7.1 MW of NVIDIA GB200. Delivered in 143 Days.
Operational in Jakarta.
In 2025, BDx Data Centers commissioned 7.1 MW of NVIDIA GB200 AI infrastructure in Jakarta, achieving Ready-for-Service (RFS) in just 143 days from mobilization.
Built for extreme-density AI workloads and certified under the NVIDIA DGX™-Ready Data Center Program, CGK3/3A supports measured peak loads of 192 kW per rack using direct-to-chip liquid cooling.
While conventional AI data center builds take 18–30 months, BDx delivered GB200-ready capacity in under five.
Proven Speed-to-Power
143 days from mobilization to RFS
7.1 MW of GB200 IT capacity delivered
192 kW* per rack sustained peak density
Week 16 Early Access milestone achieved
Zero safety incidents during energization
This outcome was driven by disciplined, parallel execution across engineering, procurement, construction, commissioning, and NVIDIA certification — fully integrated from project mobilization.
What You’ll Learn in the Full Case Study
- How BDx compressed an 18–30 month delivery model into 143 days
- GB200 liquid cooling architecture at 192 kW* per rack
- Commissioning strategy (Cx2–Cx5) under live-system conditions
- Long-lead equipment mitigation and procurement acceleration
- Risk governance and executive-level decision management
- Lessons for hyperscale and enterprise AI infrastructure in APAC
If you are planning high-density AI deployments in APAC, this case study provides a practical blueprint for accelerated execution.
* Peak electrical load per rack under full processing conditions (EDPp1 measurement basis).
Download the CGK3/3A AI Infrastructure Case Study
Discover how BDx delivers AI-ready capacity at the speed the market demands.
