RISC-V Alliance Japan

Menu
  • 日本語日本語
  • Events
    • 2026 Spring
    • 2025 Autumn
    • 2024 Summer
    • 2024 Winter
    • 2023 Summer
    • 2022 Autumn
    • 2022 Spring
    • 2022 Vietnam
    • 2021 Autumn
    • 2021 Spring
    • 2020
    • 2020 Vietnam
    • 2019 Tokyo
  • RISC-V International
  • Facebook
  • Twitter

RISC-V Day Tokyo 2026 Spring

Thursday, March 5, 2026
9:00–18:00 Japan Standard Time (UTC+9)
Click here for attendees to register (Peatix)
Click here to submit your proposal (Talks / Posters) — Proposals will undergo a selection process
Click here for the Call for Sponsors: Presentations & Exhibits

 

 

Venue Layout

CONFERFENCE A @Ito Memorial Hall (Click thumbnail to expand the picture)

時間
写真

発表題目
発表概要

発表者
経歴

13:10-13:30

Welcome Remarks
Makoto Ikeda, representing the University of Tokyo, addresses the critical shift in the semiconductor landscape where RISC-V has evolved from an academic standard into a vital industrial pillar. His remarks focus on the “democratization” of chip design, highlighting how open-standard architectures enable Japanese industry and academia to co-innovate in high-growth areas like AI, robotics, and automotive systems.

Makoto Ikeda | Professor at the University of Tokyo
Dr. Ikeda is Special Advisor to the President of the University of Tokyo and Director of d.lab at the University of Tokyo. He also serves as Science Adviser to Japan’s Ministry of Education, Culture, Sports, Science and Technology (MEXT), contributing to national policy in science, technology, and advanced research strategy.

A Professor at the University of Tokyo since 2013, he previously served as Associate Professor, Lecturer, and Research Associate, building a distinguished academic career spanning nearly three decades at Japan’s leading research university. His work bridges advanced semiconductor technology, integrated circuit design, and system-level innovation, with a strong emphasis on translating fundamental research into practical societal impact.

13:30-14:00

Standard Cell Library Development and Design Infrastructure Establishment for the 2nm Process at Rapidus — Powered by Raads
 Standard cells are small fundamental circuits composed of a single logic gate or up to a dozen gates. Only when a library containing several hundred types of such cells is properly prepared can full SoC design be realized. A complete cell library must include the various models and views required by EDA tools, such as electrical characteristic models, HDL models for logic simulation, and physical views including LEF/DEF and GDS-II for place-and-route.
 At Rapidus, within the Raads (Rapidus AI-Assisted Design Solutions) design environment, standard cell library development is a foundational element of the 2nm design infrastructure. While user-facing production libraries are outsourced to a third party (Synopsys), Rapidus also develops in-house libraries to support internal evaluation, validation, and benchmarking of external libraries. (Note: Presentation title and abstract prepared by the organizer.)

Koki Tsurusaki | Director, Design Technology Division, Engineering Center, Rapidus Corporation
 Koki Tsurusaki was involved in semiconductor physical design and floorplanning technology development at Renesas Electronics and is listed as an inventor on patents related to design optimization methodologies. He currently serves as Director of the Design Technology Division at Rapidus Corporation, where he leads the development of the AI-enabled design environment “Raads” and the establishment of design infrastructure for advanced process technologies.

 

14:00-14:30

Open Silicon for the AI Era: Tenstorrent and AI Everywhere
 As the business environment is rapidly transforming with the rise of AI, strengthening Japan’s market competitiveness has become an urgent priority. This requires advances not only in semiconductor manufacturing, but also in key chips, software, and human resource development.This talk introduces Tenstorrent’s open and advanced technology solutions and their role in building a competitive AI and semiconductor ecosystem.

Mamoru Nakano, APAC Region Manager & Japan Country Manager at Tenstorrent
Since the company’s establishment in January 2023, he has been leading its business expansion and organizational development in the Japanese market. Prior to joining Tenstorrent, he served as Country Manager of Graphcore Japan, overseeing sales, marketing, and customer engagement. Before that, he spent approximately 16 years as President and Representative Director of Cray Inc. Japan, driving the HPC business and supporting growth across the Asia region. Earlier in his career at HP, he held the role of General Manager for HPC and Linux business across Asia, building a long-standing track record at the forefront of high-performance computing and cloud infrastructure.

14:30-15:00

SiFive 2nd Generation Intelligence Family Introduction
 This talk presents the second generation of the SiFive Intelligence Family, a suite of RISC-V-based AI accelerator IP cores. These cores support efficient edge AI acceleration across applications from low-power far-edge devices to cloud-scale systems. Built on the open RISC-V standard with custom extensions, the family ensures compatibility with the RVA23 profile, enabling seamless integration with custom accelerators for AI inference, large language models, and related workloads.

