Introducing

MLSoC Modalix

Breakthrough Multi-Modal Edge AI Product Family

ONE Platform to support CNNs, Transformers, LLMs, LMMs and GenAI at the Edge.
More than 10X the Performance per Watt of alternatives.
Family of products for Edge AI your way: 25 TOPs to 200 TOPs.

Get Started

Modalix Family of Products to Suit Your Edge AI Needs

25 TOPS
MLSoC Modalix M25 Icon
  • Drones
  • Robotics
  • Healthcare Intake Triage
  • Smart Cameras (Depth/Texture)
50 & 100 TOPS
MLSoC Modalix M50 Icon MLSoC Modalix M100 Icon
  • Robotics - AGV, AMR
  • Smart Retail
  • Smart Cities
  • Industrial
  • Aerospace & Defense
  • Healthcare Diagnostic Equipment
  • Smart Cameras (People ID, Scene Analytics)
200 TOPS
MLSoC Modalix M200 Icon
  • Edge Inference Servers
  • Enterprise Vision Systems
  • AI Gateway Devices
MLSoC Modalix M25 Icon
25 TOPS
MLSoC Modalix M50 Icon
50 TOPS
MLSoC Modalix M100 Icon
100 TOPS
MLSoC Modalix M200 Icon
200 TOPS
  • Drones
  • Robotics
  • Healthcare Intake Triage
  • Smart Cameras (Depth/Texture)
  • Robotics - AGV, AMR
  • Smart Retail
  • Smart Cities
  • Industrial
  • Aerospace & Defense
  • Healthcare Diagnostic Equipment
  • Smart Cameras (People ID, Scene Analytics)
  • Edge Inference Servers
  • Enterprise Vision Systems
  • AI Gateway Devices

Breakthrough Multi-Modal Edge AI Product Family

For the first time in a system-on-chip at the Edge, the MLSoC Modalix family enables integration of multiple modalities. Text, image audio, video and haptic inputs together, performing complex ML-inference, and presenting outputs in any modality. It handles everything from CNNs, Transformers, LLMs, LMMs and GenAI, scaling with your applications from 25 to 200 TOPS.

SiMa.ai MLSoC is a Complete System-on-Chip

Unique combination of high-performance and low power for building
complete system solutions for edge ML applications.

Download Product Brief

Run Full Applications on a Single Chip

SiMa.ai Accelerates the Entire Pipeline vs. the Competition

Other ML accelerators only support this part of the pipeline

Learn About the Evolution of Our MLSoC Generations

Modalix Extends Established Industry Leadership of the First Generation MLSoC

MLSoC Modalix complements the first generation MLSoC and extends its capabilities with a lineup of products at different combinations of performance, interfaces and power so that you can always pick the right product for your application. Choose between the first generation MLSoC or from members of the MLSoC Modalix family to precisely target your application with the features and performance you need.
GEN 1
GEN 2
MLSoC
MLSoC Modalix
M25 / M50 / M100 / M200
TOPS
50
25/50/100/200
Machine Learning Accelerator (MLA)
MLA optimized for CNN-based Models. INT8 and INT16 support.
MLA enhanced with hardware blocks to accelerate GenAI computations. Supports mixed precision: Floating point (BFLOAT16) and INT8 and INT16
Application Processor
Arm A65 x 4 dual-threaded 1.15GHz; 15K DMIPS
Arm A65 x 8 dual threaded 1.5GHz; 32K DMIPS
MIPI support
N/A
4 independent CSI 2 interfaces each with 4x 2.5Gbps lanes
ISP
N/A
Arm C-71 1.2GHz, supporting 8-24 bit RAW including RGBIR
Vision processor
Synopsys ARC EV-74 800 MHz
Synopsys ARC EV-74 1GHz
Video decoder
H.264/265 4KP60 x 1
H.264/265/AV1 4KP60 x 1
Video encoder
H.264 4KP30 x 1
H.264 4KP30 x 1
PCie
Gen 4 x 8 lanes endpoint only
Gen 5 x 8 lanes Root Complex and Endpoint
Ethernet
1GbE x 4
10GbE x 4
On-chip Memory
4MB
8MB
DRAM
32-bit LPDDR4 x 4
128 bit (8 channels) 32 & 64 bit LPDDR4/4X/5
Manufacturing Process node
TSMC 16nm
TSMC 6nm
Model Types
CNN-based models
CNNs, Transformers, LLMs, LMMs & GenAI
TOPS
Machine Learning Accelerator (MLA)
Application Processor
MIPI support
ISP
Vision processor
Video decoder
Video encoder
PCie
Ethernet
On-chip Memory
DRAM
Manufacturing Process node
Model Types
Gen 1 Chip
MLSoC
50
MLA optimized for CNN-based Models. INT8 and INT16 support.
Arm A65 x 4 dual-threaded 1.15GHz; 15K DMIPS
N/A
N/A
Synopsys ARC EV-74 800 MHz
H.264/265 4KP60 x 1
H.264 4KP30 x 1
Gen 4 x 8 lanes endpoint only
1GbE x 4
4 MB
32-bit LPDDR4 x 4
TSMC 16nm
CNN-based models
Gen 2 Chip Gen 2 Chip Gen 2 Chip Gen 2 Chip
MLSoC Modalix
M25
M50
M100
M200
25
50
100
200
MLA enhanced with hardware blocks to accelerate GenAI computations. Supports mixed precision: Floating point (BFLOAT16) and INT8 and INT16
Arm A65 x 8 dual threaded 1.5GHz; 32K DMIPS
4 independent CSI 2 interfaces each with 4x 2.5Gbps lanes
Arm C-71 1.2GHz, supporting 8-24 bit RAW including RGBIR
Synopsys ARC EV-74 1GHz
H.264/265/AV1 4KP60 x 1
H.264 4KP30 x 1
Gen 5 x 8 lanes Root Complex and Endpoint
10GbE x 4
8 MB
128 bit (8 channels) 32 & 64 bit LPDDR4/4X/5
TSMC 6nm
CNNs, Transformers, LLMs, LMMs & GenAI

ONE Platform to Accelerate Your Entire Edge-AI Pipeline

Looking for ONE Platform, in one piece of silicon, with the industry’s fastest, most power-efficient and truly cost-effective edge-ML accelerator?

Why Choose SiMa.ai MLSoC for Your Edge AI Needs?

Family of offerings supports applications from CNNs through Transformers, LLMs, LMMs and GenAI
Palette software purpose-built for MLSoC provide ONE Platform for simplified development, deployment and reduced TCO
Industry leading FPS/Watt across platform offerings executes complex pipelines quickly and efficiently
Built-in powerful eight-core Arm A65, ISP, CV Processor and MLA accelerate entire ML pipelines and business logic on a single chip

The MLSoC Modalix Family and First Generation MLSoC are Compatible with SiMa.ai Palette Software

Palette is designed from the ground-up to take advantage of hardware innovations in SiMa.ai purpose-built silicon for the Edge, and accelerate complete application pipelines including AI and business logic.

Learn more about how to create, build and deploy edge ML solutions on SiMa.ai MLSoC with Palette software here.