Study on the Implementation of a Simple and Effective Memory System for an AI Chip

In this study, a simple and effective memory system required for the implementation of an AI chip is proposed. To implement an AI chip, the use of internal or external memory is an essential factor, because the reading and writing of data in memory occurs a lot. Those memory systems that are currently used are large in design size and complex to implement in order to handle a high speed and a wide bandwidth. Therefore, depending on the AI application, there are cases where the circuit size of the memory system is larger than that of the AI core. In this study, SDRAM, which has a lower performance than the currently used memory system but does not have a problem in operating AI, was used and all circuits were implemented digitally for simple and efficient implementation. In particular, a delay controller was designed to reduce the error due to data skew inside the memory bus to ensure stability in reading and writing data. First of all, it verified the memory system based on the You Only Look Once (YOLO) algorithm in FPGA to confirm that the memory system proposed in AI works efficiently. Based on the proven memory system, we implemented a chip using Samsung Electronics’ 65 nm process and tested it. As a result, we designed a simple and efficient memory system for AI chip implementation and verified it with hardware.

Research on Multifeature Data Routing Strategy in Deduplication

Deduplication is a popular data reduction technology in storage systems which has significant advantages, such as finding and eliminating duplicate data, reducing data storage capacity required, increasing resource utilization, and saving storage costs. The file features are a key factor that is used to calculate the similarity between files, but the similarity calculated by the single feature has some limitations especially for the similar files. The storage node feature reflects the load condition of the node, which is the key factor to be considered in the data routing. This paper introduces a multifeature data routing strategy (DRMF). The routing strategy is made based on the features of the cluster, including routing communication, file similarity calculation, and the determination of the target node. The mutual information exchange is achieved by routing communication, routing servers, and storage nodes. The storage node calculates the similarity between the files stored, and then the file is routed according to the information provided by the routing server. The routing server determines the target node of the route according to the similar results and the node load features. The system prototype is designed and implemented; also, we develop a system to process the feature of cluster and determine the specific parameters of various features of experiments. In the end, we simulate the multifeature data routing and single-feature data routing, respectively, and compare the deduplication rate and data slope between the two strategies. The experimental results show that the proposed data routing strategy using multiple features can improve the deduplication rate of the cluster and maintain a lower data skew rate compared with the single-feature-based routing strategy MCS; DRMF can improve the deduplication rate of the cluster and maintain a lower data skew rate.