<i>PISeL</i> : &lt;u&gt;Pi&lt;/u&gt;pelining DNN &lt;u&gt;I&lt;/u&gt;nference for &lt;u&gt;Se&lt;/u&gt;rver&lt;u&gt;l&lt;/u&gt;ess Computing

Serverless computing offers resource efficiency, cost efficiency, and a "pay-as-you-go" pricing model, which makes it highly attractive to both users and cloud providers. However, serverless computing faces serious cold start problem, especially for deep neural network (DNN) inference, which requires low latency. Existing cold start optimization focuses only on quick container start and fast runtime and library loading. However, DNN application bootstrap (DNN framework load and start, model initialization, model download, deserialization and copy) is the leading factor during the overall cold start time. As the model size grows, the application-level bootstrap becomes more severe.