View source for Intel corporation (20240126506). FLOATING-POINT DECOMPOSITION CIRCUITRY WITH DYNAMIC PRECISION simplified abstract

=FLOATING-POINT DECOMPOSITION CIRCUITRY WITH DYNAMIC PRECISION=


==Organization Name==

[[:Category:intel corporation|intel corporation]]


[[Category:intel corporation]]

==Inventor(s)==

[[:Category:Roberto Dicecco of Toronto (CA)|Roberto Dicecco of Toronto (CA)]][[Category:Roberto Dicecco of Toronto (CA)]]

[[:Category:Joshua Fender of East York (CA)|Joshua Fender of East York (CA)]][[Category:Joshua Fender of East York (CA)]]

[[:Category:Shane O'connell of Toronto (CA)|Shane O'connell of Toronto (CA)]][[Category:Shane O'connell of Toronto (CA)]]

==FLOATING-POINT DECOMPOSITION CIRCUITRY WITH DYNAMIC PRECISION - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240126506 titled 'FLOATING-POINT DECOMPOSITION CIRCUITRY WITH DYNAMIC PRECISION

==Simplified Explanation==

The circuitry described in the abstract decomposes block floating-point numbers into lower precision floating-point numbers by utilizing high precision storage circuit, input selectors, and a low precision block floating-point vector circuit.

* High precision storage circuit provides high precision floating-point numbers.
* Input selectors receive high precision floating-point numbers and generate corresponding lower precision floating-point components with adjusted exponents.
* Low precision block floating-point vector circuit combines the lower precision floating-point components generated by the input selectors.
* The lower precision floating-point components can be processed spatially or over multiple iterations over time.

== Potential Applications ==
This technology can be applied in:
* Signal processing
* Image processing
* Machine learning algorithms

== Problems Solved ==
* Efficient decomposition of block floating-point numbers
* Reduction of computational complexity
* Improved accuracy in numerical calculations

== Benefits ==
* Higher efficiency in processing floating-point numbers
* Enhanced accuracy in computations
* Potential for faster processing speeds

== Potential Commercial Applications ==
Optimized for:
* High-performance computing systems
* Scientific computing applications
* Data analytics platforms

== Possible Prior Art ==
One possible prior art for this technology could be the use of fixed-point arithmetic in signal processing applications to reduce computational complexity.

=== What is the impact of this technology on computational efficiency in signal processing applications? ===
This technology significantly improves computational efficiency by decomposing block floating-point numbers into lower precision components, reducing the overall computational complexity and enhancing processing speeds in signal processing applications.

=== How does this innovation contribute to the accuracy of numerical calculations in machine learning algorithms? ===
By providing a mechanism to adjust exponents and combine lower precision floating-point components, this innovation ensures improved accuracy in numerical calculations for machine learning algorithms, leading to more reliable results and predictions.


==Original Abstract Submitted==

circuitry for decomposing block floating-point numbers into lower precision floating-point numbers is provided. the circuitry may include a high precision storage circuit configured to provide high precision floating-point numbers, input selectors configured to receive the high precision floating-point numbers from the high precision storage circuit and to generate corresponding lower precision floating-point components with adjusted exponents, and a low precision block floating-point vector circuit configured to combine the various lower precision floating-point components generated by the input selectors. the lower precision floating-point components may be processed spatially or over multiple iterations over time.



[[Category:G06F7/485]]
[[Category:G06F7/483]]
[[Category:G06F7/487]]
[[Category:G06F7/499]]
[[Category:G06F7/544]]
[[Category:G06F17/16]]