AWS Lambda Language Performance

Abstract

Cloud services are experiencing expansive growth, and the potential uses cases for its application in the IT sector is becoming increasingly widespread. This rapid growth is accompanied by a demand for performance which necessitates systematic benchmarking. The process of setting up a cloudbased benchmark is tedious and obstructive. Therefore, frequent benchmarks must be conducted in order to inform individuals from practitioners to hobbyists alike. Amazon Web Services is a major market leader in cloud computing which offers a Function as a Service platform named AWS Lambda. The following will conduct a benchmark on the Lambda platform and its supported languages: C#, Java, Node.js, and Python. The benchmark will examine the performance of the respective languages with relation to workload input size when configured to three different memory sizes: 128, 512, and 1024MB. The results presented reveal languages such as Java and C# consistently outperform the other languages with C# being the most performant when configured to 128 and 1024MB. All languages experienced a performance increase in tandem with increasing memory size.