HP Labs is the exploratory and advanced research group for HP Inc. HP Labs' headquarters is in Palo Alto, California and the group has research and development facilities in Bristol, UK. The development of programmable desktop calculators, inkjet printing, and 3D graphics are credited to HP Labs researchers. HP Labs was established on March 3, 1966, by founders Bill Hewlett and David Packard, seeking to create an organization not bound by day-to-day business concerns. HP Labs was created with the help of former Bell Labs engineer and MOSFET (MOS transistor) inventor Mohamed Atalla, who later left in 1972.The labs have downsized dramatically; in August 2007, HP executives drastically diminished the number of projects, down from 150 to 30. As of 2018, HP Labs has just over 200 researchers, compared to earlier staffing levels of 500 researchers.With the Hewlett Packard Enterprise being spun off from Hewlett-Packard in November 1, 2015 and renamed to and HP Inc., the research lab also spun off Hewlett Packard Labs to Hewlett Packard Enterprise and HP Labs was kept for HP Inc.
Optical computing aims to leverage the large bandwidth of light and exploits its potential for low-latency signal processing to efficiently handle large-scale Machine Learning (ML) workloads. A critical component of this approach is a Photonic Tensor Core (PTC), which facilitates general matrix-vector products. However, computational accuracy can be degraded by noise and fabrication imperfections, posing challenges for practical deployment. In this work, we address the problem of designing noise-robust PTCs within footprint-constrained architectures composed of Phase Shifters (PSs), Directional Couplers (DCs), and Waveguide Crossings (CRSs). Using a recently developed automated design framework, we propose and demonstrate in simulations the efficacy of two complementary techniques: noise injection during chip topology training and penalty terms that promote noise robust and imperfection-insensitive local optima in the loss landscape of the ML task. We find that, compared to the original design, the ML task accuracy degrades significantly slower under increasing noise levels when the proposed robustness techniques are used. We increase the average accuracy in the presence of DC biases from 87.2% to 97.8% using noise injection on the MNIST classification task. Furthermore, using the gradient-based techniques, we increase the accuracy when the DCs experience high variations from ∼81% to ∼84% on the F-MNIST classification task. These advances contribute to the practical viability of photonic accelerators for ML and potentially beyond, by addressing noise and imperfections through hardware-algorithm co-design.
For over a decade, the problem of distributed cloud workload management has been studied with the goal of co-optimizing operational costs with other metrics, such as energy efficiency, using multi-objective optimization algorithms. However, there is a lack of a multi-objective algorithm that can not only provide a diverse and high-quality Pareto-optimal solution set, but also scale to different cloud management scenarios. The heterogeneity of cloud workloads and geo-distributed datacenters introduces complex operational scenarios over time and geographic locations. The introduction of emerging sustainability objectives, such as carbon emissions and wastewater generation, further aggravates the cloud management challenge. Moreover, inter-datacenter network costs must also be considered during cloud workload management to prevent unrealistic workload migration scenarios. In this article, we propose a novel cloud resource management framework called SHIELD-EB to co-optimize operational costs, operational carbon emissions, and wastewater generation in a geo-distributed cloud datacenter platform. To generate a diverse solution set with different tradeoffs, SHIELD-EB integrates a customized evolutionary strategy (CES) with eXtreme Gradient Boosting (XGB). Experimental results show that SHIELD-EB can achieve improvements of up to 33% in Pareto hypervolume, 7.1% ($0.6 M) in operational costs, 11.6% (7.6 tons) in operational carbon, and 12.0% (102.4 tons) in water usage compared to the state-of-the-art over a duration of 120 hours.
A key trait of stochastic optimizers is that multiple runs of the same optimizer in attempting to solve the same problem can produce different results. As a result, their performance is evaluated over several repeats, or runs, on the problem. However, the accuracy of the estimated performance metrics depends on the number of runs and should be studied using statistical tools. We present a statistical analysis of the common metrics, and develop guidelines for experiment design to measure the optimizer's performance using these metrics to a high level of confidence and accuracy. To this end, we first discuss the confidence interval of the metrics and how they are related to the number of runs of an experiment. We then derive a lower bound on the number of repeats in order to guarantee achieving a given accuracy in the metrics. Using this bound, we propose an algorithm to adaptively adjust the number of repeats needed to ensure the accuracy of the evaluated metric. Our simulation results demonstrate the utility of our analysis and how it allows us to conduct reliable benchmarking as well as hyperparameter tuning and prevent us from drawing premature conclusions regarding the performance of stochastic optimizers.
Resistive Random Access Memory (RRAM) based in-memory computing (IMC) accelerators offer significant performance and energy advantages for deep neural networks (DNNs), but face three major limitations: (1) they support only static dot-product operations and cannot accelerate arbitrary non-linear functions or data-dependent multiplications essential to modern LLMs; (2) they demand large, power-hungry analog-to-digital converter (ADC) circuits; and (3) mapping model weights to device conductance introduces errors from cell nonidealities. These challenges hinder scalable and accurate IMC acceleration as models grow. We propose NL-DPE, a Non-Linear Dot Product Engine that overcomes these barriers. NL-DPE augments crosspoint arrays with RRAM-based Analog Content Addressable Memory (ACAM) to execute arbitrary non-linear functions and data-dependent matrix multiplications in the analog domain by transforming them into decision trees, fully eliminating ADCs. To address device noise, NL-DPE uses software-based Noise Aware Fine-tuning (NAF), requiring no in-device calibration. Experiments show that NL-DPE delivers 28X energy efficiency and 249X speedup over a GPU baseline, and 22X energy efficiency and 245X speedup over existing IMC accelerators, while maintaining high accuracy.
Recent experiments on quantum computers have challenged the limits of classical computation in chemistry, simulating ground states of strongly correlated molecules. Many of these experiments have utilized the unitary cluster Jastrow ansatz, a quantum circuit inspired by the unitary coupled cluster ansatz that can be tailored to current quantum hardware. Notably, the largest experiment in Sci. Adv. 11, 25 (2025) executed a quantum circuit with 77 qubits and 10,570 gates on an IBM quantum computer and performed classical post-processing with up to 6400 nodes on Fugaku to compute ground state energies better than Hartree-Fock. In this work, we present a polynomial time classical algorithm to compute the energy of any single-layer unitary cluster Jastrow circuit, independent of locality constraints for quantum hardware. Our algorithm can reproduce the largest experiment from Sci. Adv. 11, 25 (2025) in less than a minute on a laptop, and through circuit optimization enabled by fast simulation we achieve a lower ground state energy than the experiment.