Yoshihito Kondo | Representative Director, SiFive Japan
 Yoshito Kondo is a semiconductor industry leader serving as Representative Director of SiFive Japan, the Japanese subsidiary of U.S.-based SiFive. He joined Sony Corporation in 1990, where he was engaged in the development of system LSIs, including media processors. In 2019, he moved to DTS Insight Corporation, where he was involved in SiFive’s distribution and technical support activities in the Japanese market. He joined SiFive Japan in October 2022, initially overseeing technical support including field application engineering, and later assumed leadership of the company, driving the adoption and business expansion of RISC-V processors.

15:00-15:30
PDF

Software Guided Intelligence for Physical AI Platform Design with RISC-V
 In this presentation, Takuya Katayama will show how the change from hardware-first to software-first platform architecture is driving changes for embedded design. With RISC-V and the power of virtual models, it is now possible to start with the workload, select and optimize the hardware, and finalize the platform without time-consuming, budget-busting tape-out cycles. The appearance of software-guided intelligence is enabling changes in design methodology that deliver faster time to market, lower total cost of development, and more efficient edge silicon. This is supported by the open, modular and extensible nature of RISC-V instruction set architecture for building workload-focused instructions and extensions.

Takuya Katayama, MIPS Technical Marketing for Software & Tools
Mr. Katayama is regarded as a business leader in the semiconductor industry who is involved in the product management domain at MIPS. Within the company’s processor and platform ecosystem, built around its RISC architecture IP, he plays a role related to software and tools strategy. He is engaged in activities that bridge technology and market needs in the embedded and semiconductor sectors.

15:30-16:00

Powering the Future of Compute with Andes RISC-V
 The future of computing is being dramatically transformed by open and scalable architectures. Evolution in the fields of AI, edge, and cloud requires flexible and optimizable designs, and RISC-V is a core technology for this.
In this presentation, Frankwell Lin, CEO of Andes Technology, will introduce the latest trends and strategies for RISC-V, which is being put into practical use at the commercial level. He will explain how customers can develop differentiated products more quickly through high-performance, low-power processor IP, vector extensions, and a powerful software ecosystem.

Frankwell Lin, Chairman & CEO of Andes Technology (Taiwan)
 Frankwell Lin, Chairman of Andes Technology, started his career being as application engineer in United Microelectronics Corporation (UMC) while UMC was an IDM with its own chip products, he experienced engineering, product planning, sales, and marketing jobs with various product lines in UMC. In 1995, after four years working on CPU chip product line as business director, he was transferred to UMC-Europe branch office to be its GM when UMC reshaped to do wafer foundry service, he lead UMC-Europe to migrate itself
from selling IDM products to selling wafer foundry service. In 1998, after 14 years working in UMC, Frankwell switched job to work in Faraday Technology Corporation (Faraday), he lead ASIC business development as starting, then on-and-off leading ASIC implementation, chip backend service, IP business development, industry relationship development (IR), as well as Faraday's spokesperson, in 2004, he started to lead the CPU project spin off operation of Faraday. Frankwell became co-founder of Andes Technology Corporation in
2005 when it was found up, he formally took position to be Andes' President since 2006 and got promoted as Chairman and CEO in 2021.
 Frankwell received BSEE degree of Electrophysics from the National Chiao-Tung University, Taiwan, and MSEE degree of Electrical and Computer Engineering from Portland State University, Oregon, USA. Under his management, Andes has been recognized as one of leading suppliers of embedded CPU IP in semiconductor industry. Andes also won the reputation of leading technology company with awards such like 2012 EE Times worldwide Silicon 60 Hot Startups to Watch, 2015 the Deloitte Technology Fast 500 Asia Pacific award, etc. Frankwell received accolade award of Outstanding Technology Management Performance, Taiwan, in 2015 and ERSO award in 2020 for his contribution to the high-tech industry. Frankwell is the Board Director of RISC-V International since 2020. Frankwell was also
elected to be the Chairman of RISC-V Taiwan Alliance (RVTA) since 2023.

16:00-16:20


A Case Study of Custom AI Accelerator Development Using RISC-V
This article uses actual examples to introduce the role RISC-V has played in developing a custom AI accelerator for FPGAs at Future, as well as the flexibility of customization made possible by RISC-V’s open architecture.

Yuki Miyashita | Future Corporation ki-labs | FPGA Lead
At a general electric manufacturer, he worked on FPGA development for autonomous driving and medical equipment, pursuing system design with high reliability, low latency, and real-time processing. Leveraging his expertise, he is currently a member of Future Corporation’s in-house R&D organization, “ki-labs,” where he leads the design and development of hardware acceleration as an FPGA lead.
Shu Kitamoto | Future Corporation ki-labs | Software Engineer
After working on business systems development at an SIer, he is now a member of Future Corporation’s “ki-labs.” As a software engineer, he is responsible for implementing the processor instruction issuing layer. He pursues faster AI inference by optimizing the boundary between higher-level software and hardware. His hobby is competitive programming (AtCoder Blue).
Ryuji Fuchikami | Future Corporation ki-labs | Architect
Graduated from Kyushu Institute of Technology in 1999. He has worked in the research and development department of a major electronics manufacturer for over 20 years, establishing image sensing and ultra-high-speed, low-latency real-time signal processing technologies. He specializes in the boundary between hardware and software, and has pursued optimization technologies that utilize the characteristics of computer architecture, including the invention of the deep learning model “LUT-Network” for FPGAs. He has written numerous articles, including in “Interface” (CQ Publishing). He is currently working at ki-labs, Future Inc., designing AI accelerator architectures.

