Pietro Mercati

Pietro Mercati

Auto-generated profile, based on publicly available information

Pietro Mercati

Intel Corporation, Hillsboro, OR, USA

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 September 2016 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Verified
Journal Article
Open access
PMPietro Mercati
Pietro Mercati
FPFrancesco Paterna
Francesco Paterna
Andrea BartoliniAndrea Bartolini
Andrea Bartolini
LBLuca Benini
Luca Benini
TRTajana Simunic Rosing
Tajana Simunic Rosing
With CMOS scaling beyond 14 nm, reliability is a major concern for IC manufacturers. Reliability-aware design has a non-negligible overhead and cannot account for user experience in mobile devices. An alternative is dynamic reliability management (DRM), which counteracts degradation by adapting the operating conditions at runtime. In this paper, for the first time we formulate DRM as an optimization problem that accounts for reliability, temperature and performance. We develop an optimal policy for multicores using convex optimization, and show that it is not feasible to implement on real systems. For this reason, we propose workload-aware reliability management (WARM), a fast DRM technique adapting to diverse workload requirements to trade reliability and user experience. WARM is implemented and tested on a real Android device. WARM approximates the solution of the convex solver within 5% on average, while executing more than $400 {\times }$ faster. WARM integrates a thermal controller that allocates tasks to meet thermal constraints. This is required since degradation strongly depends on temperature. We show that WARM meets temperature constraints within 5% in 87.5% more cases than the state-of-the-art. We show that WARM task allocation achieves up to one year lifetime improvement for a multicore platform. It can achieve up to 100% of performance improvement on cluster architectures, such as big.LITTLE, while still guaranteeing the reliability target. Finally, we show that it achieves performance in the 4% of the maximum for a broad range of a applications, while meeting the reliability constraints.

5 June 2023
Verified
Conference Paper
Open access
YNYang Ni
Yang Ni
DADanny Abraham
Danny Abraham
MIMariam Issa
Mariam Issa
YKYeseong Kim
Yeseong Kim
PMPietro Mercati
Pietro Mercati
MIMohsen Imani
Mohsen Imani
Reinforcement Learning (RL) has opened up new opportunities to enhance existing smart systems that generally include a complex decision-making process. However, modern RL algorithms, e.g., Deep Q-Networks (DQN), are based on deep neural networks, resulting in high computational costs. In this paper, we propose QHD, an off-policy value-based Hyperdimensional Reinforcement Learning, that mimics brain properties toward robust and real-time learning. QHD relies on a lightweight brain-inspired model to learn an optimal policy in an unknown environment. On both desktop and power-limited embedded platforms, QHD achieves significantly better overall efficiency than DQN while providing higher or comparable rewards. QHD is also suitable for highly-efficient reinforcement learning with great potential for online and real-time learning. Our solution supports a small experience replay batch size that provides 12.3 times speedup compared to DQN while ensuring minimal quality loss. Our evaluation shows QHD capability for real-time learning, providing 34.6 times speedup and significantly better quality of learning than DQN.

1 January 2014
Verified
Conference Paper
Closed access
PMPietro Mercati
Pietro Mercati
Andrea BartoliniAndrea Bartolini
Andrea Bartolini
FPFrancesco Paterna
Francesco Paterna
TRTajana Simunic Rosing
Tajana Simunic Rosing
LBLuca Benini
Luca Benini
Reliability is a major concern in multiprocessors. Dynamic Reliability Management (DRM) aims at trading off processor performance with lifetime. The state-of-the-art publications study only the theory supported by simulation. This paper presents the first complete software implementation, working on a real hardware, of a low-overhead, Android-compatible workload-aware DRM Governor for mobile multiprocessors. We discuss the design challenges and the run-time overhead involved. We show the effectiveness of our governor in guaranteeing the predefined target lifetime and show that it achieves up to 100% of lifetime improvement with respect to traditional governors, while providing comparable performance for critical applications.

29 July 2024 in IEEE Design & Test
Verified
Journal Article
Closed access
PMPietro Mercati
Pietro Mercati
GBGanapati Bhat
Ganapati Bhat
Wearable and Internet of Things devices (IoT) are becoming ubiquitous in health applications, such as movement disorders, rehabilitation, and activity monitoring. Wearable IoT devices are enabling these applications through low-power sensors, processors, and machine learning algorithms. However, widespread adoption of wearable IoT devices has been limited due to challenges such as energy sustainability, sensor data shift during real-world usage, missing data, privacy, and security. Multiple health organizations have stipulated that resolving these challenges is crucial to bolster the adoption of wearable IoT devices. This paper reviews the primary challenges to wearable IoT devices and current state-of-the-art in addressing these challenges. Specifically, the goal of this paper is also to review state-of-the-art research across multiple challenges to wearable IoT devices, as opposed to focusing on a single area. We also highlight some future research directions and open problems to achieve widespread adoption of wearable IoT devices.

1 August 2014
Verified
Conference Paper
Closed access
PMPietro Mercati
Pietro Mercati
Andrea BartoliniAndrea Bartolini
Andrea Bartolini
FPFrancesco Paterna
Francesco Paterna
LBLuca Benini
Luca Benini
TRTajana Rosing
Tajana Rosing
Technology scaling made reliability a primary concern for integrated circuits. Increased power and temperature exasperate the impact of degradation phenomena and shorten processors lifetime. This issue is particularly dramatic for mobile processors, characterized by variable workload and environmental conditions. Due to the different time scales at which reliability phenomena and computation happens, state-of-theartDRM solutions are evaluated using high-level workload and system models. To enable the design of workload-aware DRM with accurate reliability models, in this work we propose a software framework for virtualizing the processors reliability.Our framework captures the effect of variable workload and environmental conditions and allows to emulate longer degradation in a short time scale. We implement the framework on a realAndroid device and exploit it to enable workload-aware DynamicReliability Management (DRM).

20 September 2024 in IEEE Access
Verified
Journal Article
Open access
YNYang Ni
Yang Ni
DADanny Abraham
Danny Abraham
MIMariam Issa
Mariam Issa
AHAlejandro Hernández-Cano
Alejandro Hernández-Cano
MIMahdi Imani
Mahdi Imani
MIMohsen Imani
Mohsen Imani
In wireless networks, dynamic spectrum access is the key to improving spectrum utilization and increasing channel capacity. Since the channels in wireless networks are highly correlated, they require intelligent algorithms to dynamically handle multi-channel access. Reinforcement Learning (RL) algorithms are introduced as effective techniques to optimize network performance. However, current RL methods heavily rely on a computationally intensive deep neural network (DNN) that is not friendly for edge devices. In this paper, we propose HD-RL, a dynamic wireless channel-sharing solution that utilizes brain-inspired lightweight hyperdimensional computing as the learning engine. HD-RL mimics important brain functionalities towards high-efficiency and noise-tolerant computation. HD-RL naturally encodes and memorizes prior knowledge to provide the near-optimal policy for channel throughput and α-fairness in the wireless network. Our evaluation shows that HD-RL achieves maximum throughput and fairness while significantly improving efficiency compared to the state-of-the-art DNN-based RL algorithms. In particular, HD-RL achieves more than 20× speedup for reaching the fairness objective. On average, the speedup of convergence time is more than 10× compared to the baseline. Our results also indicate that HD-RL has substantially higher robustness against possible hardware failure, e.g., up to 40% dimension loss in the model.