network
Unleash the computational power of AI
The MangoBoost DPU can solve key bottlenecks in AI Servers and help GPUs to operate in an optimized environment
Challenge
Efficient neural network computation is crucial for AI applications. Despite advancements in GPU computational prowess, real-world AI tasks like network communication and data pre-processing limit their full potential. These overheads hinder optimal GPU utilization, leading to decreased system performance.
Solution
The MangoBoost DPU offloads AI tax operations for optimal computation. MangoBoost's DPU delivers scalable AI Serving Systems, exemplified by the NVIDIA Triton inference server.
storage
Streamline storage
The MangoBoost DPU offloads storage tax operations in disaggregated storage environments
Challenge
The disaggregated storage server effectively resolves the underutilization of storage devices in conventional architectures. However, it comes with a "storage tax" in the form of high CPU overhead for managing storage stacks and fast network transfers, exacerbated by the demand for large and fast storage demands.
Solution
The MangoBoost DPU accelerates heavy storage tax operations that reside in both NVMe-oF initiator & target servers and supports two representative network protocols (TCP and RoCEv2) in modern data centers. The SPECStorage benchmark shows that our DPU is a perfect fit for disaggregated storage solutions.
VIRTUALIZATION
Optimize virtualization
The MangoBoost DPU provides flexible and fast packet processing offload in virtualized environments
Challenge
VMs and container-based services are popular, and efficient packet processing is essential for handling their high network traffic. Datacenters manage traffic with network functions such as NAT to reach line-rate performance, but running these functions on CPUs incurs an unaffordably large overhead.
Solution
The MangoBoost DPU resolves this issue with flexible and fast packet processing offload. Users can freely configure hardware with P4 or offload OVS and VXLan, and run network functions at high throughput and low latency.
The MangoBoost DPU features simple deployment, maximum customizability, and support for a variety of workloads
Seamless DPU deployment with our SDK
The MangoBoost Software Development Kit (SDK) maximizes the potential of the MangoBoost DPU, allowing quick and easy offload and accelerating datacenter workloads. The SDK features user-friendly APIs, libraries, and industry-standard runtime environment.
Support and acceleration for key datacenter workloads
As modern datacenters transition towards virtualization and distributed models, designed to support workloads such as big data, Al, and cloud applications, MangoBoost's DPU provides comprehensive support for these complex requirements.
Customer-optimized DPUs
With our innovative composable DPU structure, we rapidly modify the design to customers' service requests. By arranging the necessary functions, we produce best-fit products with optimized performance and costs for various needs.