Pouya Fotouhi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Pouya Fotouhi has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
S. J. Ben Yoo
University of California
6 shared publications- Roberto Proietti
Politecnico di Torino
5 shared publications Roberto Proietti
Department of Electrical and Computer Engineering, University of California, Davis, CA, USA
4 shared publicationsSebastian Werner
University of California
4 shared publications- Sung Joo Ben Yoo
University of California, Davis
4 shared publications - Marjan Fariborz
University of California Davis
3 shared publications
Latest Publications
- 29 May 2022
- Published by Springer Nature
- p. 44-64
LLM: Realizing Low-Latency Memory by Exploiting Embedded Silicon Photonics for Irregular Workloads
- 29 May 2022
- Published by Springer Nature
- p. 44-64
LLM: Realizing Low-Latency Memory by Exploiting Embedded Silicon Photonics for Irregular Workloads
- 17 June 2021
- Published by Springer Nature
- p. 176-194
HTA: A Scalable High-Throughput Accelerator for Irregular HPC Workloads
- 17 June 2021
- Published by Springer Nature
- p. 176-194
HTA: A Scalable High-Throughput Accelerator for Irregular HPC Workloads
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Quantum Electronics
- Vol. 26 (2) , 1-10
Silicon Photonic Flex-LIONS for Bandwidth-Reconfigurable Optical Interconnects
This paper reports the first experimental demonstration of silicon photonic (SiPh) Flex-LIONS, a bandwidth-reconfigurable SiPh switching fabric based on wavelength routing in arrayed waveguide grating routers (AWGRs) and space switching. Compared with the state-of-the-art bandwidth-reconfigurable switching fabrics, Flex-LIONS architecture exhibits 21× less number of switching elements and 2.9× lower on-chip loss for 64 ports, which indicates significant improvements in scalability and energy efficiency. System experimental results carried out with an 8-port SiPh Flex-LIONS prototype demonstrate error-free one-to-eight multicast interconnection at 25 Gb/s and bandwidth reconfiguration from 25 Gb/s to 100 Gb/s between selected input and output ports. Besides, benchmarking simulation results show that Flex-LIONS can provide a 1.33× reduction in packet latency and >1.5× improvements in energy efficiency when replacing the core layer switches of Fat-Tree topologies with Flex-LIONS. Finally, we discuss the possibility of scaling Flex-LIONS up to N = 1024 ports (N = M × W) by arranging M 2 W-port Flex-LIONS in a Thin-CLOS architecture using W wavelengths.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Quantum Electronics
- Vol. 26 (2) , 1-10
Silicon Photonic Flex-LIONS for Bandwidth-Reconfigurable Optical Interconnects
This paper reports the first experimental demonstration of silicon photonic (SiPh) Flex-LIONS, a bandwidth-reconfigurable SiPh switching fabric based on wavelength routing in arrayed waveguide grating routers (AWGRs) and space switching. Compared with the state-of-the-art bandwidth-reconfigurable switching fabrics, Flex-LIONS architecture exhibits 21× less number of switching elements and 2.9× lower on-chip loss for 64 ports, which indicates significant improvements in scalability and energy efficiency. System experimental results carried out with an 8-port SiPh Flex-LIONS prototype demonstrate error-free one-to-eight multicast interconnection at 25 Gb/s and bandwidth reconfiguration from 25 Gb/s to 100 Gb/s between selected input and output ports. Besides, benchmarking simulation results show that Flex-LIONS can provide a 1.33× reduction in packet latency and >1.5× improvements in energy efficiency when replacing the core layer switches of Fat-Tree topologies with Flex-LIONS. Finally, we discuss the possibility of scaling Flex-LIONS up to N = 1024 ports (N = M × W) by arranging M 2 W-port Flex-LIONS in a Thin-CLOS architecture using W wavelengths.
