Intel corporation (20240334245). PROCESSING OF PACKET FRAGMENTS simplified abstract

PROCESSING OF PACKET FRAGMENTS

Organization Name

intel corporation

Inventor(s)

John J. Browne of Limerick (IE)

Andrey Chilikin of Limerick (IE)

Elazar Cohen of Haifa (IL)

Joseph Hasting of Orefield PA (US)

James Clee of Orefield PA (US)

Jerry Pirog of Easton PA (US)

Jamison D. Whitesell of Pennsburg PA (US)

Ambalavanar Arulambalam of Center Valley PA (US)

Anjali Singhai Jain of Portland OR (US)

Andrew Cunningham of Ennis (IE)

Ruben Dahan of Petah Tikva (IL)

PROCESSING OF PACKET FRAGMENTS - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240334245 titled 'PROCESSING OF PACKET FRAGMENTS

The abstract describes a network interface device that offloads processing of packet fragments to an accelerator, prioritizes dropping of fragments over non-fragmented packets, and handles congestion by dropping fragments or halting reassembly.

• Offloading processing of packet fragments to an accelerator
• Prioritizing dropping of fragments over non-fragmented packets
• Handling congestion by dropping fragments or halting reassembly
• Reassembling fragments into a reassembly packet
• Forwarding a second packet to indicate congestion and dropped fragments

Potential Applications: - Network infrastructure - Data centers - Cloud computing

Problems Solved: - Efficient packet processing - Congestion management - Accelerated data transfer

Benefits: - Improved network performance - Reduced packet loss - Enhanced data processing speed

Commercial Applications: Title: "Accelerated Network Interface Device for Efficient Packet Processing" This technology can be used in networking equipment, data centers, and cloud services to optimize packet processing and improve overall network performance.

Questions about the technology: 1. How does the device prioritize dropping fragments over non-fragmented packets? 2. What are the potential implications of using an accelerator for packet processing in terms of network efficiency and speed?

Original Abstract Submitted

examples described herein relate to a network interface device that performs: offloading processing of fragments of a packet to an accelerator; processing non-fragmented packets; and prioritizing dropping of fragments of the packet over dropping of non-fragmented packets. offloading processing of fragments of the packet to the accelerator can include: the accelerator performing: reassembling the fragments of the packet into a first reassembly packet; and based on congestion associated with at least one of the fragments of the packet of the first reassembly packet: dropping fragments of the first reassembly packet associated with one or more flows; halting reassembly of the first reassembly packet; and forwarding a second packet to a host system, wherein the second packet indicates that congestion occurred, identifies one or more impacted flows, and indicates a number of dropped packet fragments.