media
press releases
MangoBoost at SNIA Developer Conference 2024
MangoBoost Releases GPU-Storage-Boost Solution Using MI300X GPU Servers at the SNIA Developer Conference 2024
September 23, 2024
MangoBoost, a provider of advanced system solutions aimed at maximizing AI data center efficiency, and a world-leading GPU company, jointly presented on “Accelerating GPU Server Access to Network-Attached Disaggregated Storage using Data Processing Unit (DPU)” at the SNIA Development Conference (SDC) 2024. Hosted by the Storage Networking Industry Association, the SNIA Development Conference (SDC) is a premier educational event in the storage industry, attracting industry leaders, visionaries, and innovators shaping the future of data technologies.
The exponential growth of artificial intelligence (AI) has driven a fundamental shift in data center storage architectures. As GPU servers demanded access to increasingly vast amounts of data, the need for more scalable and cost-effective solutions became critical. Traditional CPU-centric servers struggled to keep pace, facing significant performance and scalability challenges. These included the substantial CPU resource consumption required to access remote storage over networks and the suboptimal data pathways between GPU and storage nodes, which failed to fully utilize high-speed network and interconnect bandwidth. To address these challenges, the MangoBoost Data Processing Units (DPUs) accelerated key infrastructure functions, including networking, storage, and peer-to-peer communications with GPUs, thereby enhancing overall system efficiency.
In their joint presentation, MangoBoost and the GPU company covered the following key topics:
The presentation was delivered by Eriko Nurvitadhi, Chief Product Officer of MangoBoost, and a Storage and Networking Architect at the global GPU company.
Before co-founding MangoBoost, Dr. Nurvitadhi served as a Principal Engineer at Intel, focusing on hardware accelerators (e.g., FPGAs, GPUs, ASICs) for server systems and AI technologies, driving R&D into product impacts. He has more than 70 peer-reviewed publications and over 120 patents granted or pending in these areas. He received a PhD in ECE from Carnegie Mellon University and an MBA from Oregon State University.
Their combined expertise and insights provided valuable perspectives on the evolving landscape of data center architecture and the role of advanced technologies in meeting the challenges posed by AI.
For more information on the co-presentation, please visit the session page.
About MangoBoost, Inc
MangoBoost is a provider of advanced system solutions maximizing AI data center efficiency. Central to our wide array of solutions, the MangoBoost DPU enables cost-efficient, standards-based AI infrastructure by providing full compatibility with all commodity GPUs, accelerators and storage products, effective utilization of legacy infrastructure and an easy-to-use network system.
Founded in 2022 with 10 years of prior research published at top-tier conferences such as OSDI and ISCA, MangoBoost consists of more than 80 system architects, software engineers and FPGA/SoC engineers, many of whom hold a PhD from the world’s top research institutions and extensive industry experience at big techs.
MangoBoost has secured more than 60 million US dollars in funding and is rapidly expanding its presence in the United States, Canada and South Korea.
For media inquiries or further information, please contact Junki Park (Chief Financial Officer) at:
contact@mangoboost.io.