Jove Enterprise Raises RMB 850M to Expand AI Server PCB Production Lines

Jove Enterprise Plans Major AI PCB Capacity Expansion

Shenzhen Jove Enterprise Limited (300814.SZ), a Chinese PCB manufacturer specializing in high-end boards for computing infrastructure, disclosed a private placement plan on May 29, 2026 targeting specific investors. The company aims to raise up to RMB 850 million (approximately USD 125.6 million) by issuing no more than 57.43 million shares to up to 35 investors.

The proceeds will be used for three primary purposes:

1. Expansion and renovation of the Heshan Jove AI PCB production line — new manufacturing equipment for high-layer-count boards
2. Digital upgrade project — Industry 4.0 systems for process monitoring and yield optimization
3. Working capital supplement — supporting raw material procurement for expanded capacity

Targeting AI Server Supply Chains

The fundraising directly targets rising demand for high-end PCBs driven by AI computing infrastructure. Jove specifically cited acceleration in:

• AI server power supply PCBs — multilayer boards handling 2-5 kW power delivery per GPU module
• High-speed communication PCBs — low-loss material boards for 400G/800G data center networking

The company confirmed it has already entered the AI server supply chains of major customers including:

These relationships position Jove as a beneficiary of the ongoing AI server buildout by hyperscalers (Microsoft, Meta, Google, Amazon) who contract through these tier-1 suppliers.

Investment Details

The private placement structure:

• Maximum raise: RMB 850 million (~USD 125.6 million)
• Maximum shares: 57.43 million new shares
• Maximum investors: 35 targeted institutional investors
• Primary use: Heshan facility AI PCB line expansion
• Expected capacity addition: Not disclosed, but industry analysts estimate 30-50% revenue capacity increase

AI Server PCB Requirements

The PCBs that Jove manufactures for AI server applications represent some of the most demanding specifications in the industry:

Power supply boards:

• 10-16 layers with 3-4oz copper weight
• Current capacity: 200-500A per power stage
• Thermal management: embedded copper slugs, heavy thermal vias
• Materials: high-Tg FR-4 (Tg > 170°C) for thermal cycling reliability

Server motherboards:

• 16-28 layers with mixed dielectric stackup
• Low-loss material (Megtron 6 or equivalent) for 56 Gbps signal integrity
• Impedance tolerance: ±7% for high-speed SerDes channels
• Via-in-pad with precision backdrilling throughout
• Large format: 400mm × 500mm+ panel utilization

Networking boards:

• 12-20 layers on ultra-low-loss material (Df < 0.003)
• Controlled impedance for 112 Gbps PAM4 channels
• Sequential lamination for dense BGA breakout
• Rogers/FR-4 hybrid stackups for RF sections

Industry Context: PCB Manufacturers Pivot to AI

Jove’s expansion mirrors a broader trend of PCB manufacturers repositioning their capacity toward AI infrastructure. Recent similar announcements include:

• Aohong Electronics — Thailand expansion targeting AI computing PCBs (May 2026)
• Kinsus Interconnect — AI-driven substrate demand extending visibility to 2029
• WUS Printed Circuit — Dedicated AI server PCB lines in Malaysia
• Shennan Circuits — Capacity expansion for high-layer server boards

The common thread: AI server demand is growing 40-60% annually, far outpacing the overall PCB market growth of 5-8%. Manufacturers without AI-focused capacity face market share erosion to those investing aggressively.

What This Means for PCB Buyers

The concentration of PCB capacity toward AI creates both opportunities and risks for other segments:

Opportunities:

• More advanced manufacturing capability becoming available (higher layer counts, finer features)
• Investment in digital upgrades improves overall quality for all customers
• Competition among manufacturers keeps pricing competitive

Risks:

• Lead time pressure as factories prioritize large AI orders
• Material competition (low-loss laminates allocated to AI programs first)
• Engineering resources focused on AI applications may reduce support for smaller orders

For hardware companies designing AI-adjacent products (networking, power, edge computing), now is the time to qualify PCB suppliers with demonstrated AI manufacturing capability — these suppliers have the equipment, materials, and process expertise your high-performance designs require.

At AtlasPCB, we manufacture the same classes of high-performance PCBs that AI servers demand: 20-40+ layers, low-loss materials, heavy copper, and precision impedance control. Our capacity serves a diverse customer base across AI infrastructure, telecommunications, and aerospace — providing the quality standards of AI manufacturing for projects of all sizes.

Source: IC&PCB Union, May 29, 2026

Image: Alex Knight via Unsplash

Need high-performance PCBs for AI or data center applications? AtlasPCB delivers 20-40+ layer boards on low-loss materials with impedance control and heavy copper — the same specifications demanded by AI server programs. Get a quote →

Related Articles

• AI-Driven Substrate Demand: Kinsus Sees Visibility to 2029
• PCB High-Speed Material Dk/Df Comparison Guide
• Heavy Copper PCB Design for Power Electronics