18491246. METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS simplified abstract (Intel Corporation)
Contents
- 1 METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Original Abstract Submitted
METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS
Organization Name
Inventor(s)
Amit Bleiweiss of Yad Binyamin (IL)
Eliran Zimmerman of Maalot (IL)
METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18491246 titled 'METHODS, SYSTEMS, ARTICLES OF MANUFACTURE AND APPARATUS TO MAP WORKLOADS
Simplified Explanation
The patent application describes a method and apparatus for mapping workloads in a neural network by defining performance characteristic targets, applying resource configurations, calculating results metrics, and generating resource mapping files.
- The apparatus includes a constraint definer, action determiner, reward determiner, and layer map generator to optimize resource assignments for different layers of the neural network based on performance targets and resource performance metrics.
- The method aims to improve the efficiency and performance of neural networks by mapping resources effectively to meet performance goals.
Potential Applications
This technology could be applied in various fields such as:
- Artificial intelligence
- Machine learning
- Data processing
Problems Solved
This technology helps solve the following problems:
- Resource allocation inefficiencies in neural networks
- Performance bottlenecks in complex workloads
Benefits
The benefits of this technology include:
- Improved performance of neural networks
- Efficient resource allocation
- Enhanced scalability of neural network applications
Potential Commercial Applications
Potential commercial applications of this technology could include:
- Cloud computing services
- Data centers
- AI hardware development
Possible Prior Art
One possible prior art could be research on resource allocation optimization in neural networks, such as studies on workload mapping and performance optimization strategies.
Unanswered Questions
How does this technology compare to existing workload mapping techniques in neural networks?
This article does not provide a direct comparison to existing workload mapping techniques in neural networks. It would be interesting to see a detailed analysis of how this method differs from or improves upon current approaches.
What are the specific performance metrics used to calculate the results metric in this method?
The article mentions calculating a results metric based on resource performance metrics and performance characteristic targets, but it does not specify the exact performance metrics used in the calculation. Understanding the specific metrics involved could provide insights into the effectiveness of this approach.
Original Abstract Submitted
Methods, apparatus, systems and articles of manufacture are disclosed to map workloads. An example apparatus includes a constraint definer to define performance characteristic targets of the neural network, an action determiner to apply a first resource configuration to candidate resources corresponding to the neural network, a reward determiner to calculate a results metric based on (a) resource performance metrics and (b) the performance characteristic targets, and a layer map generator to generate a resource mapping file, the mapping file including respective resource assignments for respective corresponding layers of the neural network, the resource assignments selected based on the results metric.