs, and professional computing infrastructure.

NVIDIA

A semiconductor company that designs high-performance graphics processing units and AI-optimized computing architectures.

The details

The ZGX Fury workstation utilizes NVIDIA's architecture to handle machine learning inference locally, bypassing the latency and connectivity requirements of cloud servers. RTX Spark integrates the graphics processing unit, AI acceleration cores, and shared memory architecture directly into the Windows environment. This unified memory approach allows the system to access high-speed data buffers, which is necessary for processing large language models on local hardware.

Timeline

  1. October 7, 2026: HP announced its AI solution portfolio expansion.

  2. October 16, 2026: The HP OmniBook Ultra 16 and X 14 are expected to be available for purchase.

  3. November 2026: Perplexity Pro trial offers will become available on consumer devices.

The Tech Race

This development follows the industry-wide trend of incorporating the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip to shift AI workloads from server clusters to local hardware. It marks a departure from cloud-dependent model inference toward edge-based enterprise processing.

Business users and developers can purchase the new OmniBook models starting October 16, 2026, to run AI tasks without cloud dependency. Additionally, users will gain access to a one-month trial of Perplexity Pro on consumer devices beginning in November 2026.

The takeaway

HP is moving to capture the professional AI market by prioritizing local hardware capabilities over cloud-based infrastructure. Watch for actual performance benchmarks to see if these systems can effectively handle the high-memory requirements of modern local AI models.

Further reading

For more on how local hardware affects processing, see the Artificial Intelligence section.

Live Poll

Would you prefer using local AI-enabled PCs over cloud-based AI services for your daily tasks?