- Open Access
- 30 September 2019
- Published by Association for Computing Machinery (ACM)
- p. 206-221
3D photonics as enabling technology for deep 3D DRAM stacking
3D stacking improves bandwidth, energy, and latency of DRAMs by exploiting shorter and more abundant wiring in three dimensions. While future stacks are predicted to provide tens of DRAM layers, TSV pitches are bound to stop decreasing due to physical limitations, requiring either large area overheads or energy for higher pin data rates to preserve bandwidth scaling. In addition, deep 3D DRAM stacks increase the average number of TSVs hops to reach the appropriate DRAM layer. Recent advances in vertical silicon-photonic interconnects now allow diameters of 1-2μm which results in high bandwidth density for optics. In this paper, we explore the benefits and architectural implications of vertical optical interconnects in terms of area, power, and performance. We propose a hierarchical approach to stacking 3D DRAM to tens of layers by utilizing sub-stacks which are optically-interconnected to a memory interface on the processor die. Our results show that photonics could be a key enabler for deep-3D DRAM offering at least 2× interconnect area savings compared to TSVs for the same bandwidth with comparable performance and less power.
- 30 September 2019
- Published by Association for Computing Machinery (ACM)
- p. 206-221
Open Access3D photonics as enabling technology for deep 3D DRAM stacking
3D stacking improves bandwidth, energy, and latency of DRAMs by exploiting shorter and more abundant wiring in three dimensions. While future stacks are predicted to provide tens of DRAM layers, TSV pitches are bound to stop decreasing due to physical limitations, requiring either large area overheads or energy for higher pin data rates to preserve bandwidth scaling. In addition, deep 3D DRAM stacks increase the average number of TSVs hops to reach the appropriate DRAM layer. Recent advances in vertical silicon-photonic interconnects now allow diameters of 1-2μm which results in high bandwidth density for optics. In this paper, we explore the benefits and architectural implications of vertical optical interconnects in terms of area, power, and performance. We propose a hierarchical approach to stacking 3D DRAM to tens of layers by utilizing sub-stacks which are optically-interconnected to a memory interface on the processor die. Our results show that photonics could be a key enabler for deep-3D DRAM offering at least 2× interconnect area savings compared to TSVs for the same bandwidth with comparable performance and less power.
- Open Access
- 30 September 2019
- Published by Association for Computing Machinery (ACM)
- p. 222-334
Enabling scalable chiplet-based uniform memory architectures with silicon photonics
Chiplet-based systems have recently received much attention for scaling-up processing power in HPC systems due to their high energy efficiency and low cost manufacturing; however, large inter-chiplet NUMA latencies, distance-related energy overheads, and limited IO bandwidth caused by state-of-the-art packaging and interconnect technologies substantially limit their scalability. Large last level caches (up to 16MiB/chiplet and 40% of chiplet area) of current systems can only temporarily hide these limitations and come at the large cost and leakage power of SRAM cells. In this paper, we propose the use of integrated silicon-photonic (SiPh) interconnects on an organic package substrate which combines low material costs with a high IO bandwidth, distance-independent energy consumption, and low-latency point-to-point interconnection fabric to effectively overcome current interconnect and packaging limitations. We exploit the properties of this fabric to propose a scalable uniform memory architecture (S-UMA) that overcomes all NUMA-related performance challenges. Moreover, we propose exploiting our low-latency SiPh fabric to remove the large LLC caches from the processor chiplets and re-integrate them into separate chiplets, increasing manufacturing yield by using smaller chiplets, allowing to use the most efficient process for SRAM circuits, or easing integration of alternative memory technologies without performance hits. Compared to state-of-the-art architectures, S-UMA offers 23% performance speed-up and 30% network power savings on average across HPC workloads for a 8-chiplet 64-core system.
