Specifications
Model
Prime Enterprise PX-3200
Form Factor
ATX 3.1 / PCIe 5.x ready
Output Class
3,200 W (240 V input)
115 V Output
Limited — full wattage requires 240 V mains
Efficiency Target
80 PLUS Platinum / Titanium class platform
AC Input
200–240 V AC for the full 3,200 W envelope
12 V Rail
Single high-current rail sized for multi-GPU systems
GPU Connector
12V-2x6 (12VHPWR) with transient headroom
Protection
OptiGuard — GPU transient / power-spike protection
EDPP
NVIDIA Electrical Design Point Peak validated on the 1600 W model
Cooling
135 mm fluid-dynamic-bearing fan, semi-passive below load threshold
Modularity
Fully modular cable set
Topology
LLC resonant + DC-DC VRM, Japanese capacitors
Operating Temperature
0 °C to 50 °C
Certifications
cTUVus · CE · CB · RoHS
Typical Use
AI training workstations, 4× GPU deskside servers, render nodes
Overview
The Prime Enterprise PX-3200 is Seasonic's enterprise ATX 3.1 platform aimed at the transient behaviour of modern AI accelerators. It adds OptiGuard, a protection scheme that rides out the large, fast microsecond-scale load steps a GPU can demand, and it carries the 12V-2x6 (12VHPWR) connector required by current-generation accelerators.
Note the input-voltage caveat common to all 3 kW-class ATX units: the headline 3,200 W figure is achieved on 240 V mains. On 115 V supplies the usable continuous output is substantially lower, which is why NVIDIA's EDPP validation is quoted for the 1600 W model. Verify the mains voltage at the deployment site before specifying.
Key Benefits
Sized for accelerator transients — OptiGuard + 12V-2x6, ATX 3.1 compliant for current GPU power behaviour, fully modular for short cable runs in dense chassis, and a semi-passive fan that stays quiet below the load threshold.
Applications
AI training and inference workstations, deskside 4-GPU servers, simulation and render nodes, and any ATX platform that needs >2 kW of sustained 12 V delivery on 240 V mains.
Request a Quote — SEASONIC PRIME ENTERPRISE PX-3200 — 3200 W ATX 3.1 SERVER POWER SUPPLY
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →