-
Towards Generalized On-Chip Communication for Programmable Accelerators in Heterogeneous Architectures
Authors:
Joseph Zuckerman,
John-David Wellman,
Ajay Vanamali,
Manish Shankar,
Gabriele Tombesi,
Karthik Swaminathan,
Kevin Lee,
Mohit Kapur,
Robert Philhower,
Pradip Bose,
Luca P. Carloni
Abstract:
We present several enhancements to the open-source ESP platform to support flexible and efficient on-chip communication for programmable accelerators in heterogeneous SoCs. These enhancements include 1) a flexible point-to-point communication mechanism between accelerators, 2) a multicast NoC that supports data forwarding to multiple accelerators simultaneously, 3) accelerator synchronization leve…
▽ More
We present several enhancements to the open-source ESP platform to support flexible and efficient on-chip communication for programmable accelerators in heterogeneous SoCs. These enhancements include 1) a flexible point-to-point communication mechanism between accelerators, 2) a multicast NoC that supports data forwarding to multiple accelerators simultaneously, 3) accelerator synchronization leveraging the SoC's coherence protocol, 4) an accelerator interface that offers fine-grained control over the communication mode used, and 5) an example ISA extension to support our enhancements. Our solution adds negligible area to the SoC architecture and requires minimal changes to the accelerators themselves. We have validated most of these features in complex FPGA prototypes and plan to include them in the open-source release of ESP in the coming months.
△ Less
Submitted 4 July, 2024;
originally announced July 2024.
-
Anticipate & Collab: Data-driven Task Anticipation and Knowledge-driven Planning for Human-robot Collaboration
Authors:
Shivam Singh,
Karthik Swaminathan,
Raghav Arora,
Ramandeep Singh,
Ahana Datta,
Dipanjan Das,
Snehasis Banerjee,
Mohan Sridharan,
Madhava Krishna
Abstract:
An agent assisting humans in daily living activities can collaborate more effectively by anticipating upcoming tasks. Data-driven methods represent the state of the art in task anticipation, planning, and related problems, but these methods are resource-hungry and opaque. Our prior work introduced a proof of concept framework that used an LLM to anticipate 3 high-level tasks that served as goals f…
▽ More
An agent assisting humans in daily living activities can collaborate more effectively by anticipating upcoming tasks. Data-driven methods represent the state of the art in task anticipation, planning, and related problems, but these methods are resource-hungry and opaque. Our prior work introduced a proof of concept framework that used an LLM to anticipate 3 high-level tasks that served as goals for a classical planning system that computed a sequence of low-level actions for the agent to achieve these goals. This paper describes DaTAPlan, our framework that significantly extends our prior work toward human-robot collaboration. Specifically, DaTAPlan planner computes actions for an agent and a human to collaboratively and jointly achieve the tasks anticipated by the LLM, and the agent automatically adapts to unexpected changes in human action outcomes and preferences. We evaluate DaTAPlan capabilities in a realistic simulation environment, demonstrating accurate task anticipation, effective human-robot collaboration, and the ability to adapt to unexpected changes. Project website: https://dataplan-hrc.github.io
△ Less
Submitted 4 April, 2024;
originally announced April 2024.
-
BERRY: Bit Error Robustness for Energy-Efficient Reinforcement Learning-Based Autonomous Systems
Authors:
Zishen Wan,
Nandhini Chandramoorthy,
Karthik Swaminathan,
Pin-Yu Chen,
Vijay Janapa Reddi,
Arijit Raychowdhury
Abstract:
Autonomous systems, such as Unmanned Aerial Vehicles (UAVs), are expected to run complex reinforcement learning (RL) models to execute fully autonomous position-navigation-time tasks within stringent onboard weight and power constraints. We observe that reducing onboard operating voltage can benefit the energy efficiency of both the computation and flight mission, however, it can also result in on…
▽ More
Autonomous systems, such as Unmanned Aerial Vehicles (UAVs), are expected to run complex reinforcement learning (RL) models to execute fully autonomous position-navigation-time tasks within stringent onboard weight and power constraints. We observe that reducing onboard operating voltage can benefit the energy efficiency of both the computation and flight mission, however, it can also result in on-chip bit failures that are detrimental to mission safety and performance. To this end, we propose BERRY, a robust learning framework to improve bit error robustness and energy efficiency for RL-enabled autonomous systems. BERRY supports robust learning, both offline and on-board the UAV, and for the first time, demonstrates the practicality of robust low-voltage operation on UAVs that leads to high energy savings in both compute-level operation and system-level quality-of-flight. We perform extensive experiments on 72 autonomous navigation scenarios and demonstrate that BERRY generalizes well across environments, UAVs, autonomy policies, operating voltages and fault patterns, and consistently improves robustness, efficiency and mission performance, achieving up to 15.62% reduction in flight energy, 18.51% increase in the number of successful missions, and 3.43x processing energy reduction.
△ Less
Submitted 19 July, 2023;
originally announced July 2023.
-
Signatures of many-body localization of quasiparticles in a flat band superconductor
Authors:
Koushik Swaminathan,
Poula Tadros,
Sebastiano Peotta
Abstract:
We construct a class of exact eigenstates of the Hamiltonian obtained by projecting the Hubbard interaction term onto the flat band subspace of a generic lattice model. These exact eigenstates are many body states in which an arbitrary number of localized fermionic particles coexist with a sea of mobile Cooper pairs with zero momentum. By considering the dice lattice as an example, we provide evid…
▽ More
We construct a class of exact eigenstates of the Hamiltonian obtained by projecting the Hubbard interaction term onto the flat band subspace of a generic lattice model. These exact eigenstates are many body states in which an arbitrary number of localized fermionic particles coexist with a sea of mobile Cooper pairs with zero momentum. By considering the dice lattice as an example, we provide evidence that these exact eigenstates are in fact manifestation of local integrals of motions of the projected Hamiltonian. In particular the spin and particle densities retain memory of the initial state for a very long time, if localized unpaired particles are present at the beginning of the time evolution. This shows that many-body localization of quasiparticles and superfluidity can coexist even in generic two-dimensional lattice models with flat bands, for which it is not known how to construct local conserved quantities. Our results open new perspectives on the old condensed matter problem of the interplay between superconductivity and localization.
△ Less
Submitted 22 December, 2023; v1 submitted 13 February, 2023;
originally announced February 2023.