kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Optimizing Neural Network Inference: A RISC-V Co-Processor Framework for 5G Communications
KTH, School of Electrical Engineering and Computer Science (EECS).
2024 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Optimera neurala nätverksslutningar : Ett RISC-V Co-Processor Framework för 5G-kommunikation (Swedish)
Abstract [en]

Artificial Intelligence (AI) is advancing rapidly, with neural network inference continually being optimized for high performance, compact size, and low power consumption. This progress creates diverse opportunities for integrating AI capabilities into the Reduced Instruction Set Computing - V (RISC-V) hardware ecosystem. Originating from UC Berkeley, RISC-V emphasizes a simplified Instruction Set Architecture (ISA) that enhances execution efficiency and scalability, making it suitable for the varied computational demands of neural network processing across different environments. In this research, we utilize the Spike Instruction Set Simulator (ISS) to evaluate the performance of the RISC-V architecture. We propose integrating a RISC-V core with a custom co-processor, optimizing and evaluating a Neural Processing Unit (NPU) architecture based on the RISC-V core. This evaluation uses a pre-trained neural network model provided by the Huawei HiSilicon BB IC team, aimed at efficiently executing neural networks within embedded 5G modems. The proposed architecture incorporates three types of Tightly-Coupled Memory (TCM)s to minimize data access to the main memory. Additionally, it employs a row-stationary-like data flow method within the Process Element (PE) array. This thesis details the matrix operation logic within the PE array of the RISC-V core architecture, the methods for executing and optimizing neural network models on RISC-V, and the results of our evaluations.

Abstract [sv]

Artificiell intelligens (AI) går snabbt framåt, med inferens av neurala nätverk som ständigt optimeras för hög prestanda, kompakt storlek och låg strömförbrukning. Dessa framsteg skapar olika möjligheter för att integrera AI-kapacitet i ekosystemet runt RISC-V (Reduced Instruction Set Computing V). RISC-V kommer från UC Berkeley och betonar en förenklad instruk- tionsuppsättningsarkitektur (ISA) som förbättrar skalbarheten och effektivitet i exekveringen, vilket gör den lämplig för de olika beräkningskraven för neurala nätverk i olika miljöer. I denna forskning använder vi Spike Instruction Set Simulator (ISS) för att utvärdera prestandan hos RISC-V-arkitekturen. Vi föreslår att integrera en RISC-V-kärna med en anpassad co-processor, optimera och utvärdera en neural processing unit (NPU)-arkitektur baserad på RISC-V-kärnan. Denna utvärdering använder ett förtränat neuralt nät som tillhandahålls av Huawei HiSilicon BB IC-teamet, och syftar till att effektivt köra neurala nätverk inom inbyggda 5G-modem. Den föreslagna arkitekturen innehåller tre typer av tätt kopplade minnen (TCM) för att minimera antalet accesser till huvudminnet. Dessutom använder den ett radstationär-liknande dataflöde inom PE-Arrayen (Processing Element Array). Artikeln beskriver logiken för matrisoperationer inom PE-arrayen, metoderna för att exekvera och optimera neurala nätverk på RISC-V, och resultaten av våra utvärderingar.

Place, publisher, year, edition, pages
2024. , p. 62
Series
TRITA-EECS-EX ; 2024:780
Keywords [en]
RISC-V, AI co-processor, PE Array, Convolution
Keywords [sv]
RISC-V, AI co-processor, PE Array, Convolution
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:kth:diva-360210OAI: oai:DiVA.org:kth-360210DiVA, id: diva2:1939024
External cooperation
Huawei HiSilicon BB IC-teame
Supervisors
Examiners
Available from: 2025-02-24 Created: 2025-02-20 Last updated: 2025-02-24Bibliographically approved

Open Access in DiVA

fulltext(3467 kB)245 downloads
File information
File name FULLTEXT01.pdfFile size 3467 kBChecksum SHA-512
319b89fa50984a4faa8c15d41430e972baead2715f7a1a3015308b48198e28915c6c69cd942a28a2a9d6a46109e81fcb0ce59e9b398360be67b5463b2250eedb
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Electrical Engineering, Electronic Engineering, Information Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 245 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 588 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf