Add initial KV Cache benchmark implementation for MLPerf Storage v3 #214
Conversation
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
This commit introduces a comprehensive KV Cache benchmark suite designed to measure storage system performance under AI/ML inference workloads, specifically targeting Large Language Model (LLM) key-value cache operations. Key components added: - Core benchmark scripts (kv-cache.py, kv-cache_sharegpt_replay.py) - Benchmark wrapper and validation tools (kv-cache-wrapper.sh, validate.sh) - Comprehensive proposal documentation for MLPerf Storage v3 integration - README with benchmark overview and usage guidelines The benchmark simulates realistic LLM inference patterns including: - Key-value cache read/write operations - Mixed sequential and random access patterns - Multi-threaded concurrent access scenarios - Conversation-based workload replay using ShareGPT dataset This work addresses the growing need to standardize storage performance measurements for AI inference workloads and provides a foundation for MLPerf Storage v3.0 KV cache benchmark specification.
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
Contributor
FileSystemGuy
left a comment
FileSystemGuy
left a comment
There was a problem hiding this comment.
Initial version being imported.
FileSystemGuy
approved these changes
Nov 25, 2025
wvaske
approved these changes
Nov 25, 2025
Contributor
wvaske
left a comment
wvaske
left a comment
There was a problem hiding this comment.
We will turn comments into issues once this is merged.
| total_prob = sum(chunk_probabilities) | ||
| chunk_probabilities = [p / total_prob for p in chunk_probabilities] | ||
|
|
||
| retrieved_indices = np.random.choice( |
Contributor
There was a problem hiding this comment.
TODO: Add support for different random distributions (random, uniform, zipfian)
|
|
||
| # --- Tiering Logic --- | ||
| # Decide which tier to write to based on available memory. | ||
| with self.memory_lock: |
Contributor
There was a problem hiding this comment.
New KVs should be written to the top layer and trigger eviction from a tier if sufficient space doesn't exist.
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
3 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 commit introduces a comprehensive KV Cache benchmark suite designed to measure storage system performance under AI/ML inference workloads, specifically targeting Large Language Model (LLM) key-value cache operations.
Key components added:
The benchmark simulates realistic LLM inference patterns including:
This work addresses the growing need to standardize storage performance measurements for AI inference workloads and provides a foundation for MLPerf Storage v3.0 KV cache benchmark specification.