Loading article…
Pegatron partners with Together AI and 5C to ship NVIDIA GB300 NVL72 and HGX B200 racks, delivering L12 capacity across Texas and Maryland data centers.
Pegatron announced a three‑way partnership with Together AI and 5C to manufacture and deploy NVIDIA‑based AI servers—including the RA4802‑72N2 on the GB300 NVL72 and the AS401‑2T0‑8H1 on the HGX B200—in U.S. data centers, a move that adds an L12 total‑solution capacity and promises faster, localized service for American AI workloads [1].
| At a glance | |
|---|---|
| Collaboration | Pegatron + Together AI + 5C |
| Systems launched | RA4802‑72N2 (NVIDIA GB300 NVL72) & AS401‑2T0‑8H1 (NVIDIA HGX B200) |
| Deployment sites | Texas (manufacturing) & Maryland (liquid‑cooled rack) |
| Solution capacity | L12 total‑solution |
The RA4802‑72N2 servers, built on NVIDIA’s GB300 NVL72 accelerator, are being produced in Texas, reinforcing Pegatron’s domestic manufacturing footprint and aiming to cut latency for U.S. customers. Meanwhile, the AS401‑2T0‑8H1 racks, featuring liquid‑cooled HGX B200 modules with in‑row coolant distribution units, have been installed at a Maryland data center alongside air‑cooled AS800‑2T0‑8H1 units. Together, these deployments illustrate a hybrid cooling strategy that balances performance density with energy efficiency.
Pegatron’s end‑to‑end offering—spanning rack‑level system engineering, networking, and cooling integration—places it among a limited set of providers capable of delivering full AI‑factory solutions at scale. By coupling NVIDIA’s latest accelerators with 5C’s 2 GW of planned data‑center capacity, the partnership targets the same high‑performance segment that rivals such as Dell EMC and HPE are courting with their own NVIDIA‑powered clusters. The L12 capacity metric signals a sizable deployment, though the press release does not disclose exact GPU counts, making direct performance comparison with competitors pending.
Pegatron’s move underscores a broader shift toward domestically produced, tightly integrated AI hardware, suggesting that future AI workloads will increasingly rely on specialized, locally serviced infrastructure rather than generic cloud resources. The partnership’s success will hinge on how quickly the L12 capacity can be expanded to meet rising demand.
Coverage is mostly measured — 10 of 10 reports stay neutral.
Every Monday — the token unlocks, Fed dates & catalysts set to move crypto and markets this week. So you’re never blindsided.
Free · 3-min read · one-click unsubscribe
AI-assisted synthesis by the TrendWatcher Editorial Desk · sourced from 2 outlets · Jul 4, 2026 · How we report
DLSS 5 is planned for release in the fall of 2026 on Nvidia's RTX 50 series GPUs.
The collaboration uses Nvidia GB300 NVL72, HGX B200, and HGX B300 systems, with both liquid‑cooled and air‑cooled rack configurations.
They target workloads such as climate modeling, computational fluid dynamics, energy exploration, brain mapping, and AI research for 6G networks.
DLSS 5 uses a neural network to process color and motion vectors, replacing standard lighting models while preserving geometry, textures, and materials to achieve photorealistic illumination.
Pegatron will showcase its solutions at Supercomputing 2025 (SC25) in St. Louis, Missouri.