As generative AI, autonomous driving simulation, and scientific computing become core engines of the digital economy, global data center and AI server deployments are accelerating at an extraordinary pace.
Rack power density has surpassed 20 kW, while single GPU modules are approaching kilowatt-level consumption. Behind these numbers lies an unprecedented challenge for power delivery infrastructure.
Systems must withstand higher transient currents, greater power density, and increasingly complex EMI environments — all while maintaining thermal control and long-term reliability.
In response, Microgate introduces the MPF cofiring power inductor series, built on a metallic composite magnetic powder material platform. Designed for high-density power zones in AI servers, data centers, GPU VRMs, and PoL architectures, the MPF series delivers optimized solutions for high current, high frequency, strong transient performance, and complex electromagnetic environments.
Across AI server and data center motherboards, GPU accelerator cards, and high-speed interconnect boards, the power delivery zone typically exhibits four defining characteristics:
Frequent GPU load transitions
High transient di/dt makes the inductor a critical variable in ripple control and overall voltage stability.
Widespread multiphase architectures
GPU VRMs are predominantly multiphase designs, amplifying component consistency requirements and increasing layout complexity.
Complex electromagnetic environment
High-speed signals, strong switching nodes, and dense copper routing make EMI behavior and thermal hotspots more sensitive.
High maintenance and reliability expectations
Data centers demand long-term operational stability. Even minor temperature rise or parameter drift can escalate into system-level risk.
The MPF series is engineered to serve as a more controllable inductor foundation for these demanding scenarios — supporting the high current density required by AI compute servers while addressing the high-frequency switching and strong transient conditions of GPU power delivery in data center platforms.
(Integrated multiphase inductors for GPU VRM and AI Server PoL applications)
microgate solution | size(mm) | L @ 1 MHz/1V (nH) | DCR (mΩ) | Max Isat (A) | Max Irms (A) |
MPF050560HFR12L**DLF | 4.8 * 4.8 * 5.8 | 120±15% | 0.235±10% | 45 | 45 |
MPF080560HF85NM**DLF | 7.8 * 4.8* 5.8 | 85±20% | 0.25±10% | 80 | 50 |
MPF100760HF85NM**DLF | 9.8 * 7.0* 5.8 | 85±20% | 0.16±10% | 75 | 55 |
Structural Feature: Integrated multiphase architecture.The multi-terminal design enables closer VRM placement and optimized current loop routing, improving layout efficiency in high-phase-count power stages.With a more compact footprint and layout-friendly phase positioning, this structure directly addresses real-world challenges in AI server and data center GPU power zones — tight board space, aggressive transient behavior, and heightened thermal sensitivity.
As AI servers, data centers, and GPU platforms continue to evolve, the role of inductors has shifted from simply meeting specifications to actively influencing system limits.
Microgate’s MPF cofiring power inductors, built on a composite magnetic material platform and integrated structural architecture, are engineered for high-density GPU VRM and PoL power zones.
They deliver a more efficient, more compact, more stable, and production-ready inductor solution — providing a stronger power foundation for next-generation AI Compute and High-Performance Computing platforms.