Have a use case in mind?
Let us know.
Get pricing, evaluation kits and engineering support to start building.
More ways to connect
Technology Partners
Build hardware, software or models that should run with Mobilint NPUs? Join our ecosystem and develop with us.
Become a Partner
Channel Partners
Bring Mobilint products to your market as an authorized reseller or distributor.
Join the Channel
Press & Media
Covering edge AI? Reach our team for interviews, press assets and company news.
Contact Press Team
Something else on your mind?
Contact the team
Have a use case in mind?
Let us know.
Get pricing, evaluation kits and engineering support to start building.
More ways to connect
Technology Partners
Build hardware, software or models that should run with Mobilint NPUs? Join our ecosystem and develop with us.
Become a Partner
Channel Partners
Bring Mobilint products to your market as an authorized reseller or distributor.
Join the Channel
Press & Media
Covering edge AI? Reach our team for interviews, press assets and company news.
Contact Press Team
Something else on your mind?
Contact the team
Have a use case in mind?
Let us know.
Get pricing, evaluation kits and engineering support to start building.
More ways to connect
Technology Partners
Build hardware, software or models that should run with Mobilint NPUs? Join our ecosystem and develop with us.
Become a Partner
Channel Partners
Bring Mobilint products to your market as an authorized reseller or distributor.
Join the Channel
Press & Media
Covering edge AI? Reach our team for interviews, press assets and company news.
Contact Press Team
Something else on your mind?
Contact the team
Developer Resources
Developer Resources
Developer Resources
Everything you need to build with Mobilint NPUs, in one place.
Everything you need to build with Mobilint NPUs, in one place.
Core Technology
To fully realize the potential of AI, we co-design and co-optimize every layer of the NPU stack in-house.
NPU Architecture
Compiler
Quantizer
Runtime
Runtime
Flexible environment for large-scale deployment
Quantizer
Compatibility verified across LLM, Vision, Action and other model architectures
Compiler
Co-designed with the hardware to maximize model-to-hardware mapping efficiency
NPU Architecture
Linear scaling from a single core to multi-chip deployments
Customer Stories
Meet the teams building AI products on Mobilint NPUs.
XSLAB
Maximum AI Performance on a Limited Power Budget | XSLAB X Mobilint
How XSLAB built a 50W TDP Arm AI server around Mobilint’s MLA100 PCIe Card, straight from CEO Myung-hwan Yoo.
Intellivix
4,000 Highway Camera Channels Nationwide — Now AI Sees It First | Intellivix X Mobilint
Why Intellivix chose Mobilint's NPU for the Korea Expressway Corporation's AI CCTV conversion project, straight from AI Researcher Kang Changhee.
View all customer stories
Core Technology
To fully realize the potential of AI, we co-design and co-optimize every layer of the NPU stack in-house.
NPU Architecture
Compiler
Quantizer
Runtime
Runtime
Flexible environment for large-scale deployment
Quantizer
Compatibility verified across LLM, Vision, Action and other model architectures
Compiler
Co-designed with the hardware to maximize model-to-hardware mapping efficiency
NPU Architecture
Linear scaling from a single core to multi-chip deployments
Customer Stories
Meet the teams building AI products on Mobilint NPUs.
XSLAB
Maximum AI Performance on a Limited Power Budget | XSLAB X Mobilint
How XSLAB built a 50W TDP Arm AI server around Mobilint’s MLA100 PCIe Card, straight from CEO Myung-hwan Yoo.
Intellivix
4,000 Highway Camera Channels Nationwide — Now AI Sees It First | Intellivix X Mobilint
Why Intellivix chose Mobilint's NPU for the Korea Expressway Corporation's AI CCTV conversion project, straight from AI Researcher Kang Changhee.
View all customer stories
Core Technology
To fully realize the potential of AI, we co-design and co-optimize every layer of the NPU stack in-house.
