Computex 2026: Pegatron Accelerates the Future of AI Factories with Nvidia

Key Takeaways

• Pegatron unveiled Nvidia Vera Rubin NVL72 liquid‑cooled supercomputer.
• System integrates 72 GPUs, 36 CPUs, NVLink 6, and BlueField‑4 DPUs.
• Claims 10× token efficiency and 50% faster workloads versus prior generation.
• Introduced DSX‑aligned AI factory workflow using SimReady digital twins.
• Pegaverse platform delivers continuous AI‑driven manufacturing intelligence.

Pulse Analysis

Pegatron’s partnership with Nvidia marks a strategic shift in how AI hardware is packaged and delivered. By centering its portfolio on the Vera Rubin platform, the company taps into Nvidia’s third‑generation MGX architecture, which promises tighter integration of GPUs, CPUs, and DPUs. This convergence reduces latency and simplifies rack‑scale deployment, a critical advantage for enterprises racing to operationalize generative AI models that demand massive compute and memory bandwidth. Pegatron’s liquid‑cooled NVL72 also addresses the escalating power and thermal challenges that have plagued earlier Blackwell‑based systems.

The technical specs of the NVL72 underscore its performance edge. With 72 Rubin GPUs and 36 purpose‑built Vera CPUs linked via NVLink 6, the system can train mixture‑of‑experts models using only a quarter of the GPUs required by legacy setups, slashing token costs by a factor of ten and delivering ten times more tokens per megawatt. Nvidia’s Quantum‑X800 InfiniBand and Spectrum‑X Ethernet further ensure that data moves at petabit‑scale speeds, enabling real‑time inference for trillion‑parameter models. For data‑center operators, these efficiencies translate into lower total cost of ownership and faster time‑to‑value for AI initiatives.

Beyond hardware, Pegatron’s integration of Nvidia DSX and SimReady digital twins introduces a validation layer rarely seen in server manufacturing. By simulating thermal, power, and workload dynamics in the Omniverse environment, customers can pre‑emptively identify bottlenecks and optimize cooling before physical installation. The Pegaverse platform extends this capability with continuous telemetry and AI‑driven analytics, turning operational data into actionable insights. Together, these tools lower deployment risk, accelerate scaling, and set a new benchmark for AI‑factory readiness in a market where speed and reliability are paramount.