• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Shuttle Unveils New DN11-Series, High-Performance AI-PCs for Demanding Applications

GFreeman

News Editor
Staff member
Joined
Mar 6, 2023
Messages
1,849 (2.40/day)
We are excited to announce the upcoming availability of the new Shuttle DN11-Series, featuring the DN11H5, DN11H7, and DN11H9. These AI-powered, high-performance barebone PCs are designed to meet the needs of modern business, industrial, and creative applications.

Powerful AI Performance in a Compact Form
At the heart of the DN11-Series are Intel Core Ultra processors, offering a significant performance boost with their hybrid core architecture, Intel Arc graphics, and a dedicated NPU (Neural Processing Unit) for AI acceleration. These systems provide up to 34 TOPS of AI performance, making them ideal for machine learning, data analysis, multimedia processing, and industrial automation.



The DN11H5 is an efficient AI-PC with solid performance for standard computing tasks and AI applications, while the DN11H7 delivers enhanced multi-core performance, making it well-suited for professional workloads. The DN11H9, with its higher base TDP and superior power range, is ideal for users requiring maximum AI processing and graphics performance.



Exceptional Connectivity & Expansion Options
The DN11H-Series stands out with its quad 2.5 Gbit/s LAN ports, allowing seamless integration into high-speed networks, firewall setups, or virtualized environments. It also supports up to four UHD displays via two HDMI 2.1 and two DisplayPort 1.4a outputs, ensuring crystal-clear visuals for data visualization, surveillance, and content creation.

Additionally, each model includes:
  • Three M.2 slots for high-speed NVMe SSD storage with RAID support
  • Up to 96 GB DDR5 memory for maximum multitasking capability
  • Multiple USB 3.2 Gen 2 (10 Gbit/s) and USB-C Gen 2x2 (20 Gbit/s) ports for fast data transfer and power delivery
  • Optional 4G/5G connectivity for remote applications

Tailored to Your Needs
Whether you need an AI-powered computing solution, a high-performance networking gateway, or an edge computing platform, the DN11-Series adapts to various professional environments. Its compact 1.35-liter metal chassis supports VESA mounting, making it suitable for space-saving installations in offices, industrial setups, or embedded applications.

The DN11H5, DN11H7, and DN11H9 will be available soon through Shuttle's official partners and online retailers. Stay tuned for more details on availability and pricing!

View at TechPowerUp Main Site | Source
 
Joined
Feb 12, 2025
Messages
29 (0.46/day)
Location
EU
Processor AMD 5600X
Motherboard ASUS TUF GAMING B550M-Plus WiFi
Cooling be quiet! Dark Rock 4
Memory G.Skill Ripjaws 2 x 32 GB DDR4-3600 CL18-22-22-42 1.35V F4-3600C18D-64GVK
Video Card(s) Sapphire Pulse RX 7800XT 16GB
Storage Kingston KC3000 2TB + QNAP TBS-464
Display(s) LG 35" LCD 35WN75C-B 3440x1440
Case Kolink Bastion RGB Midi-Tower
Power Supply Enermax Digifanless 550W
Mouse Razer Deathadder v2
Benchmark Scores phi4 - 42.00 tokens/s
When running 4 byte parameter pre-trained models aka inferencing, about 3 CPU cores worth of compute will saturate RAM bandwidth. Rest of the cores and Xe AI accelerators give no meaningful performance increase here. Maybe running you LLM on Xe accelerators will consume less power, but ultimately all those X FLOPS of performance is capped by bandwidth of two DDR5 memory channels. I very much like the idea of AI/LLM focused hardware, but unfortunately this not the hardware design for serious AI/LLM enthusiast.
 
Top