16:20-16:40

TRACE32: Advanced Debug and Trace Solutions for RISC-V
 RISC-V adoption is rapidly expanding across a wide range of domains, from embedded devices to high-performance SoCs. As systems become increasingly complex through multi-core integration and heterogeneous architectures, the importance of advanced development and analysis environments continues to grow.
 This talk introduces the latest trends in high-end debug and trace technologies for RISC-V systems. It explains how TRACE32’s comprehensive debug solutions—such as non-intrusive observation and synchronized multi-core debugging—can contribute to efficient analysis and resolution of complex system-level issues.

Yuji Mori | Representative Director, Lauterbach Japan Co., Ltd. (Germany)
 Yuji Mori serves as Representative Director of Lauterbach Japan Co., Ltd., the Japanese subsidiary of Germany-based Lauterbach, a global leader in debug and trace solutions for embedded systems. Over the course of his career, he has been involved in development and technical support within Japan’s major semiconductor and embedded industries, including companies such as Hitachi and Renesas Electronics, building extensive experience in SoC and embedded system design and debugging. At Lauterbach, Mori oversees business development and customer engagement in the Japanese market, promoting more efficient development for increasingly complex next-generation systems such as multi-core SoCs and RISC-V platforms through advanced technologies including non-intrusive tracing and synchronized multi-core debugging.

16:40-17:00
PDF

Debugging Environments Supporting RISC-V Development: Tools and Practical Challenges in Real Hardware Development
 This session addresses common challenges encountered in RISC-V software development—such as initial boot bring-up, RTOS debugging, and trace analysis—and introduces practical development and verification methodologies using ICE.
 As RISC-V adoption accelerates across industrial applications, the most critical requirement for development teams is an environment that enables efficient bring-up and in-depth analysis on real hardware. DTS INSIGHT provides embedded developers with robust debug and trace environments through its RISC-V–compatible ICE (In-Circuit Emulator) products.

Tetsuo Asada | DTS Insight Corporation (Japan)
 After joining DTS Insight Corporation (formerly Yokogawa Digital Computer Corporation), he worked on the development of the debugger microVIEW-PLUS. He later engaged in customer support and subsequently contributed to the development of the dynamic analysis tool TRQer. He is currently responsible for the development of adviceXross.
 For many years, he has been involved in both the development and support of embedded development support tools, gaining deep expertise in the debugging mechanisms provided by CPU architectures. Leveraging this experience, he is engaged in the development of debugging and dynamic analysis tools.

17:00-17:10

Closing Remarks: Leveraging RISC-V to Revitalize Japan’s Digital Sector
 Japan’s digital sector is highly dependent on overseas sources for cloud infrastructure, AI computing resources, and core server technologies, and the import structure of data center technology is restricting industrial competitiveness and digital sovereignty. RISC-V represents one of the few strategic opportunities to bring standardization and implementation capabilities domestically through an open ecosystem, and could become an important pillar for rebuilding the semiconductor and digital industrial infrastructure.
 In closing, I would like to reorganize today’s discussion as the intersection of policy, standardization, implementation, and human resource development, and confirm that industry, government, and academia working together to establish the RISC-V ecosystem in Japan is essential for revitalizing Japan’s digital sector and establishing next-generation industrial competitiveness.

Takuhiro Nakajo | Associate Professor, Institute of Engineering, Graduate School of Tokyo University of Agriculture and Technology
 Takuhiro Nakajo graduated from the Department of Electrical Engineering, Faculty of Engineering, Kobe University in 1985, and received his Master’s degree in Electronic Engineering from the Graduate School of Engineering at Kobe University in 1987. Since 1999, he has served as an Associate Professor in the Division of Advanced Information Technology and Computer Science, Institute of Engineering, Graduate School of Tokyo University of Agriculture and Technology.
 His research areas include processor architecture, embedded systems, and reconfigurable computing, and he has been engaged for many years in research aimed at advancing computer systems.
 In addition to his research activities, he is widely known for his significant contributions to computer science education in Japan. He has served as a co-translator of the Japanese editions of the internationally renowned textbook Computer Architecture: A Quantitative Approach by John L. Hennessy and David A. Patterson (commonly known in Japan as “HenePata”), from the 4th through the 6th editions, as well as Digital Design and Computer Architecture by David Money Harris and others (including the 1st edition, ARM edition, and 2nd edition). Through these translations, he has helped disseminate modern architectural theory in Japan.
 He is a member of the Information Processing Society of Japan (IPSJ), the Institute of Electronics, Information and Communication Engineers (IEICE), the IEEE Computer Society, and ACM, and holds a Ph.D. in Engineering.