- 30 September 2019
- Published by Association for Computing Machinery (ACM)
- p. 222-334
Open AccessEnabling scalable chiplet-based uniform memory architectures with silicon photonics
Chiplet-based systems have recently received much attention for scaling-up processing power in HPC systems due to their high energy efficiency and low cost manufacturing; however, large inter-chiplet NUMA latencies, distance-related energy overheads, and limited IO bandwidth caused by state-of-the-art packaging and interconnect technologies substantially limit their scalability. Large last level caches (up to 16MiB/chiplet and 40% of chiplet area) of current systems can only temporarily hide these limitations and come at the large cost and leakage power of SRAM cells. In this paper, we propose the use of integrated silicon-photonic (SiPh) interconnects on an organic package substrate which combines low material costs with a high IO bandwidth, distance-independent energy consumption, and low-latency point-to-point interconnection fabric to effectively overcome current interconnect and packaging limitations. We exploit the properties of this fabric to propose a scalable uniform memory architecture (S-UMA) that overcomes all NUMA-related performance challenges. Moreover, we propose exploiting our low-latency SiPh fabric to remove the large LLC caches from the processor chiplets and re-integrate them into separate chiplets, increasing manufacturing yield by using smaller chiplets, allowing to use the most efficient process for SRAM circuits, or easing integration of alternative memory technologies without performance hits. Compared to state-of-the-art architectures, S-UMA offers 23% performance speed-up and 30% network power savings on average across HPC workloads for a 8-chiplet 64-core system.
- 13 June 2019
- Published by Optica Publishing Group in Journal of Optical Communications and Networking
- Vol. 11 (7) , 333-346
Enabling Scalable Disintegrated Computing Systems With AWGR-Based 25D Interconnection Networks
2.5D integrated systems exploiting electronic interposers to tightly integrate multiple processor dies into the same package suffer from significant performance degradation caused by the large latency overheads of their die-to-die multihop electrical interconnection networks. Silicon-photonic interposers with wavelength-routed interconnects can overcome this issue by enabling directly connected, scalable topologies while exhibiting low-energy optical communication even at large distances. This paper studies the use of an arrayed waveguide grating router (AWGR) as a scalable, low-latency silicon-photonic interconnection fabric for computing systems with up to 256 cores. Our results indicate that AWGRs could be a key enabler for large-scale interposer systems, offering an average performance speed-up of at least 1.25× with 1.32× lower power for 256 cores compared to state-of-the-art electrical networks, while offering a more compact solution compared to alternative photonic interconnects.
- 13 June 2019
- Published by Optica Publishing Group in Journal of Optical Communications and Networking
- Vol. 11 (7) , 333-346
Enabling Scalable Disintegrated Computing Systems With AWGR-Based 25D Interconnection Networks
2.5D integrated systems exploiting electronic interposers to tightly integrate multiple processor dies into the same package suffer from significant performance degradation caused by the large latency overheads of their die-to-die multihop electrical interconnection networks. Silicon-photonic interposers with wavelength-routed interconnects can overcome this issue by enabling directly connected, scalable topologies while exhibiting low-energy optical communication even at large distances. This paper studies the use of an arrayed waveguide grating router (AWGR) as a scalable, low-latency silicon-photonic interconnection fabric for computing systems with up to 256 cores. Our results indicate that AWGRs could be a key enabler for large-scale interposer systems, offering an average performance speed-up of at least 1.25× with 1.32× lower power for 256 cores compared to state-of-the-art electrical networks, while offering a more compact solution compared to alternative photonic interconnects.
- 1 January 2019
- Published by Optica Publishing Group
- p. SM3G.2
FlexLION: A Reconfigurable All-to-All Optical Interconnect Fabric with Bandwidth Steering
We demonstrate an all-to-all interconnect fabric with bandwidth steering by wavelength routing, wavelength add/drop filtering and spatial switching. We report a proof-of-concept experiment and demonstrate scalability up to 32 ports with aggregated bandwidth of 51.2 Tb/s under worst-case crosstalk.
