DEEPBOX

DEEPBOX

RUN AI MODELS

AT THE EDGE

From ultra-compact to desktop-sized, all DeepBoxes come rugged and optimized for heavy industrial usage

From ultra-compact to desktop-sized, all DeepBoxes come rugged and optimized for heavy industrial usage

DEEPBOX

RUN AI MODELS

AT THE EDGE

From ultra-compact to desktop-sized, all DeepBoxes come rugged and optimized for heavy industrial usage

Choose the right

DEEPBOX

for you

Hardened units for industrial environments

AI INFERENCE (EGDE)

Industrial and rugged design supports wide operating temperatures, power input, and vibration tolerance

NVIDIA Jetson Module Compatible: Jetson Orin™ NX 8GB/16GB and Nano 4GB/8GB

AIM-Linux software service with peripheral driver integration supports NVIDIA JetPack SDK

Accepts flexible application-oriented UIO40-Express I/O expansion cards

Versatile wireless/storage expansions with M.2 B key, M.2 E key, and M.2 key M socket

WISE-DeviceOn supported for quick AI deployment at scale

AI INFERENCE (EGDE)

Industrial and rugged design supports wide operating temperatures, power input, and vibration tolerance

NVIDIA Jetson Module Compatible: Jetson Orin™ NX 8GB/16GB and Nano 4GB/8GB

AIM-Linux software service with peripheral driver integration supports NVIDIA JetPack SDK

Accepts flexible application-oriented UIO40-Express I/O expansion cards

Versatile wireless/storage expansions with M.2 B key, M.2 E key, and M.2 key M socket

WISE-DeviceOn supported for quick AI deployment at scale

AI INFERENCE

Certified for heavy industry (IEC 61000-6-4)

NVIDIA® Jetson AGX Orin 64GB and 32GB Modules

9~36V wide power and -10~60 °C wide temp. supported

Multiple IO ports: LAN, DIO, COM, CANBus, USB 3.2 and USB type C

Versatile expansions with M.2 B key, M.2 E key, SD, PCIe x16(optional) support

Support Linux Ubuntu 20.04 LTS and JetPack SDK 5.0 and above

AI INFERENCE

Certified for heavy industry (IEC 61000-6-4)

NVIDIA® Jetson AGX Orin 64GB and 32GB Modules

9~36V wide power and -10~60 °C wide temp. supported

Multiple IO ports: LAN, DIO, COM, CANBus, USB 3.2 and USB type C

Versatile expansions with M.2 B key, M.2 E key, SD, PCIe x16(optional) support

Support Linux Ubuntu 20.04 LTS and JetPack SDK 5.0 and above

AI TRAINING

Support Intel® 12th&13th Gen Core™ i3/i5/i7/i9 processor

Triple independent display: HDMI + HDMI + Optional Display

DDR5 SO-DIMM ECC/non-ECC memory support up to 64 GB

9-36VDC wide range power input

Up to 3sets 2.5" hard drive bays, support Intel® SW RAID

Support up to 4GbE, 8USB, 8COM, 16bit DIO, 2CANBus, TPM2.0

SUSI API, WISE-DeviceOn, McAfee and Acronis bundled

AI TRAINING

Support Intel® 12th&13th Gen Core™ i3/i5/i7/i9 processor

Triple independent display: HDMI + HDMI + Optional Display

DDR5 SO-DIMM ECC/non-ECC memory support up to 64 GB

9-36VDC wide range power input

Up to 3sets 2.5" hard drive bays, support Intel® SW RAID

Support up to 4GbE, 8USB, 8COM, 16bit DIO, 2CANBus, TPM2.0

SUSI API, WISE-DeviceOn, McAfee and Acronis bundled

Where should I start?

Where should I start?

Where should I start?

Find which integration best fits your needs

Find which integration best fits your needs

Find which integration best fits your needs

FAQ

What is a DeepBox™?

A DeepBox™ is a packaged AI hardware and software solution made by ThinkDeep AI. All DeepBoxes are built with industrial and rugged design to support wide operating temperatures, power input, and vibration tolerance. For constrained environments DeepBoxes are available in compact and ultra-compact formats.

What are the applications for DeepBox™?

DeepBoxes are multi-purpose and can be used for applications such as defect detection in quality control and threat recognition in security. They are optimized for real-time processing of visual data with AI and adapt to a wide variety of use cases. The architectures make them ideal for either open or closed-loop deployments (where internet access is not permitted). Another benefit is cost control, where running AI algorithms on device is used to avoid going over budget in terms of compute and energy costs.

© 2023 ThinkDeep

© 2023 ThinkDeep

© 2023 ThinkDeep