 

CONFERENCE B @Multi-Purpose Space (Click thumbnail to expand the picture)

Time
Pic

Presentation Title
Abstract

Presenter
Bio

14:45-15:05

2nd Generation Intelligence Products Technology Explainer 
 The SiFive 2nd Generation Intelligence technology family introduces a comprehensive suite of RISC-V products, including the X100, X200, X300, and XM series, designed to solve the growing fragmentation and complexity within modern AI compute requirements. This architecture employs a unified Scalar, Vector, and Matrix (SVM) compute strategy that efficiently manages diverse tasks ranging from basic control flow to intensive convolutional processing.
 A key technical breakthrough is the implementation of memory latency tolerance, which utilizes a refined pipeline allowing scalar units to pre-fetch data and effectively eliminate vector pipeline stalls. Furthermore, the inclusion of direct core connectivity through the Scalar Coprocessor Interface (SSCI) and Vector Co-processor Interface (VCIX) enables high-bandwidth integration with custom accelerators and companion cores. The performance is further augmented by a dedicated hardware Exponential Unit specifically engineered to accelerate Softmax operations by significantly reducing the cycles required for exponential functions. Together, these features establish a scalable and low-latency framework capable of handling a wide array of AI workloads across the entire compute spectrum, from power-efficient IoT devices to high-performance datacenter environments.

Kunio Uchiyama | Lab Master (Invited Senior Researcher)  | AI Chip Design Open Innovation Laboratory |  National Institute of Advanced Industrial Science and Technology (AIST)
Kunio Uchiyama currently serves as an Invited Senior Researcher at the AI Chip Design Open Innovation Laboratory within the National Institute of Advanced Industrial Science and Technology (AIST). He is also prominently recognized as the Chief Executive Officer (CEO) and Director of the AI Chip Design Center (AIDC), a joint initiative between AIST and the University of Tokyo that transitioned from a NEDO-funded project to a shared facility of AIST in 2023. With a professional background that includes serving as a Corporate Officer and Chief Scientist at Hitachi, Ltd., he is a Fellow of both the IEEE and IEICE and has held a position on the Board of Governors of the IEEE Computer Society. His recent work focuses on promoting the democratization of chip design and accelerating the development of AI-specific hardware and post-Moore semiconductor technologies within the Japanese startup ecosystem.

15:10-15:30

AIST’s Initiatives in Advanced Semiconductor Research and Development
 Semiconductors are utilized in various aspects of our daily lives and have become an indispensable presence for addressing social challenges and strengthening industrial competitiveness. Consequently, securing a stable supply of semiconductors has become a critical issue from the perspective of economic security, much like energy resources.
In addition, the drive for higher performance in AI chips—led by generative AI—has resulted in ever-increasing demands for improvements in semiconductor performance itself. In other words, the further evolution of semiconductors is now essential. In this presentation, I will introduce the initiatives being undertaken by AIST to drive this continued evolution of semiconductor technology.

Dr. Meishoku Masahara | National Institute of Advanced Industrial Science and Technology (AIST)
Director, Semiconductor Frontier Research CenterDr. Masahara completed the doctoral program at the Graduate School of Science and Engineering, Waseda University, in 1995, earning a Ph.D. in Engineering. In 2002, he joined the National Institute of Advanced Industrial Science and Technology (AIST), where he has been consistently dedicated to research in advanced semiconductors. Currently, as the Director of the Advanced Semiconductor Research Center, he leads AIST’s research and development initiatives in the field of advanced semiconductors. He also served as a visiting researcher at imec in Belgium from 2005 to 2006. Since 2022, he has participated as a member of the Leading-edge Semiconductor Technology Center (LSTC) and serves as the General Manager of the Technology Development Management Division.

15:35-15-55
 

Accelerating RISC-V Development via Hybrid Co-Emulation and Shift-Left Methodologies
S2C EDA Solutions offers an integrated verification ecosystem designed to accelerate RISC-V development through a “Shift-Left” methodology, which allows software and hardware teams to work in parallel. The platform is built on three core pillars: Genesis Architect for early microarchitecture analysis and virtual modeling, OmniArk Emulation for scalable system integration and compliance testing, and Prodigy FPGA Prototyping for high-performance software development and hardware evaluation. A critical technical component of this suite is the Hybrid Co-Emulation Bridge, which synchronizes non-RTL virtual models (like QEMU or SystemC) with RTL hardware designs. This is achieved using a sophisticated interface of proxies for untimed software transactions, transactors that convert these into cycle-accurate hardware events, and dedicated communication channels to ensure seamless data flow. By bridging the gap between high-level architectural exploration and physical prototyping, this technology enables developers to achieve early software maturity and rigorous certification before final silicon is produced.

