用于张量编译器的配置交叉注意力的图形神经网络
Dmitrii Khizbullin1, Eduardo Rocha de Andrade2, Thanh Hau Nguyen2
1King Abdullah University of Science and Technology (KAUST), Thuwal, Saudi Arabia.
Frontiers in artificial intelligence
|September 5, 2025
概括
通过智能选张量配置, 加快神经网络推断. 这种AI张量编译器显著提高了AI数据中心的性能并降低了能源消耗.
科学领域:
- 人工智能
- 计算机科学
- 机器学习
背景情况:
- 神经网络推断需要高效的张量运算.
- 优化张量布局 (转换,) 是一个复杂的组合.
- 现有的编译器通常依赖于启发式,限制性能增长.
研究的目的:
- 介绍TGraph,一种用于优化神经网络推理的神经图架构.
- 开发一个能够选快速张量配置的AI张量编译器.
- 减少人工智能推断的计算成本和能源足迹.
主要方法:
- 用计算图形表示神经网络推断.
- 使用图形架构来探索张量排列.
- 开发一个与传统方法不同的人工智能张量编译器框架.
主要成果:
- 与基线的29.8%相比,TGraph显著提高了整个布局集合的平均Kendall的t.
- 提出的方法作为一个AI张量编译器,优于基于启发式的方法.
- 估计潜在的二氧化碳排放量相当于人工智能数据中心地区50%以上的家庭排放量.
结论:
- TGraph提供了一种强大的新方法来优化神经网络推断.
- 人工智能张量编译器框架显示了显著的性能改进.
- 这项工作对人工智能的能源效率有重大影响.
相关概念视频
Neural Circuits
1.5K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
1.5K
Cross Product
325
The cross product is a fundamental concept in vector algebra that is a vector operation on two different vectors to obtain a third vector. Unlike the scalar product, the cross product results in a vector quantity perpendicular to both the original vectors.
The magnitude of the cross product is obtained by multiplying the magnitude of both the vectors and the sine of the angle between them. This means that a larger angle between the vectors will lead to a greater magnitude of the cross product.
The magnitude of the cross product is obtained by multiplying the magnitude of both the vectors and the sine of the angle between them. This means that a larger angle between the vectors will lead to a greater magnitude of the cross product.
325
Multi-input and Multi-variable systems
147
Cruise control systems in cars are designed as multi-input systems to maintain a driver's desired speed while compensating for external disturbances such as changes in terrain. The block diagram for a cruise control system typically includes two main inputs: the desired speed set by the driver and any external disturbances, such as the incline of the road. By adjusting the engine throttle, the system maintains the vehicle's speed as close to the desired value as possible.
In the absence...
In the absence...
147
Vector Product (Cross Product)
12.4K
Vector multiplication of two vectors yields a vector product, with the magnitude equal to the product of the individual vectors multiplied by the sine of the angle between both the vectors and the direction perpendicular to both the individual vectors. As there are always two directions perpendicular to a given plane, one on each side, the direction of the vector product is governed by the right-hand thumb rule.
Consider the cross product of two vectors. Imagine rotating the first vector about...
Consider the cross product of two vectors. Imagine rotating the first vector about...
12.4K
Vector Algebra: Graphical Method
13.5K
Vectors can be multiplied by scalars, added to other vectors, or subtracted from other vectors. The vector sum of two (or more) vectors is called the resultant vector or, for short, the resultant.
We use the laws of geometry to construct resultant vectors, followed by trigonometry to find vector magnitudes and directions. For a geometric construction of the sum of two vectors in a plane, we follow the parallelogram rule. Suppose two vectors are at arbitrary positions. Translate either one of...
We use the laws of geometry to construct resultant vectors, followed by trigonometry to find vector magnitudes and directions. For a geometric construction of the sum of two vectors in a plane, we follow the parallelogram rule. Suppose two vectors are at arbitrary positions. Translate either one of...
13.5K
Sequence Networks of Rotating Machines
140
A Y-connected synchronous generator, grounded through a neutral impedance, is designed to produce balanced internal phase voltages with only positive-sequence components. The generator's sequence networks include a source voltage that is exclusively in the positive-sequence network. The sequence components of line-to-ground voltages at the generator terminals illustrate this configuration.
Zero-sequence current induces a voltage drop across the generator's neutral impedance and other...
Zero-sequence current induces a voltage drop across the generator's neutral impedance and other...
140


