Jose Luquin

Jose Luquin

Auto-generated profile, based on publicly available information

Jose Luquin

IBM Research—Almaden, San Jose, USA

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
19 December 2025 in Nature Electronics
Verified
Journal Article
Closed access
PDF file
ASAbhairaj Singh
Abhairaj Singh
MGManuel Le Gallo
Manuel Le Gallo
AVAthanasios Vasilopoulos
Athanasios Vasilopoulos
JLJose Luquin
Jose Luquin
PNPritish Narayanan
Pritish Narayanan
ASAbu Sebastian
Abu Sebastian

30 September 2025 in Nature Communications
Verified
Journal Article
Open access
PDF file
ACAn Chen
An Chen
SAStefano Ambrogio
Stefano Ambrogio
PNPritish Narayanan
Pritish Narayanan
AOAtsuya Okazaki
Atsuya Okazaki
CMCharles Mackin
Charles Mackin
GBGeoffrey W. Burr
Geoffrey W. Burr
A Lite Bidirectional Encoder Representations from Transformers model is demonstrated on an analog inference chip fabricated at 14nm node with phase change memory. The 7.1 million unique analog weights shared across 12 layers are mapped to a single chip, accurately programmed into the conductance of 28.3 million devices, for this first analog hardware demonstration of a meaningfully large Transformer model. The implemented model achieved near iso-accuracy on the General Language Understanding Evaluation benchmark of seven tasks, despite the presence of weight-programming errors, hardware imperfections, readout noise, and error propagation. The average hardware accuracy was only 1.8% below that of the floating-point reference, with several tasks at full iso-accuracy. Careful fine-tuning of model weights using hardware-aware techniques contributes an average hardware accuracy improvement of 4.4%. Accuracy loss due to conductance drift – measured to be roughly 5% over 30 days – was reduced to less than 1% with a recalibration-based “drift compensation” technique.

30 October 2025 in Research Square
Verified
Preprint
Open access
PDF file
PNPritish Narayanan
Pritish Narayanan
SAStefano Ambrogio
Stefano Ambrogio
CMCharles Mackin
Charles Mackin
MIMasatoshi Ishii
Masatoshi Ishii
BKBenjamin Killeen
Benjamin Killeen
GBGeoffrey Burr
Geoffrey Burr
Analog AI implements the multiply-accumulate operations that dominate deep learning at the location of the weight data, offering orders of magnitude performance and energy improvements over conventional digital systems. However, translating these benefits to the Convolutional Neural Networks (CNN) widely used in applications such as image and speech processing is non-trivial. Significant reuse of both weights and activations, together with the need to extensively rearrange activations between layers, require micro-architectural solutions that span across weight mapping strategy, activation positioning, pipelining between stages, and data transport. In this paper, we describe a weight-stationary Analog AI micro-architecture for Convolutional Neural Networks (CNNs), called Row-By-Row- (RBR-) CNNs and its associated circuits and pipelines. We show a hardware demonstration of RBR-CNNs on a 14nm Analog AI inference chip with Phase Change Memory (PCM), achieving software-equivalent accuracy on the ML Perf Benchmark “Key Word Spotting” task using 4 RBR CNN layers. We show RBR CNNs using Analog AI can achieve 7x–15x latency improvements vs. high performance chips reported on ML Perf, while offering extremely high energy-efficiency – at least two orders of magnitude better than published low-power edge chip results, indicating strong applicability in embedded and mobile settings.