S2C EDA Inc. (Shanghai)

16:00-16:20

Specification-accurate customizable RISC-V cores and accelerators
 Keysom provides a silicon-proven RISC-V and accelerator IP ecosystem centered on theorem-proving formal verification, which mathematically guarantees equivalence between hardware specifications and implementation. Their modular architecture supports custom instructions and optional features, integrated via a Smart LLVM compiler and standard interfaces like AXI and CV-X-IF.
 The CoreXplorer no-code EDA tool automates Design Space Exploration, allowing users to generate application-specific cores and full HDK/SDK packages in minutes. By tailoring hardware to specific requirements, Keysom achieves significant performance gains, including up to a 30% increase in frequency and a 22% reduction in energy consumption. This automated approach bridges the gap between hardware and software, delivering rigorous correctness and optimized efficiency for edge and AI applications.

Luca TESTA, PhD – Co-founder and COO | Keysom Inc. 
 Luka is the Co-founder and COO of Keysom, where he leads the development of innovative, application-specific RISC-V processor architectures. With a career spanning over 15 years in high-end hardware engineering, he specializes in bridging the gap between complex hardware specifications and efficient software execution. Under his leadership, Keysom has pioneered a unique methodology—rooted in formal mathematical verification—to deliver silicon-proven RISC-V cores that are automatically optimized from application source code.
 Luka’s technical foundation is built on deep expertise in analog and RF circuit design. He earned his PhD in France in 2010 in collaboration with STMicroelectronics, focusing on 65nm CMOS and BiCMOS technologies for aerospace and telecommunications. His professional journey includes serving as an Analog/RF Design Engineer for Ku-band satellite systems and a five-year tenure at Amplitude Laser, where he rose to become the Head of the Electronics Department. In 2018, he transitioned into entrepreneurship by co-founding Koncepto.io, an engineering firm dedicated to custom embedded hardware and Linux solutions. 
   At Keysom, Luka leverages this extensive background in board-level design and system architecture to provide the industry with highly configurable cores. By integrating advanced EDA tools like CoreXplorer with a Smart LLVM compiler, he enables developers to achieve significant gains in frequency, energy efficiency, and silicon area. Based in Bordeaux, Luka remains a key figure in the European RISC-V ecosystem, focusing on the deployment of verified, high-performance silicon for the next generation of edge and AI computing.

16:25-16:45

Development and Performance Evaluation of a RISC-V Based CPU Chiplet at Tenstorrent
We’ll provide an overview of the development policy and progress of chiplet technology in Tenstorrent, updating the current status of Tenstorrent chiplet development being advanced by AIDC. Furthermore, we’ll discuss efforts regarding correlation with model simulators aimed at improving the accuracy of performance evaluation, focusing on the execution environment of the CPU “Ascalon” currently in the integration phase.

Masayuki Kimura | Sr.Staff Engineer | Tenstorrent Japan  
 Masayuki Kimura is a veteran RISC-V CPU Architect at Tenstorrent, where he specializes in high-performance hardware engineering and processor microarchitecture. With a career rooted in deep technical expertise, he has extensive experience in the development of RISC-V microprocessors, the integration of MIPS processors for High-Performance Computing (HPC), and the creation of instruction set simulators for the RH850, MIPS, and RISC-V architectures. His background also includes pioneering work on Through Chip Communication (TCI) for inter-chip wireless communication devices.

 A long-standing leader in the Japanese RISC-V community, Kimura-san has been a frequent contributor and speaker at industry forums since 2017, widely sharing his technical insights through an extensive library of public presentations. Prior to his current role at Tenstorrent, he held key engineering positions at Renesas Electronics, contributing to the evolution of embedded and automotive processor technology. He holds his academic credentials from the University of Tokyo, combining a rigorous theoretical foundation with decades of hands-on experience in silicon design.

 

TUTORIAL A @Multi-Purpose Space (Click thumbnail to expand the picture)

Time
Pic

Presentation Title
Abstract

Presenter
Bio

10:00-10:10

Accelerating AI via “Architecture” RV23
 The RVA23 profile currently progressing within RISC-V leverages its inherent openness to further advance the “Data Short-circuiting” equation pioneered by NVIDIA. Existing “off-the-shelf” processors, such as x86 and ARM, are burdened by a “generality tax” in the form of legacy cache hierarchies and rigid bus structures.
 Through highly flexible standardization, RISC-V enables the direct integration of AI engines into the CPU pipeline and allows for the efficient execution of confidential computing on AI accelerators. To integrate these functional implementations into the software ecosystem, the implementation of standardized architectural features that transcend individual microarchitectures is essential.
 From standard CPU IP to hyperscale custom silicon, mechanisms that reduce data movement costs through unified memory management are also effective for power-constrained edge devices. The new era of AI-native hardware will be realized through the “democratization of architecture.”

