Periodic Reporting for period 1 - WINC (Wireless Networks within Next-Generation Computing Systems)
Periodo di rendicontazione: 2022-10-01 al 2025-03-31
Computing systems are ubiquitous in our daily life and have transformed the way we learn, work, or communicate with each other, to the point that progress is intimately tied to the improvements brought by new generations of the processors that lie at the heart of these systems. A common trait of current computing systems is that their internal data communication has become a fundamental bottleneck. The anticipated death of Moore’s Law has forced computer scientists and architects to find new ways to build faster processors, which include massive parallelization, specialized accelerator design, and disruptive technologies such as quantum computing. These trends cause an exponential increase in the volume and variability of data transfers within computing systems, rendering traditional interconnects insufficient and threatening to halt progress unless fast and versatile communication alternatives are developed. In this light, the WINC project envisions a revolution in computer architecture enabled by the integration of wireless networks within computing systems. The main hypothesis is that wireless terahertz technology will lead to at least a tenfold improvement in the speed, efficiency, and scalability of both non-quantum and quantum systems. With a cross-cutting approach, WINC aims to validate the hypothesis by (i) revealing the fundamental limits of wireless communications within computing packages, (ii) developing antennas and protocols that operate close to those limits while complying with the stringent constraints of the scenario, and (iii) developing radically novel architectures that translate the unique benefits of the wireless vision into order-of-magnitude improvements at the system level. If successful, WINC will be the seed of a new generation of non-quantum and quantum systems and foster progress in the computing field for the decades to come, with huge and all-pervasive impact in fields such as AI, biology, medicine, or finance.
During the first half of the project, the WINC research team has worked in four different fronts, namely:
(i) Towards demonstrating the capacity of terahertz technology to provide the bandwidth, versatility, and compactness required at the computing package scale. In this respect, we have simulated wireless links within the complex structure of a computing package and observed several impairments that affect the speed and efficiency of the wireless communication. In response to these studying a technique called time reversal that allows us to combat the detrimental effects of the many reflections happening inside these enclosed environments.
(ii) Designing communication protocols capable of providing high speed and energy efficient operation with small area. In this case, we are currently developing area and power models of the circuits needed to implement the wireless links, which depend on the chosen modulation and other design factors. On top of it, we developed protocols to control the access to the shared wireless medium in this very particular scenario of communications.
(iii) Towards proving a significant improvement in the performance, efficiency, and scalability of non-quantum and quantum systems. On the classical side, we have initially demonstrated that we can speedup multi-chip accelerators specialized in AI computations by 10-20% with a single broadband wireless channel, and we are analyzing other scenarios such as general-purpose CPUs. On the quantum side, we are developing algorithms that map quantum algorithms within the specific type of quantum computers that we are developing in this project. Our current solution is three times faster and 4 times more effectively than existing techniques.
(iv) Developing a vertical simulation framework for the cross-layer design and assessment of wireless-enabled computing systems. In particular, we have successfully modeled wireless links (including its medium access control protocols) inside existing computer architecture simulators, which will be released as open source code, and will continue updating them as new models and protocols are developed in the project.
(i) Towards demonstrating the capacity of terahertz technology to provide the bandwidth, versatility, and compactness required at the computing package scale. In this respect, we have simulated wireless links within the complex structure of a computing package and observed several impairments that affect the speed and efficiency of the wireless communication. In response to these studying a technique called time reversal that allows us to combat the detrimental effects of the many reflections happening inside these enclosed environments.
(ii) Designing communication protocols capable of providing high speed and energy efficient operation with small area. In this case, we are currently developing area and power models of the circuits needed to implement the wireless links, which depend on the chosen modulation and other design factors. On top of it, we developed protocols to control the access to the shared wireless medium in this very particular scenario of communications.
(iii) Towards proving a significant improvement in the performance, efficiency, and scalability of non-quantum and quantum systems. On the classical side, we have initially demonstrated that we can speedup multi-chip accelerators specialized in AI computations by 10-20% with a single broadband wireless channel, and we are analyzing other scenarios such as general-purpose CPUs. On the quantum side, we are developing algorithms that map quantum algorithms within the specific type of quantum computers that we are developing in this project. Our current solution is three times faster and 4 times more effectively than existing techniques.
(iv) Developing a vertical simulation framework for the cross-layer design and assessment of wireless-enabled computing systems. In particular, we have successfully modeled wireless links (including its medium access control protocols) inside existing computer architecture simulators, which will be released as open source code, and will continue updating them as new models and protocols are developed in the project.
The project has so far led to significant results with large potential impact in the wireless communications, computer architecture, and quantum computing domains. In particular, we highlight:
(i) Time reversal within computing packages: our early explorations of the wireless channels within packages clearly indicated the difficulty of sustaining high data rates due to the reverberant nature of the scenario. In order to mitigate the inter-symbol interference without having to induce high losses, time reversal was explored. The results assuming an ideal time reversal filter suggest not only that speeds in excess of 50 Gb/s are possible, but also that multiple parallel channels could be created thanks to the spatiotemporal focusing achieved by the technique. Further research and demonstrations are required to confirm the potential of this technique, which could also have an impact if used within quantum computing.
(ii) Cryogenic antennas and wireless channels: making the case and assessing, for the first time, the potential capacity of wireless channels at cryogenic temperatures inside the cryocooler of a quantum computer, using cryogenic (possibly superconducting) wireless antennas is at reach for this project. This has a very large potential impact.
(iii) Mapping algorithms for quantum multicore architectures: our work in on mapping algorithms for multicore architectures has led to the proposal of techniques that significantly outperformed those of the state of the art, including the one used in the world-leading IBM Qiskit compiler. Hence, this result has the potential to be a breakthrough as the concept of multicore quantum computing gains popularity, and could be exploited in future’s quantum computing ecosystem.
(iv) Design space exploration of quantum systems: the unique combination of models and simulation tools being developed for cross-layer model-based exploration of quantum systems can be a crucial asset for exploration of future architectures, hence putting WINC in an advantageous position to achieve significant breakthroughs in the realm of quantum computer architecture. Further research should be done to materialize such a potential impact.
(i) Time reversal within computing packages: our early explorations of the wireless channels within packages clearly indicated the difficulty of sustaining high data rates due to the reverberant nature of the scenario. In order to mitigate the inter-symbol interference without having to induce high losses, time reversal was explored. The results assuming an ideal time reversal filter suggest not only that speeds in excess of 50 Gb/s are possible, but also that multiple parallel channels could be created thanks to the spatiotemporal focusing achieved by the technique. Further research and demonstrations are required to confirm the potential of this technique, which could also have an impact if used within quantum computing.
(ii) Cryogenic antennas and wireless channels: making the case and assessing, for the first time, the potential capacity of wireless channels at cryogenic temperatures inside the cryocooler of a quantum computer, using cryogenic (possibly superconducting) wireless antennas is at reach for this project. This has a very large potential impact.
(iii) Mapping algorithms for quantum multicore architectures: our work in on mapping algorithms for multicore architectures has led to the proposal of techniques that significantly outperformed those of the state of the art, including the one used in the world-leading IBM Qiskit compiler. Hence, this result has the potential to be a breakthrough as the concept of multicore quantum computing gains popularity, and could be exploited in future’s quantum computing ecosystem.
(iv) Design space exploration of quantum systems: the unique combination of models and simulation tools being developed for cross-layer model-based exploration of quantum systems can be a crucial asset for exploration of future architectures, hence putting WINC in an advantageous position to achieve significant breakthroughs in the realm of quantum computer architecture. Further research should be done to materialize such a potential impact.