Dedicated Computing Announces M1000, Offering High-Performance AI Acceleration Based on NVIDIA Clara Holoscan MGX Platform

Integrating hardware and software, the M1000 platform is optimized for medical device developers augmenting traditional healthcare tools with AI

 MILWAUKEE – March 22, 2022 – Dedicated Computing, a leading original design manufacturer (ODM) of proprietary, embedded computing systems, today announced M1000, its high-performance artificial intelligence acceleration platform. The M1000 is an optimized building block for medical device developers focused on data-driven healthcare applications. System longevity is a key value in the high-performance platform, with a purchase lifecycle of five years and availability to medical device manufacturers. Incorporating a software stack developed and supported by NVIDIA, the M1000 platform contains the supporting software components necessary, a long-term support software stack with board support package, operating system, and libraries. Runtime layers are included as the primary interface to the end-customer applications.

Artificial intelligence is a new tool in the treatment provider’s toolbox, supporting their experience with more advanced imaging that enhances decisions and outcomes. The M1000 platform is initially targeted at real-time inference augmentation for endoscopy applications. Dedicated’s AI acceleration platform is derived from the NVIDIA Clara Holoscan MGX platform and is based on NVIDIA Orin to deliver 250 AI TOPS and scaling to over 600 AI TOPS with an NVIDIA RTX A6000 GPU. NVIDIA ConnectX7 delivers streaming I/O with up to 200 GbE and GPUDirect RDMA for ultra low-latency processing. Supported by this AI computer and software framework, medical device developers are empowered to envision and create advanced autonomous systems and AI-powered medical devices that feature superior image, video, and signal processing. Applications are global and diverse, ranging from endoscopes to advanced surgical displays and more.

“As longtime developers of robust, high-performance hardware solutions, our building block approach recognizes the diverse needs of medical device developers. Systems are designed for the footprint limitations and environmental demands of healthcare anywhere, even as they meet the performance that breaks new ground in AI implementation,” said Rich Ross, VP of Technology and Product Development, Dedicated Computing. “Patient care improves when skilled clinicians are armed with continually smarter and more capable tools. The M1000 platform is another step forward in supporting the needs of medical device developers, blending high-performance capabilities with state-of-the-art graphics performance powered by teaming with NVIDIA and its continually advancing technologies.”

Dedicated Computing’s building block hardware designs are diverse and proven, keeping the focus on robust data performance and physical durability for the environmental rigors of healthcare computing. Decades of expertise includes applications featuring AI, machine vision, and advanced GPU processing in patient care scenarios such as data acquisition for intravascular ultrasound, advanced eye measurements that eliminate subjective processes, lab and cyclotron automation, fluorescence image-guided, and full color 3D image reconstruction based on light instead of radiation.

About Dedicated Computing

As an Original Design Manufacturer (ODM) of proprietary, embedded computing systems, Dedicated Computing is a partner to Original Equipment Manufacturers (OEMs) worldwide. Focused on healthcare, life sciences, training and simulation, and industrial markets, Dedicated Computing employs over 200 associates collectively supporting its mission of Powering the World’s Most Important Devices®.

To connect with the Dedicated Computing team for support in design, development, or deployment, call 877.333.4848 or connect via email at support@dedicatedcomputing.com.

#  #  #

Media Contact
Jeff Durst, Sr Product Manager
262-953-1532
jeff.durst@dedicatedcomputing.com

Comments are closed