Shumpei Kawasaki | Representative Director, RISC-V Association Japan 
 A semiconductor strategist whose career spans the intersection of processor architecture design and international technical alliances in the software sector. After completing his studies in Computer Science at the University of Illinois, he joined the Semiconductor Division of Hitachi, Ltd.
 In collaboration with Xerox PARC, he developed the “AI32,” a 32-bit AI processor capable of directly executing Smalltalk-80. This pioneering endeavor to integrate object-oriented environments with dedicated hardware formed the foundation for subsequent RISC design philosophies.
 Later, he was involved in the legal mediation of the Motorola lawsuit concerning instruction sets, gaining firsthand experience in the dynamics between architecture and intellectual property. He designed the instruction sets for SEGA’s consumer gaming consoles, playing a central role in the system architecture of the Saturn and Dreamcast generations, and led the concrete implementation of Japan-U.S. technical collaborations.
 During his approximately 20-year residency in Silicon Valley, he participated in projects for leading US Automotive OEM, a router manufacturer, a smart phone, and automotive manufacturer. He drove system designs integrating security, graphics, and QNX, and as an original development member of automotive Google navigation system, he realized a vehicular information foundation by merging NVIDIA graphics with the SH4A platform.
 Currently, at the RISC-V Association, he focuses on conceptualizing use cases for cutting-edge semiconductors while working toward the reconstruction of Japan’s semiconductor technology foundation.

10:10-10:40

SiFive 2nd Generation Intelligence Products Technology Explainer 
 Dive into the technical details of the cutting-edge world of RISC-V and AI with SiFive Senior Principal Architect John Simpson. In this technology explainer video, he breaks down some of the key innovations in our new 2nd Generation Intelligence™ products, to explain how they are optimized for AI workloads from edge devices to the cloud.

John Simpson | Senior Principal Architect | SiFive Inc. 
John Simpson is a Senior Principal Architect at SiFive, working on the definition of RISC-V vector AI extensions and performance event sampling. He is responsible for the 2nd Generation of SiFive Intelligence products and has previously presented at the RISC-V North America Summit. John is a seasoned semiconductor architect with three decades of industry experience.
 

10:40-11:10

Invited Talk: Current Status of the “RISC-V Server SoC Specification”
 In recent years, data center and server systems have been required to meet a wide range of platform-level demands, including Reliability, Availability, Serviceability (RAS), security, guaranteed performance, and long-term maintainability.

 This talk provides an overview of the concept of a standardized server platform foundation that enables RISC-V to expand into the data center–scale server market, and explains its technical significance.
 The primary goal of the RISC-V Server SoC Specification is to allow OS and hypervisor vendors to boot and run a single binary OS image on RISC-V servers without modification. In other words, the specification seeks to standardize hardware interfaces and functional requirements in areas where differentiation or vendor-specific implementation should not be necessary, thereby improving software compatibility and portability across RISC-V server platforms.
 This specification is positioned as one component of a broader server platform standard. When combined with essential elements such as UEFI/ACPI-based boot and runtime services, security models, platform firmware, and management mechanisms including BMC, it forms the basis of a complete RISC-V server platform standard.
 The presentation will cover: (a) Clock, timer, and interrupt control, (2) DMA protection and virtualization support through the RISC-V IOMMU, (3) PCIe subsystem integration rules, (4) Standardized RAS error record formats, (5) Mitigation of noisy-neighbor effects through QoS guarantees, (6) Industry-standard APIs for remote AI server management such as Redfish, (7) Common server management specifications including PLDM (Platform Level Data Model), (8) Management communication protocols such as MCTP (Management Component Transport Protocol). Through these topics, the talk highlights the essential building blocks required for robust and interoperable RISC-V server infrastructure. (Title and abstract prepared by the organizers.)

Vedvyas Shanbhogue | Semiconductor Engineer | Meta (formerly Facebook) | Vice Chair, RISC-V ISA Security Horizontal Technical Committee | Chair, Confidential Computing SIG/TG (RISC-V International)
 Vedvyas (“Ved”) Shanbhogue is a semiconductor architect at Meta, where he develops compute platforms for next-generation data center infrastructure. His work spans critical domains for cloud and AI systems, including security, virtualization, power management, and new ISA extensions across both RISC-V and x86 architectures. Ved has deep expertise in hardware security technologies that enable Confidential Computing, as well as in the integration of virtualization stacks such as KVM and Xen with key platform mechanisms including PCIe Integrity and Data Encryption (IDE), IOMMU, and Intel virtualization technologies (VT-x/VT-d).
 Beginning in 2021, Ved served as a Member of Technical Staff at the RISC-V startup Rivos Inc., contributing to the development of next-generation CPUs and AI data center system foundations. Following Meta’s acquisition of Rivos in 2025, he joined Meta’s data center silicon organization. Ved spent 21 years at Intel, ultimately as a Senior Principal Engineer, where he led the design of advanced processor security mechanisms that strengthen trust and isolation, including Control-flow Enforcement Technology (CCE/CET) and Trust Domain Extensions (TDX). He also contributed to sophisticated virtualization enhancements such as VMFUNC, EPT-switching, #VE, and HLAT, as well as microarchitectural design incorporating functional safety considerations.
 Beyond hardware architecture, Ved has extensive experience in system software, including Linux and Windows driver development, microcode design for network processors, high-availability middleware for telecom infrastructure, and protocol implementations such as SS7 and GSM.
 With a full-stack perspective spanning CPU design, OS and virtualization foundations, networking, and media processing technologies (MPEG-TS, RTP, H.264), Ved is a technical leader driving the evolution of next-generation computing platforms.