- 1 January 2019
- Published by Optica Publishing Group
- p. SM3G.2
FlexLION: A Reconfigurable All-to-All Optical Interconnect Fabric with Bandwidth Steering
We demonstrate an all-to-all interconnect fabric with bandwidth steering by wavelength routing, wavelength add/drop filtering and spatial switching. We report a proof-of-concept experiment and demonstrate scalability up to 32 ports with aggregated bandwidth of 51.2 Tb/s under worst-case crosstalk.
- Open Access
- 1 October 2018
- Published by Association for Computing Machinery (ACM)
- p. 269-278
AWGR-based optical processor-to-memory communication for low-latency, low-energy vault accesses
- 1 October 2018
- Published by Association for Computing Machinery (ACM)
- p. 269-278
Open AccessAWGR-based optical processor-to-memory communication for low-latency, low-energy vault accesses
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Towards Energy-Efficient High-Throughput Photonic NoCs for 2.5D Integrated Systems: A Case for AWGRs
Silicon Photonics (SiPs) can overcome the energy and bandwidth limitations of electrical interconnects in networks-on-chip (NoCs) and enable efficient global all-to-all connectivity-the ideal from a performance perspective. Unfortunately, state-of-the-art SiP switching fabrics impose power overheads for thermo-optical control of microring resonators (MRs), excessive crosstalk, or challenging physical layout. The Arrayed Waveguide Grating Router (AWGR) is a SiP device that provides layout-efficient and scalable all-to-all connectivity through wavelength-routing on a passive and compact platform. Recent technological advances now enable AWGR integration with significantly reduced footprint (<;1mm 2 ), crosstalk (<;-38dB), and loss (<;2dB), making AWGRs emerge as a major enabler for energy-efficient all-to-all connectivity in NoCs. This paper, for the first time, populates the design space of AWGR-based NoC topologies and compares AWGRs to state-of-the-art SiP fabrics and aggressive electrical baselines. Our results make a compelling case for AWGRs in interposer-based systems with processor disintegration, which have high bandwidth demands, large on-chip distances, and less stringent area constraints. AWGR-based NoCs can offer an average of 1.67× power savings, 1.23× speed-up, and 1.5× lower energy-delay-product (EDP) on PARSEC3.0/SPLASH-2× workloads compared to electrical baselines, and fewer waveguides, wavelengths, or MRs than alternative SiP crossbar fabrics.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Towards Energy-Efficient High-Throughput Photonic NoCs for 2.5D Integrated Systems: A Case for AWGRs
Silicon Photonics (SiPs) can overcome the energy and bandwidth limitations of electrical interconnects in networks-on-chip (NoCs) and enable efficient global all-to-all connectivity-the ideal from a performance perspective. Unfortunately, state-of-the-art SiP switching fabrics impose power overheads for thermo-optical control of microring resonators (MRs), excessive crosstalk, or challenging physical layout. The Arrayed Waveguide Grating Router (AWGR) is a SiP device that provides layout-efficient and scalable all-to-all connectivity through wavelength-routing on a passive and compact platform. Recent technological advances now enable AWGR integration with significantly reduced footprint (<;1mm 2 ), crosstalk (<;-38dB), and loss (<;2dB), making AWGRs emerge as a major enabler for energy-efficient all-to-all connectivity in NoCs. This paper, for the first time, populates the design space of AWGR-based NoC topologies and compares AWGRs to state-of-the-art SiP fabrics and aggressive electrical baselines. Our results make a compelling case for AWGRs in interposer-based systems with processor disintegration, which have high bandwidth demands, large on-chip distances, and less stringent area constraints. AWGR-based NoCs can offer an average of 1.67× power savings, 1.23× speed-up, and 1.5× lower energy-delay-product (EDP) on PARSEC3.0/SPLASH-2× workloads compared to electrical baselines, and fewer waveguides, wavelengths, or MRs than alternative SiP crossbar fabrics.