Parallel Attention Computation With Blockwise GPU Pipelining

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Attention mechanisms in machine learning models, particularly for large language models (LLMs), are resource-intensive due to sequential execution of matrix multiplication and element-wise operations, leading to inefficient use of GPUs and increased time and memory requirements as models process longer input sequences.

Innovation Solution

An accelerated attention mechanism enables parallel execution of matrix multiplication and element-wise operations through pipelining, allowing certain matrix multiplication operations to be performed concurrently with element-wise operations, thereby overlapping work inside and outside GPU dedicated units for matrix multiplication.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If matrix multiplication and element-wise operations are executed sequentially in attention mechanisms, then computational accuracy is maintained, but processing speed decreases and GPU efficiency is reduced

Engineering Contradiction:
Improveprocessing speedVSAvoidcomputational complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent segments the attention mechanism computation into distinct blocks: query blocks, key blocks, and value blocks. This segmentation allows independent parallel processing of different block combinations, enabling matrix multiplication and element-wise operations to execute concurrently without interfering with each other, thus improving processing speed while maintaining computational accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary segmentation of input sequences into blocks before computation begins. By pre-organizing the data into query blocks, key blocks, and value blocks with predetermined dimensions, the system prepares the computational structure in advance, allowing subsequent matrix multiplication and element-wise operations to proceed in parallel without sequential dependencies.

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If attention mechanisms process longer input sequences, then model capability improves, but time and memory requirements increase

Engineering Contradiction:
Improvemodel capabilityVSAvoidprocessing time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent divides long input sequences into smaller blocks (query blocks, key blocks, value blocks) with fixed maximum dimensions. This segmentation allows the model to handle arbitrarily long sequences by processing them in parallel blocks, reducing the time and memory overhead associated with processing entire long sequences as single units while maintaining the model's ability to capture long-range dependencies.

Inventive Principle:
Principle #1Segmentation

3Adaptability or versatility

If attention mechanisms process longer input sequences, then model capability improves, but memory requirements increase

Engineering Contradiction:
Improvemodel capabilityVSAvoidmemory requirements
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent segments the computation into blocks that can be processed independently, allowing memory to be reused across different block computations. Instead of allocating memory for entire long sequences at once, the system processes blocks in parallel with controlled memory footprint, reducing overall memory requirements while maintaining the ability to process long sequences through repeated use of the same memory resources.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent enables memory recovery by processing blocks in parallel and independent batches. After each batch of block computations is completed, the intermediate results are discarded and memory is recovered for reuse in the next batch, allowing the system to handle longer sequences without proportionally increasing peak memory requirements.

Inventive Principle:
Principle #34Discarding and recovering

Data Source

PatentEP4651031A1Accelerated attention mechanism with parallel operations
Publication Date: 2025.11.19 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP4651031A1 patent drawingFigure 1A~1B
  • EP4651031A1 patent drawingFigure 2A
  • EP4651031A1 patent drawingFigure 2B

AI summary

An accelerated attention mechanism with parallel operations can improve machine learning technology by enabling execution of certain matrix multiplication operations in parallel with element-wise operations, leading to an increase in speed without quality loss. To compute attention values in a machine learning model, the mechanism can receive a query vector, key vector, and value vector and split each of these vectors into blocks. For a given query block, the mechanism can determine attention values by performing element-wise operations to update the attention values for the given query block based at least in part on previously computed attention scores for the given query block and a given key block. Concurrent with performance of at least some of the element-wise operations, the mechanism can perform a matrix multiplication operation using given query block and a next key block to determine attention scores for the given query block and next key block.