# \[2018 NIPS] Dynamic Space-Time Scheduling for GPU Inference

## Summary

The authors evaluated different multiplexing (time & space) techniques for ML inferences on GPUs and proposed ideas to achieve the best tradeoff across criterias.

## Background & Motivation

Almost all cloud inference service providers/frameworks assign each model an exclusive GPU. This, combined with the small batch sizes used in an online setting, results in low hardware utilization. Current approaches that multiplex workloads have different tradeoffs, and there is no single solution that wins on all criteria.

<table><thead><tr><th>Approach</th><th width="150">Utilization</th><th width="179.18672199170126">Performance (throughput/latency)</th><th width="223">Predictability/Performance Isolation</th></tr></thead><tbody><tr><td>Exclusive access</td><td>Poor</td><td>Good</td><td>Good</td></tr><tr><td>Time multiplexing (CUDA context switching)</td><td>Average</td><td>Poor</td><td>Good</td></tr><tr><td>Spatial multiplexing</td><td>Good</td><td>Average</td><td>Poor</td></tr></tbody></table>

## Design & Implementation

![](/files/ppSB52QnYqqRbFCUogwv)

The authors proposed software-level fusion of kernel operators across multiple inference jobs to get the best of all worlds.

## Links & References

* [Paper PDF](http://learningsys.org/nips18/assets/papers/102CameraReadySubmissionGPU_Virtualization%20\(8\).pdf)


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://blog.ruipan.xyz/machine-learning-systems/machine-learning-systems-index/2018-nips-dynamic-space-time-scheduling-for-gpu-inference.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.