SoftBank Group Corp: Shackelford Data Center

The Shackelford Data Center is a large-scale AI and hyperscale cloud facility under development by SoftBank Group Corp., OpenAI, and Oracle Corporation as part of the global Stargate Project — an ambitious initiative to create a distributed 10-gigawatt (GW) network of AI-optimized data centers across the United States. Located in Shackelford County, Texas, this site is one of five initial Stargate facilities designed to power the next generation of artificial intelligence, machine learning, and supercomputing applications.

Positioned in the heart of West Texas, the Shackelford site benefits from the region’s vast land availability, abundant renewable energy potential, and strong electrical grid interconnectivity via the ERCOT network. Its geographic proximity to energy generation hubs, fiber networks, and renewable sources makes it ideal for scaling GPU-intensive compute infrastructure while meeting sustainability targets.

This development underscores SoftBank’s strategic transition from a global investment group to a foundational builder of AI and cloud infrastructure, aligning with its vision to enable global-scale computation through intelligent, sustainable, and distributed facilities.

⚙️ Facility Highlights

• Operator: Developed under the Stargate partnership (SoftBank Group Corp., OpenAI, and Oracle Corporation).
• Location: Shackelford County, Texas, United States.
• Facility Role: Core U.S. hyperscale and AI data center powering the Stargate distributed compute network.
• Power Capacity: Estimated hundreds of megawatts, contributing to the overall 10 GW Stargate buildout.
• Energy Source: Connected to the ERCOT grid, integrating solar and wind power for renewable operations.
• High-Density Compute: Purpose-built for AI and GPU servers exceeding 50 kW/rack, optimized for large-scale model training and inference.
• Cooling Systems: Advanced liquid cooling and heat reuse technologies engineered for high-performance computing.
• Facility Scale: Multi-building hyperscale campus with modular expansion capability.
• Design Standards: Tier IV-equivalent redundancy and fault-tolerant electrical design.
• Operational Timeline: Initial phases expected to go online by 2026, with full capacity scaling through 2030.

🔐 Security & Compliance

• 24/7 Facility Security: Continuous onsite and remote monitoring by integrated SOC and AI analytics systems.
• Perimeter & Access Control: Secure fencing, guarded entry, mantrap vestibules, and biometric authentication.
• AI-Assisted Surveillance: Predictive video analytics and anomaly detection integrated with SoftBank’s monitoring network.
• Power & Cooling Redundancy: Dual substations, N+N UPS systems, and multiple mechanical plants for critical load separation.
• Disaster Resilience: Designed to withstand Texas weather extremes, including high wind and temperature variation.
• Certifications (Planned):
• ISO/IEC 27001 — Information Security Management
• ISO 22301 — Business Continuity
• SOC 2 Type II — Data Security and Controls
• Uptime Institute Tier IV (Design Objective)
• Data Protection: Implements full encryption, network segmentation, and audit-compliant operational protocols.

🌐 Connectivity & Carrier Access

• Carrier-Neutral Facility: Multiple Tier 1 network providers with redundant fiber entrances.
• Regional Network Reach: Low-latency connectivity to Dallas, Austin, Houston, and other Stargate nodes.
• Stargate Backbone Integration: Optical mesh connectivity linking all Stargate sites — Ohio, New Mexico, Milam County, and a Midwest location — for synchronous AI workload distribution.
• Cloud Integration: Direct interconnect with Oracle Cloud Infrastructure (OCI) and distributed OpenAI compute clusters.
• Private Networking Options: Dedicated dark fiber, MPLS, and encrypted enterprise circuits available.
• Global Connectivity: Linked to SoftBank’s transpacific network, enabling Japan–U.S. data exchange for cross-regional AI operations.

👥 Who It Serves

The Shackelford Data Center is designed for the most compute-intensive organizations and next-generation workloads requiring massive scale and efficiency.

• AI & Machine Learning Companies: Large-scale training and inference for LLMs and generative AI models.
• Cloud & Hyperscale Providers: Partners deploying AI clusters and HPC environments through Oracle Cloud and SoftBank’s infrastructure.
• Telecommunications & Edge Providers: Expanding low-latency 5G and edge computing services across the southern U.S.
• Energy & Research Institutions: Utilizing HPC and AI compute for analytics, simulations, and smart energy management.
• Enterprise & Financial Sectors: Hosting private cloud infrastructure with secure, compliant colocation environments.
• Government & Defense Applications: High-assurance hosting for regulated or mission-critical systems.