Conversation
…g BTBHistory with FTQ
…n with branch metrics reported by PMU events in A64FX
ABenC377
added
enhancement
ABenC377
requested review from
FinnWilkinson,
JosephMoore25,
dANW34V3R and
jj16791
and removed request for
FinnWilkinson
dANW34V3R
approved these changes
Aug 23, 2024
FinnWilkinson
approved these changes
Aug 23, 2024
JosephMoore25
approved these changes
Aug 28, 2024
jj16791
approved these changes
Aug 29, 2024
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
5 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This update to the branch predictors (both Perceptron and Generic) replaces the hashmap that was previously being used to keep previous-state information between predict and update with a queue (called a Fetch Target Queue or FTQ) of previous-state information for the in-flight branches. The FTQ allows speculative updating of the global history (which is also included in this PR).
In terms of performance, it seems to be quite lateral. a full breakdown of stats across our benchmark suite is provided below. To summarise the results, most benchmarks have an improved misprediction rate with the update, but the changes in execution time are hit and miss with many becoming a bit slower (probably due to an increased number of interactions with the branch predictor through the pipeline).
I think this is still a worthwhile change to make to SimEng even though there isn't a clear-cut performance upgrade because it makes the flow of branches through the branch predictor more regular and so should facilitate more complicated branch predictors in the future (e.g., TAGE).