NPU Architecture
Compiler
Quantizer
Runtime
Runtime
Flexible environment for large-scale deployment
Quantizer
Compatibility verified across LLM, Vision, Action and other model architectures
Compiler
Co-designed with the hardware to maximize model-to-hardware mapping efficiency
NPU Architecture
Linear scaling from a single core to multi-chip deployments
Customer Stories
Meet the teams building AI products on Mobilint NPUs.
XSLAB
Maximum AI Performance on a Limited Power Budget | XSLAB X Mobilint
How XSLAB built a 50W TDP Arm AI server around Mobilint’s MLA100 PCIe Card, straight from CEO Myung-hwan Yoo.
Intellivix
4,000 Highway Camera Channels Nationwide — Now AI Sees It First | Intellivix X Mobilint
Why Intellivix chose Mobilint's NPU for the Korea Expressway Corporation's AI CCTV conversion project, straight from AI Researcher Kang Changhee.
View all customer stories
Based on internal benchmark results (Sep. 2026)
・CNN : YOLO26n, s, m and l (B=1, NPU core mode single)
・ARIES measured on MLA100 PCIe Card (80 TOPS INT8, 25 W TDP). GPU A, B and C are anonymized GPUs
・ARIES price based on large-quantity order; GPU prices based on Amazon listings
Based on internal benchmark results (Sep. 2026)
・CNN : YOLO26n, s, m and l (B=1, NPU core mode single)
・ARIES measured on MLA100 PCIe Card (80 TOPS INT8, 25 W TDP). GPU A, B and C are anonymized GPUs
・ARIES price based on large-quantity order; GPU prices based on Amazon listings
Performance
FPS
Up to 151% higher
991
874
412
459
YOLO26n
609
546
242
264
YOLO26s
267
306
139
122
YOLO26m
223
237
112
99
YOLO26l
Power Efficiency
FPS/W
Up to 1722% higher
45.73
5.58
2.51
8.40
YOLO26n
26.08
2.69
1.45
3.81
YOLO26s
11.03
1.23
0.81
1.77
YOLO26m
9.03
0.94
0.62
1.43
YOLO26l
Cost Efficiency
FPS/$
Up to 738% higher
1.32
0.51
0.16
0.38
YOLO26n
0.81
0.32
0.10
0.22
YOLO26s
0.36
0.18
0.06
0.10
YOLO26m
0.30
0.14
0.04
0.08
YOLO26l
Mobilint ARIES
GPU A
GPU B
GPU C
Beyond the limits of power, heat, and cost AI processors built for the field
Beyond the limits of power, heat, and cost AI processors built for the field
Beyond the limits of power, heat, and cost AI processors built for the field
Chip Lineup
From powerful on-device intelligence to scalable on-premises AI inference
From powerful on-device intelligence to scalable on-premises AI inference
From powerful on-device intelligence to scalable on-premises AI inference
Built on a powerful, low-power NPU stack, available in multiple form factors for flexible AI deployment.
Built on a powerful, low-power NPU stack, available in multiple form factors for flexible AI deployment.
AI Chips Built for Real-World Tasks
AI Chips Built for Real-World Tasks
AI Chips Built for Real-World Tasks
Mobilint’s high-performance, low-power NPU architecture and full-stack software give AI a platform to run where the work happens.
Built for the real world
Built for the real world
Built for the real world
From high-intensity AI inference to on-device AI where space and power are limited, Mobilint hardware is built to match each deployment.
From high-intensity AI inference to on-device AI where space and power are limited, Mobilint hardware is built to match each deployment.
Power & Cost Efficiency
More AI performance per watt and per dollar, within the power and cost limits of real deployments.
Model & Software Versatility
From CNNs to LLMs, VLMs and VLAs, one SDK supports a wide range of models and the major AI frameworks.
Platform Compatibility
Works with the operating systems, CPUs and servers you already run, so it fits into existing systems.
Field-Proven Reliability
Built for demanding real-world environments, and validated through environmental testing and international certifications.
Read our vision and mission