11:10-11:40

Invited Talk: Trusted Foundations for the AI and Cloud Era — Secure Large-Scale Data Analytics with RISC-V Confidential Computing
Confidential Computing is not merely an additional security feature; it is becoming a fundamental prerequisite for entry into the next-generation data center market. By providing mechanisms for protecting data in use and enabling isolated execution environments, Confidential Computing can be realized with properly designed hardware support at an overhead of roughly 10% in computational cost.
 This talk explains the comprehensive development of Confidential Computing capabilities as a core element of the next-generation RISC-V server platform for AI cloud infrastructure. In addition to foundational technologies such as the privileged ISA, the Hypervisor Extension, the Advanced Interrupt Architecture (AIA), and the IOMMU, RISC-V is actively advancing key isolation and compartmentalization features, including the Supervisor Domains Extension and the definition of the CoVE (Confidential VM Extension) Application Binary Interface (ABI). Together, these efforts enable the realization of Trusted Execution Environment Virtual Machines (TVMs) that are isolated from both the host OS and the hypervisor.
 Furthermore, the talk introduces critical ecosystem technologies required for secure accelerator-based AI computing. These include PCIe Integrity and Data Encryption (IDE), which protects CPU–device communications (GPU/NIC/SSD) from eavesdropping and tampering, as well as protocols for mutual authentication between the host CPU and devices such as SPDM (Security Protocol and Data Model). It also covers PCIe-standard mechanisms for assigning devices securely to Confidential VMs, including TDISP (TEE Device Interface Security Protocol).
 By combining these technologies, RISC-V Confidential Computing enables the secure direct assignment of AI accelerators, remote attestation rooted in hardware Root-of-Trust, and improved interoperability through open security silicon—establishing a robust trusted foundation for large-scale AI and cloud data analytics.

Ravi Sahita | Security Engineer | Meta (formerly Facebook) | Vice Chair, RISC-V International Security Horizontal Technical Committee | Chair, Confidential Computing SIG/TG (RISC-V International)  
 Ravi Sahita is a technical leader in computer security, instruction set architectures (ISAs), open-source system software, virtualization, platform design, and distributed systems. He has extensive experience across the full lifecycle of secure system development—from research and architecture through implementation and product deployment.
 Ravi has contributed to the development of industry-standard specifications in areas such as software ABIs, security models, network security, and Quality of Service (QoS). He also has hands-on expertise with FIPS (NIST) security certification processes and has been involved in open-source virtualization implementations including Xen and KVM/Linux. He is an author of numerous technical papers and industry standards within organizations such as RISC-V International, the IETF, and the Trusted Computing Group (TCG).
 Ravi possesses deep expertise in platform security architecture spanning software, processors, chipsets, and devices. He has led data center–scale Confidential Computing initiatives, including major projects such as RISC-V CoVE, OpenTitan, and Intel TDX.
 His contributions include advanced security innovations such as ROP defense through Intel CET, defining VM introspection ISA mechanisms within Intel VT-x, implementing the first security hypervisor designed to protect guest runtime integrity (Deepsafe), and developing system defense capabilities for Intel vPro. In recent years, Ravi has also focused on protecting privacy-sensitive AI training and inference infrastructure, as well as countermeasures against adversarial machine learning. He is the holder of more than 240 patents.

11:40-12:30

TEE (Trusted Execution Environment) + Attestation + RoT (Root of Trust) on RISC-V
 Trusted Execution Environments (TEEs), which provide isolated execution environments separate from the normal operating system, are widely utilized across cloud platforms (Intel SGX, TDX, AMD SEV-SNP), smartphones (Arm Cortex-A TrustZone), and embedded systems (Arm Cortex-M TrustZone). Numerous experimental implementations have also been proposed for RISC-V.
 In recent years, the application of TEE concepts has expanded to include GPUs, and further developments are anticipated.
 This lecture will introduce these technologies, along with the underlying mechanisms that support TEEs, such as Root of Trust (RoT) and attestation. It will also explain the essential security requirements necessary to build trusted systems.

Kuniyasu Suzaki | Professor, Institute of Information Security
 After serving at the National Institute of Advanced Industrial Science and Technology (AIST), he was appointed Professor at the Institute of Information Security in 2022.
 He has been recognized as an Invited Expert of the Trusted Computing Group (TCG) since 2019, and has served as an Advisor to the Privacy Tech Association since 2024.

 

 

Speakers are listed without honorifics.

Time
Photo

Presentation Title
Abstract

Speaker
Bio

Main Hall
9:00-13:00

Turing Machine ASIC “Hands-On” TinyTapeout Tutorial — From RTL to GDSII and Fabrication
This tutorial provides a fully hands-on introduction to ASIC design by implementing a Turing Machine using open-source EDA tools and an educational fabrication framework. The tutorial uses OpenROAD/OpenLane, SkyWater’s SKY130 open PDK, and the TinyTapeout platform, enabling attendees to experience how modern open-source tools can be used to design real chips. Throughout the session, participants will interactively learn each stage of the flow—design automation, physical implementation, and verification—while building an appreciation of Turing’s computational model through silicon realization. The goal of this 3-hour hands-on session is to give participants practical exposure to open-source ASIC development, deepen their understanding of computer architecture through the Turing Machine example, and guide them toward submitting a complete GDSII design via TinyTapeout’s fabrication shuttle. Participants must bring a personal Windows laptop with 50GB of free space (WSL2 + Ubuntu 24.04 recommended) or a Mac with Docker pre-installed. Tools will be installed on-site using a prepared Docker image.

