Cascade 100M
M.2 AI Accelerator Module
Scalable AI for systems with a standard M-key M.2 interface.
At a glance
Chips
4× MX3 AI Accelerator Chips
AI performance
Weight storage
Power
Form factor
M.2 2280 M-Key
Best for
Embedded and edge systems
The Cascade 100M adds scalable AI acceleration to systems with a standard M-key M.2 interface.
Powered by four MX3 accelerators, the compact M.2 module offloads AI inference from the host processor while fitting into a wide range of embedded and compute platforms. It provides a flexible way to add dedicated AI performance without redesigning an existing system around a specialized accelerator.
PRODUCT PROOF
Purpose-built AI acceleration.
The key hardware characteristics of MX3 Edge AI Accelerator at a glance.
25
TFLOPS
AI performance
4x
MX3
AI accelerator chips
42 MB
On-chip
model weight storage
What you get
Dedicated AI inference that works alongside your host.
Designed for flexible integration
Bring dedicated MemryX AI acceleration into your system with an efficient design that balances performance, power, and thermal requirements.
Dedicated AI offload
Offload neural-network inference to MX3 while preserving host compute for the rest of your application. Unified software experience
Unified software experience
Use the same MemryX software workflow—from model compilation through runtime deployment—across single- and multi-chip configurations.
Real-time batch-1 inference
MX3’s pipelined dataflow architecture delivers low-latency inference without relying on large batch sizes.
Inside the architecture
Why MX3 works differently
At-memory compute and a pipelined dataflow architecture deliver efficient, consistent AI inference without requiring developers to redesign their applications around the accelerator.
Accuracy
No retraining required
BFloat16 activations preserve model accuracy through direct recompilation, without requiring retraining or hand-tuning.
Deterministic
Consistent batch-1 latency
A pipelined dataflow architecture keeps multiple inputs in flight at once, maintaining high hardware utilization without batching for asynchronous, low-latency inference.
Integration
Designed to work with your host
MemryX operates as dedicated AI acceleration, offloading neural-network inference from the host processor.
Architecture
At-memory dataflow
The architecture keeps model data close to compute, enabling high-throughput AI processing by minimizing costly data movement.
Flexibility
Build beyond a single model
Run models individually or combine them through concurrent, pipelined, or conditional execution so developers can build around the application — not the accelerator.
Deployment
Built to scale
Scale from compact deployments to multi-chip configurations with a common software foundation as performance requirements grow.
Specifications
Cascade 100M at full detail.
Product
Part Number
Accelerators
AI performance
Weight storage
Form factor
M.2 2280 M-Key
Dimensions
Input Voltage
Power
1 W – 12 W
interface
PCI Express Gen 3 and above
Frameworks
Certification
Operating Temperature
Operating Humidity
Target Systems
Applications
Built for real-world AI
Video Management Systems
Industrial & Robotics
Transportation
Smart Devices
Start building
Everything you need to get started.
Documentation
Get Started
Install the SDK, run your first inference, and compile your own model.
CODE
eXamples
Working reference applications you can clone and run on MemryX hardware.
Support
Community forum
Ask questions, share builds and get help from MemryX engineers.
Add dedicated AI to an M.2-equipped system.
Talk to MemryX about Cascade 100M availablility, integration, and volume requirements.
