EmerGing trEnds iN qUantum computing, crYptography aNd arithmEtic (GENUYNE)

Principal Investigators: José Luis Imaña and Carlos González Calvo

Start and end dates: 01/09/2025-31/08/2028

Code: PID2024-158203OB-I00

Summary: In the era of Artificial Intelligence (AI), computing systems face a very different landscape. Whereas low power and high performance have been the main requirements of a computing system, the growth of learning-based applications and large volumes of data pose unprecedented challenges that cannot be addressed by traditional methodologies. In addition, the increase in computing performance is limited by energy dissipation. All this, together with the fact that with the end of Dennards law reducing the power consumption of integrated circuits and improving performance through technological miniaturization poses considerable challenges, makes innovative approaches to computing necessary. Different emerging computing paradigms with the potential to significantly accelerate performance and increase efficiency, while reducing cost and power consumption have been proposed. Some of these next-generation computing paradigms will be studied in the GENUYNE project.

One of the most important emerging paradigms is quantum computing (QC). Quantum computers introduce a novel mode of computation having an enormous computational power. Significant trends in QC are the development of quantum algorithms and the integration of QC with IA. QC is not only developed on quantum computers, so the establishment of simulation techniques on classical circuits requires the use of High-Performance Computing (HPC) systems. One possibility for simulating quantum circuits in classical devices is the combination of classical simulation with hybrid quantum computers, existing different ways of integrating both technologies. FPGAs are ideal candidates for integration into HPC systems, helping as accelerators due to their high level of parallelization.

The area where quantum computers poses a serious threat is in cybersecurity. Cryptographic keys or data encrypted might be compromised whenever quantum computers become a reality. For this, Post-Quantum Cryptography (PQC) has become a trending research topic due to its resistance against quantum computers. Some PQC algorithms have recently been standardized while other schemes are still under evaluation. Therefore, significant efforts must be made in applied engineering and research to efficiently implement these new cryptographic primitives.

Another emerging paradigm in computation is the Next Generation Arithmetic (NGA) that is linked to the use of non-standard numerical formats tailored to specific problems. To accelerate the deployment of novel arithmetic approaches it is necessary to provide system prototypes. Hence, it is essential to automate the process via code-generators, High-Level Synthesis and Electronic Design Automation tools. The appearance of the RISC-V open-source ISA is critical to construct processors and System-on-Chips proving the advances with respect to the state of the art.

In GENUYNE project, the new emerging paradigms of QC and NGA will be explored. For QC, we will design faster quantum algorithms for different technologies of quantum computers and the integration of FPGAs as quantum accelerators will be developed. Furthermore, we will design cryptographic accelerators for the PQC algorithms recently standardized and for the upcoming schemes under evaluation. For NGA, we will design new specialized hardware units for emerging numerical formats and provide open-source tools in combination with RISC-V-based architectures.

Overall goal: To explore the new emerging paradigms of QC and NGA. For QC and for PQC arising from the threat of QC, to design faster quantum algorithms, quantum accelerators on FPGAs and cryptographic accelerators for the standardized and upcoming PQC schemes (General Objective-1). For NGA, to design new specialized hardware units for emerging numerical formats and provide open-source tools in combination with RISC-V-based architectures (General Objective-2).

The overall goal of the project has been divided into two General Objectives (GO). Each GO is further broken down into Specific Objectives (SO), which are defined below.

• General Objective-1 (GO1). Quantum computing and post-quantum cryptography.
    – SO1.1. Quantum computing with different technologies.  
    – SO1.2. Quantum accelerators on FPGAs.  
    – SO1.3. Acceleration of Post-quantum cryptography.  
• General Objective-2 (GO2). Classical fast code generation and deployment.
    – SO2.1. Next Generation Arithmetic and Approximate Computing.
    – SO2.2. Classical fast code generation of datapath at circuit level using HLS.
    – SO2.3. System prototyping.