Munetomo Maruyama
Click for Preview Material
Munetomo Maruyama is a veteran semiconductor design engineer with decades of experience in microcontroller (MCU) and SoC development in Japan. He began his career at Hitachi in 1986, contributing to the development of the SH-family microcontrollers, and later worked across multiple major semiconductor companies and ventures, where he led projects involving imaging SoCs, embedded MCUs, and advanced semiconductor device architectures. Maruyama is also widely known for developing and publishing the open-source mmRISC series of RISC-V CPU cores, contributing significantly to the dissemination of open hardware design methodologies in Japan. He has authored numerous technical articles and educational materials and is recognized for his clear, practical teaching style. With experience dating back to 1978 in microcomputers, FPGA, and computer architecture, Maruyama expertise: (1) MCU / SoC Architecture & Design (2) Open-source CPU Development (RISC-V mmRISC series) (3) ASIC/FPGA Design & Verification (4) Hands-on Engineering Education. 

 

A physical model of a Turing machine displayed at the “Go Ask Alice” exhibit of the Harvard Collection of Historical Scientific Instruments.
The device visually demonstrates how an abstract Turing machine operates by manipulating symbols on a tape.
Photograph by GabrielF/Wikimedia Commons(CC BY-SA 3.0)


Alan Turing (1912-1954), 1951. Photograph by Elliott & Fry. Courtesy of Wikimedia Commons. Public Domain.

 

About RISC-V Association

RISC-V-logo-figonly-mod-2

RISC-V Association is an association among RISC-V Foundation member companies. RISC-V Foundation defines Member Sponsored Events and Meetups as regional efforts and RISC-V Association supports these events. Presently Software Hardware Consulting handles business ends of this activity.

Recent Posts

  • The frontier of data center architecture, live in Tokyo.
  • RISC-V Day Tokyo 2025 Autumn Tutorial “Building a Device ‘ID Card’—From Power-On to Tamper Detection and Remote Proof (Tentative Title)” :RISC-V Day Tokyo 2025 Autumn • Instructor: H. Andrés Lagar-Cavilla, Principal Engineer, Google)
  • NVIDIA adds support for the CUDA platform on RISC-V in addition to x86 and Arm for its AI platform.
  • PBS WATCH: Trump hosts top tech CEOs, not including Elon Musk, at White House dinner
  • The design for the RISC-V 15th anniversary mug has been completed.
  • Announcement of “RISC-V DAY TOKYO 2025 AUTUMN”
  • In May 2025, the Trump administration officially announced the repeal of the “AI Diffusion Rule”.
  • Codasip selected to design a high-end RISC-V processor for the EU-funded DARE project
  • February 27, 2025 Thursday RISC-V DAY TOKYO 2025 SPRING Presentation 4:Kubuds(China)
  • 2025年2月27日(木)RISC-V DAY TOKYO 2025 SPRING 出展者情報2:アンデス・テクノロジーズ(台湾)

Archives

  • February 2026 (1)
  • September 2025 (4)
  • July 2025 (1)
  • May 2025 (1)
  • March 2025 (1)
  • February 2025 (3)
  • December 2024 (1)
  • November 2024 (3)
  • May 2024 (1)
  • January 2024 (6)
  • December 2023 (8)
  • November 2023 (2)
  • October 2023 (1)
  • June 2023 (4)
  • May 2023 (3)
  • April 2023 (2)
  • March 2023 (1)
  • January 2023 (1)
  • November 2022 (1)
  • October 2022 (2)
  • May 2022 (2)
  • April 2022 (2)
  • January 2022 (1)
  • December 2021 (3)
  • November 2021 (2)
  • June 2021 (1)
  • April 2021 (5)
  • February 2021 (2)
  • October 2020 (2)
  • September 2020 (1)
  • August 2020 (2)
  • July 2020 (1)
  • May 2020 (1)
  • April 2020 (1)
  • November 2019 (3)
  • October 2019 (15)
  • September 2019 (5)
  • August 2019 (1)
  • June 2019 (1)
  • 日本語日本語
  • Events
    • 2026 Spring
    • 2025 Autumn
    • 2024 Summer
    • 2024 Winter
    • 2023 Summer
    • 2022 Autumn
    • 2022 Spring
    • 2022 Vietnam
    • 2021 Autumn
    • 2021 Spring
    • 2020
    • 2020 Vietnam
    • 2019 Tokyo
  • RISC-V International
  • Facebook
  • Twitter
Privacy Policy
© 2021-2026 RISC-V Alliance Japan