Cosmic: Cost-Effective Support for Cloud-Assisted 3D Printing
summary
Cloud controller called Cosmic partitions serverless workers into groups and issues timely speculation to hide ~10 ms invocation delays measured on AWS Lambda, running a physics-based PVF scan optimizer that must reply in tens of milliseconds. Evaluation uses a local printer emulator plus a VM coordinator and 30 diverse print jobs; Cosmic meets all timing windows and cuts cost versus VM and naïve serverless baselines by about 2.8x–3.5x, while keeping computation on AWS Lambda.
AI-written summary · about
chapters (10)
- 0:00Cosmic project intro
- 1:45Cloud control for 3D printing
- 4:39Metal 3D printing time limits
- 6:03Serverless vs VMs
- 7:47Serverless real-time challenges
- 9:37Cosmic’s function grouping
- 11:24Cosmic’s speculation mechanism
- 12:52Cosmic performance modeling
- 13:57Evaluation results and savings
- 15:02Conclusion and future
related talks
17mCloud Computing Made Better: An AWS View on Serverless Computing
46mServerless for Data Scientists
36mQuantum & High-Performance Computing in The Cloud
28mServerless Computing for Mathematical Optimization
54mServerless beyond Functions
54mContainerless in the cloud with AWS Lambda
42mServerless Systems That Won't Burn a Hole in Your Pocket
35mEfficiency of Serverless Architectures with AWS Solutions