Qualcomm Dragonfly AI200
Active
Qualcomm Dragonfly™
AI200
Contact Sales
chip image

Purpose-built rack-scale AI inference with the industry's highest memory capacity per accelerator¹ — engineered for generative and agentic AI at scale.

Qualcomm Dragonfly™ AI200 is a purpose-built, rack-scale AI inference accelerator engineered for the generative and agentic AI era. Each AI200 card delivers an industry-leading 768 GB of LPDDR memory1, while a single 140 kW liquid-cooled ORv3-compliant rack provides industry-leading 43 TB of memory capacity1  — enabling efficient inference up to 10T parameter models with compelling total cost of ownership (TCO).

Product license agreement

Benefits

default icon alt text
Best-in-class Memory Capacity¹
With 768 GB per card and 43 TB per rack, AI200 serves models from 7B up to 10 trillion parameters, supporting long-context, RAG, and agentic workloads without compromise.
default icon alt text
Rack-Scale Deployment
A fully integrated OCP ORv3-compliant, direct-liquid-cooled rack with cableless backplane simplifies deployment, accelerates time-to-production, and improves reliability across enterprise and hyperscale data centers.
default icon alt text
Leading Total Cost of Ownership
Purpose-built for inference economics, AI200 delivers compelling performance-per-dollar, driving down cost-per-token for generative and agentic AI workloads.
default icon alt text
Complete Inference Software Stack
The Qualcomm® AI Inference Suite provides end-to-end tooling for model deployment as bare-metal, cloud VM, or inference-as-a-service — reducing complexity and accelerating time-to-value.
default icon alt text
Direct Liquid-Cooled
A 140 kW direct-liquid-cooled OCP ORv3-compliant rack maximizes compute density and energy efficiency, lowering power usage effectiveness (PUE) and enabling sustainable AI infrastructure at scale.

1Compared to commercially-available competing published product specifications.

Features

  • Rack-scale solution purpose-built for AI inference economics 
  • 768 GB LPDDR memory per card — highest per-accelerator capacity in the industry1
  • Industry-leading 43 TB memory capacity per rack1
  • Supports SOTA models up to 10T parameters with context lengths up to 128K 
  • PCIe Gen6 scale-up; Ethernet-with-RoCE scale-out 
  • Qualcomm® AI Inference Suite for bare-metal, VM, or IaaS deployment 
  • Designed to accelerate various model categories - reasoning, vision, agentic, LLM, text-to-image, and video processing 
  • Air and direct liquid cooling
  • OCP ORv3-compliant rack with cableless backplane
Specifications
Rack
Form Factor
Single Wide, Open Rack v3 (ORv3) compliant
Number of Cards
56
Memory
Capacity
43 TB
Effective Bandwidth
0.414 PB/s1
Scale-Up
Interface Type
PCIe 6.0
Scale-Out
Interface Type
Ethernet w/RoCE
Thermal Management
Cooling
Direct Liquid Cooling (DLC), Air Cooling
Thermal Design Power
140 kW
  1. With LPDDR5x

QUALCOMM DRAGONFLY AI INFRASTRUCTURE MANAGEMENT SUITE

Orchestrate and
scale efficiency

The suite provides provisioning, monitoring, orchestration and fault handling across rack‑scale deployments. Together, hardware, connectivity and software form the foundation of a cohesive data center platform approach — one designed to scale with customers as AI workloads evolve. 

Qualcomm relentlessly innovates to deliver intelligent computing everywhere, helping the world tackle some of its most important challenges. Our leading-edge AI, high performance, low-power computing, and unrivaled connectivity deliver proven solutions that transform major industries. At Qualcomm, we are engineering human progress.

Stay connected

Get the latest Qualcomm and industry information delivered to your inbox.

Subscribe
Manage your subscription
Also of Interest

© Qualcomm Technologies, Inc. and/or its affiliated companies.

Snapdragon and Qualcomm branded products are products of Qualcomm Technologies, Inc. and/or its subsidiaries. Qualcomm patented technologies are licensed by Qualcomm Incorporated.

Note: Certain services and materials may require you to accept additional terms and conditions before accessing or using those items.

References to "Qualcomm" may mean Qualcomm Incorporated, or subsidiaries or business units within the Qualcomm corporate structure, as applicable.

Qualcomm Incorporated includes our licensing business, QTL, and the vast majority of our patent portfolio. Qualcomm Technologies, Inc., a subsidiary of Qualcomm Incorporated, operates, along with its subsidiaries, substantially all of our engineering, research and development functions, and substantially all of our products and services businesses, including our QCT semiconductor business.

Materials that are as of a specific date, including but not limited to press releases, presentations, blog posts and webcasts, may have been superseded by subsequent events or disclosures.

Nothing in these materials is an offer to sell or license any of the services or materials referenced herein.