Resource Center - Success Stories

MiTAC Computing and DDN Deliver High-Performance AI Storage Infrastructure

MiTAC Computing and DDN Deliver High-Performance AI Storage Infrastructure

MiTAC Computing and DDN have partnered to deliver turnkey, end-to-end AI solutions, from data processing, model training, inferencing, RAG, and AI agents, helping enterprise customers build faster, more scalable infrastructure for demanding, data-intensive workloads.

At the center of the solution is the DDN Infinia™ data engine paired with MiTAC’s high-density server platforms—including the MiTAC GC68C-B8056 1U storage server and high-density GPU racks powered by either NVIDIA RTX PRO™ GPUs or AMD Instinct™ GPUs. Together, they eliminate end-to-end data bottlenecks, minimize data movement, and maximize GPU utilization across modern AI environments.

Customer Challenge

AI, machine learning, and HPC workloads depend on fast, reliable access to data, and storage bottlenecks frequently limit the performance of even the most powerful compute environments. Addressing the intensive data demands of multimodal RAG pipelines, enterprise LLM inference, and agentic AI requires an infrastructure that can keep GPUs saturated while maintaining predictable performance under heavy multi-tenant demand. As enterprise AI moves from assistive generative AI toward task-based and persistent AI agents, infrastructure must handle far more retrieval and inference calls per user request.

To achieve instant responses and lower total cost of ownership, organizations need an integrated AI infrastructure that can:

  • Provide ultra-low latency document retrieval for instant AI inferencing responses and responsive multi-step AI agents.
  • Sustain massive throughput under pressure without starving high-performance GPUs, even as multiple coordinated agents access shared data at once.
  • Scale seamlessly from server racks to clusters without adding integration complexity.

The Joint Solution: Dense Infrastructure Meets DDN Infinia™

The collaboration merges MiTAC’s high-density hardware foundation with the DDN Infinia data engine to deliver a turnkey solution for AI data lakes, inference pipelines, agentic AI workflows, and real-time analytics.

Hardware Architecture: Storage and Dense AI Compute Rack-scale Solutions

  • MiTAC GC68C-B8056 Storage Server: A high-density 1U storage platform built for rapid data ingestion and high-speed storage deployment. Powered by a single AMD EPYC™ 9005 Series processor, it features 24 DDR5 DIMM slots, 12 hot-swap NVMe U.2 drive bays, PCIe 5.0 expansion, dual NVMe M.2 boot slots, and OCP v3.0 networking—making it exceptionally well suited for data-heavy AI applications.

MiTAC GC68C-B8056 Storage Server

  • MiTAC GC68C-B8056 Storage Rack: Engineered as an optimal configuration to run the DDN Infinia data engine, the integrated storage rack features six GC68C-B8056 storage servers paired with one dedicated GC68C-B8056 management server.

MiTAC GC68C-B8056 Storage Rack

  • MiTAC AI Compute Racks: Built for enterprise AI applications, the solution supports scalable GPU racks configured with either NVIDIA RTX PRO™ GPUs (paired with MiTAC’s G4827Z5 MGX™ AI servers) or AMD Instinct™ GPUs (paired with MiTAC’s G4826Z5 or G8825Z5 high-density, liquid- or air-cooled AI servers)—providing maximum computational density and memory bandwidth for real-time inference, RAG, multimodal AI processing, and agentic AI.

NVIDIA RTX PRO GPU Rack

NVIDIA RTX PRO™ GPU Rack

AMD Instinct GPU Rack

AMD Instinct™ GPU Rack

Software Platform: DDN Infinia™

  • Eliminates Bottlenecks: Eliminates end-to-end AI data bottlenecks and maximizes GPU performance by minimizing data movement and delivering sub-millisecond document retrieval.
  • Workload Acceleration: Purpose-built to accelerate turnkey AI inference, multimodal RAG pipelines, and agentic AI.
  • Seamless Scalability: Software-defined architecture allows flexible expansion across racks and clustered deployments to grow with evolving enterprise requirements.
  • Enhances GPU/Token Productivity: By leveraging KV caching, the DDN Infinia™ solution offloads context from GPU HBM, eliminating GPU re-computation, accelerating TTFT, and increasing token throughput across distributed deployments, giving persistent AI agents fast access to long-running context and memory.

Results and Value

“DDN and MiTAC Computing bring complementary strengths together to help customers deploy AI infrastructure with greater speed, efficiency, and confidence. That combination is especially important as data-intensive workloads continue to grow in complexity and scale.” – Andrew Murphy, Sr. Director, Product Management, DDN

“By combining our high-density storage platform with DDN’s Infinia capabilities, we can help customers build AI infrastructure that is fast, scalable, and ready for data-intensive workloads.” – Raymond Huang, General Manager, MiTAC Computing

Together, MiTAC and DDN help customers simplify deployment, improve storage throughput, and build a more scalable foundation for AI and data-centric workloads:

  • Maximized GPU Utilization: Minimizes data movement so high-performance GPUs remain continuously saturated, maximizing return on investment.
  • Ultra-Low Latency: Instant document retrieval enables rapid AI inference responses and responsive AI agents for real-time enterprise applications.
  • Reduced Risk & Faster Time-to-Market: Validated hardware architecture paired with optimized storage software simplifies deployment and reduces integration complexity.
  • Future-Proof Growth: A flexible, high-density platform that grows effortlessly alongside expanding AI, analytics, and data lake requirements, from today’s generative AI and RAG to tomorrow’s agentic AI.

 

SUBSCRIBE NOW to get the latest news.

Clicking "OK" confirms your acceptance of our Terms of Use, Privacy Policy, and Cookie Policy.