calender
Date & Time
Search
Datum
{{range.dates[index].day}}
{{range.dates[index].date}}
Time
Mornings Noon Afternoons Evenings
  • from
  • to
  • o'clock
Topic
Event
Properties
{{item.name}}
{{item.name}}
Exhibition venue

(please choose the desired areas)

Lecture language
Format

Event database

The event database contains all event-related information for the digital electronica conference and supporting event program.

Back to the EventList

Architectural Exploration for AI and Machine Learning with the flexibility of RISC-V

NOV
09
2020
09. NOV 2020

Lecture Embedded Forum > Tools & Software

12:20-12:40 h | Virtual

A modem SoC project is a combination of hardware and processor resources with the software application that together provide an optimal solution for the targeted end market use case. The AI and machine learning algorithms have been developed and perfected in cloud-based platforms and training datasets base on significant real-world data. To further enhance the performance for applications in edge devices the use of dedicated hardware can be explored to fine tune the optimum configuration of many core processors. RISC-V offers not just the flexibility to configure each node to match the performance requirements but can also support custom instructions and extensions. Custom instructions can help address the needs of the target applications by accelerating common routines or with dedicated core to core lightweight communication in many core arrays. This talk with cover the migration of cloud-based algorithms to dedicated hardware acceleration with the design flexibility RISC-V now offers system designs for SoC, FPGA and chiplet designs.

Subjects: Embedded systems

Speaker: Simon Davidmann (Imperas)

Type: Lecture

Speech: English

A modem SoC project is a combination of hardware and processor resources with the software application that together provide an optimal solution for the targeted end market use case. The AI and machine learning algorithms have been developed and perfected in cloud-based platforms and training datasets base on significant real-world data. To further enhance the performance for applications in edge devices the use of dedicated hardware can be explored to fine tune the optimum configuration of many core processors. RISC-V offers not just the flexibility to configure each node to match the performance requirements but can also support custom instructions and extensions. Custom instructions can help address the needs of the target applications by accelerating common routines or with dedicated core to core lightweight communication in many core arrays. This talk with cover the migration of cloud-based algorithms to dedicated hardware acceleration with the design flexibility RISC-V now offers system designs for SoC, FPGA and chiplet designs.

Speaker,
Imperas

Simon Davidmann

Simon Davidmann

Imperas

Show Profile

Informations

Speaker,
Imperas

Simon Davidmann
Simon Davidmann
CEO

More Events