EPAC: The Last Dance
Paper in proceeding, 2026

This paper presents EPAC, a RISC-V-based accelerator chip developed within the European Processor Initiative (EPI) as part of a multi-year, multi-partner effort to build a European HPC processor ecosystem. EPAC is implemented in GlobalFoundries 22FDX (GF22FDX) technology, covers an area of 27 mm2 with approximately 0.3 billion transistors, and integrates three distinct RISC-V compute tiles targeting different workload classes: VEC, a vector processing tile for double-precision HPC workloads; STX, a many-core tile optimized for stencil and machine learning computations; and VRP, a variable-precision tile for iterative numerical solvers requiring extended floating-point formats. All tiles are connected through a Coherent Hub Interface (CHI) based network-on-chip with a distributed L2 cache system and communicate with external memory via a SerDes link. The chip was taped out in GF22FDX technology and successfully brought up, with all major IP blocks validated. This paper describes the architecture of each tile and the uncore infrastructure, the integration and physical implementation process, and the board-level bring-up activities. It also reflects on the engineering and coordination lessons learned from a full chip design effort distributed across academic and industrial partners in Europe.

Stencil

Variable Precision

EPAC

RISC-V

European Processor Initiative

Chip Tapeout

Vector Computing

Author

Filippo Mantovani

Centro Nacional de Supercomputacion

Fabio Banchelli

Centro Nacional de Supercomputacion

Pablo Vizcaino

Centro Nacional de Supercomputacion

Roger Ferrer

Centro Nacional de Supercomputacion

Oscar Palomar

Centro Nacional de Supercomputacion

Francesco Minervini

Centro Nacional de Supercomputacion

Jesus Labarta

Centro Nacional de Supercomputacion

Mauro Olivieri

Centro Nacional de Supercomputacion

Sapienza University of Rome

Sebastiano Pomata

SemiDynamics

Pedro Marcuello

SemiDynamics

Jordi Cortina

SemiDynamics

Alberto Moreno

SemiDynamics

Josep Sans

SemiDynamics

Roger Espasa

SemiDynamics

Vasileios Papaefstathiou

Foundation for Research and Technology Hellas (FORTH)

Nikolaos Dimou

Foundation for Research and Technology Hellas (FORTH)

Georgios Ieronymakis

Foundation for Research and Technology Hellas (FORTH)

A. Psathakis

Foundation for Research and Technology Hellas (FORTH)

Michalis Giaourtas

Foundation for Research and Technology Hellas (FORTH)

Iasonas Mastorakis

Foundation for Research and Technology Hellas (FORTH)

Manolis Marazakis

Foundation for Research and Technology Hellas (FORTH)

Eric Guthmuller

Grenoble Alpes University

Andrea Bocco

Grenoble Alpes University

Jérôme Fereyre

Grenoble Alpes University

César Fuguet

Grenoble Alpes University

Mate Kovač

University of Zagreb

Mario Kovač

University of Zagreb

Luka Mrković

University of Zagreb

Josip Ramljak

University of Zagreb

Luca Bertaccini

Swiss Federal Institute of Technology in Zürich (ETH)

Tim Fischer

Swiss Federal Institute of Technology in Zürich (ETH)

Frank K. Gurkaynak

Swiss Federal Institute of Technology in Zürich (ETH)

Paul Scheffler

Swiss Federal Institute of Technology in Zürich (ETH)

Luca Benini

Swiss Federal Institute of Technology in Zürich (ETH)

Bhavishya Goel

Chalmers, Computer Science and Engineering (Chalmers), Computer Engineering (Chalmers)

Madhavan Manivannan

Chalmers, Computer Science and Engineering (Chalmers), Computer Engineering (Chalmers)

Tiago Rocha

University of Lisbon

Nuno Neves

University of Lisbon

Jens Krüger

Fraunhofer Society

Proceedings of the 23rd ACM International Conference on Computing Frontiers 2026 Workshops and Special Sessions Cf 2026 Companion

155-162
9798400725692 (ISBN)

23rd ACM International Conference on Computing Frontiers 2026 Workshops and Special Sessions, CF 2026 Companion
Catania, Italy,

Subject Categories (SSIF 2025)

Computer Systems

DOI

10.1145/3801488.3806242

More information

Latest update

7/29/2026