Add Dragonfly to I/O bottlenecks in AI Scheduling Challenges whitepaper - #2272
Open
pmady wants to merge 1 commit into
Open
Add Dragonfly to I/O bottlenecks in AI Scheduling Challenges whitepaper#2272pmady wants to merge 1 commit into
pmady wants to merge 1 commit into
Conversation
…a tools Chapter 4 had an empty placeholder bullet under "Model loading for inference", following the existing "Cold start" entry. Fill it with the distribution-at-scale case: many replicas starting together each pull the same weights independently, saturating registry egress. Cold-start streaming tools address per-pod latency but not this aggregate load. Chapter 5 adds Dragonfly to Data and Storage Tools and to the I/O Bottlenecks row of the challenge-to-solution mapping table. Both are ordered alphabetically, consistent with the rest of the section. Dragonfly is a CNCF Graduated project for peer-to-peer distribution of container images and AI model artifacts. Signed-off-by: Pavan Madduri <pavan4devops@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Follow-up to #2164. Marlow asked me to send this straight to cncf/toc.
Ch 4, I/O Bottlenecks: the "Model loading for inference" list has an empty bullet right after "Cold start" looks like a placeholder that shipped blank. Filled it with the scale-out case: when a lot of replicas start at once they each pull the same weights independently and hammer the registry. The streaming tools in the cold start bullet help per-pod latency but don't do anything for that.
Ch 5: added Dragonfly (CNCF graduated) to Data and Storage Tools and to the I/O Bottlenecks row of the mapping table. Alphabetical in both, to match what's already there.
Happy to reword or make any changes.