Memory Characterization Of Big Data Applications
Garvit Tanwar, Insha K Nadaf, Anusha Manami, Vandita Gowda, Govind Sreekar Shenoy · 2022
This paper explores Memory Characterization of Big Data Applications for numerous big data applications. The team use Docker to run them. To get pertinent information, the team used our samples extensively with the cache simulator Dinero IV. In a cache memory, our main goal is to cut down on the amount of miss-rate while also speeding up fetching. So, using these tools and the rigorous standards that Cloud Suite supplied as big data apps, the team produced a large amount of data. Then, in order to plot the findings and observations, the team examined the data and looked for a relationship among the different components, such as cache size, block size, and associativity. Finding a pattern in the miss rates across various combinations of sizes and associativity in the benchmarks is our major goal. By doing so, the team will be able to identify a general pattern in how the demand miss rates change under the limits mentioned above. Demand miss rates fall when the team raise the associativity, cache size and line size, where d size is the cache size and I size is the line size.