Running AI
A new standard for edge computingAI inside Cube is an edge computer created to achieve any sort of AI operation through powerful GPU hardware and AI inside’s proprietary AI management OS and software.
Features
Ultimate AI performance
The hardware and software of the AI inside Cube series were simultaneously designed to deliver a balanced integrated system. To run the powerful GPU, thermal management and noise reduction measures are required and so, below the seamless design of the AI inside Cube featuring no visible fasteners, integrated cooling equipment, and an airflow control system keep levels of heat and noise extremely low.
Everything is simple
As with previous AI inside products, use of the AI inside Cube series starts simply with plug and play. Under the subscription model, activity begins from the day of arrival, without hardware selection, purchase, system integration, customization, AI learning, or verification experiments.
You alone handle the data
The AI inside Cube series is designed to let users control their own information. AI processing can be performed at any time without providing data to AI inside, while AI technology updates from AI inside are available with no special limitations.
Series
World-changing creativity “Pro”
The highest-performing edge computer in this series boosts your creativity and empowers you to make real the AI training and inference ideas you have in mind.
Powerful “Cube”
Achieves powerful and high AI processing power while keeping power consumption and heat emissions low.
Valuable “mini”
With a size that can fit into both hands, the AI inside mini can be used in a variety of locations and scenarios.
Specifications
Product name |
AI inside Cube Pro |
AI inside Cube |
AI inside Cube mini |
External dimensions |
450×178×715mm |
308×300×300mm |
156×150×154mm |
Weight |
Approx. 25 kg |
Approx. 11 kg |
Approx. 2.5 kg |
Storage |
2TB |
1TB |
512GB |
Processing performance* |
Please contact us for details. |
Approx. 25,000 items/hour |
Approx. 7,200 items/hour |
* Processing performance of the integrated model when not equipped with the AI-OCR feature for semi-structured form