Related papers: GPU performance in Run3 ALICE online/offline recon…

As exascale systems reach unprecedented concurrency, traditional performance analysis tools struggle with the overhead of massive-scale telemetry. We present an accelerated infrastructure for the hpcanalysis framework that leverages a…

The ALICE experiment profited of the Long Shutdown during 2019-2021 in order to expand its physics capabilities and fully profit from the increased LHC luminosity in Run 3. The Inner Tracking System has been replaced with a new silicon…

Analytical hardware performance models yield swift estimation of desired hardware performance metrics. However, developing these analytical models for modern processors with sophisticated microarchitectures is an extremely laborious task…

ALICE is the CERN LHC experiment optimised for the study of the strongly interacting matter produced in heavy-ion collisions and devoted to the characterisation of the quark-gluon plasma. To achieve the physics program for LHC Run 3, a…

Starting in 2022, the upgraded LHCb detector will collect data with a pure software trigger. In its first stage, reducing the rate from 30MHz to about 1MHz, GPUs are used to reconstruct and trigger on B and D meson topologies and high-pT…

The ALICE experiment at the Large Hadron Collider (LHC) proposes major detector upgrades to fully exploit the increase of the luminosity of the LHC in RUN~3 and to extend the physics reach for rare probes at low transverse momentum. The…

The ALICE Experiment has replaced its Inner Tracking System with a 7-layer pixel-only tracker made out of more than 24000 monolithic active pixel sensor chips, in order to fulfill the requirements of the physics program of the LHC Run 3.…

Performance optimization can be a daunting task especially as the hardware architecture becomes more and more complex. This paper takes a kernel from the Materials Science code BerkeleyGW, and demonstrates a few performance analysis and…

Graphics Processing Units allow for running massively parallel applications offloading the CPU from computationally intensive resources, however GPUs have a limited amount of memory. In this paper a trie compression algorithm for massively…

The recently proposed open-source KAZE image feature detection and description algorithm offers unprecedented performance in comparison to conventional ones like SIFT and SURF as it relies on nonlinear scale spaces instead of Gaussian…

The torrential influx of floating-point data from domains like IoT and HPC necessitates high-performance lossless compression to mitigate storage costs while preserving absolute data fidelity. Leveraging GPU parallelism for this task…

Data compression and decompression have become vital components of big-data applications to manage the exponential growth in the amount of data collected and stored. Furthermore, big-data applications have increasingly adopted GPUs due to…

High throughput and low latency data processing is essential for systems requiring live decision making, control, and machine learning-optimized data reduction. We focus on two distinct use cases for in-flight streaming data processing for…

The ALICE Collaboration will undertake a major upgrade of the detector apparatus during the second LHC Long Shutdown LS2 (2019-2020) in view of the Runs 3 and 4 (2021-2029). The objective of the upgrade is two-fold: i) an improvement of the…

Pipelining between data loading and computation is a critical tensor program optimization for GPUs. In order to unleash the high performance of latest GPUs, we must perform a synergetic optimization of multi-stage pipelining across the…

Large inter-GPU all-reduce operations, prevalent throughout deep learning, are bottlenecked by communication costs. Emerging heterogeneous architectures are comprised of complex nodes, often containing $4$ GPUs and dozens to hundreds of CPU…

Simulations of physical phenomena are essential to the expedient design of precision components in aerospace and other high-tech industries. These phenomena are often described by mathematical models involving partial differential equations…

The ALICE Collaboration completed the upgrade of the detector and is now commissioning for the beginning of the data taking during LHC Run 3. In parallel, R&D activities and simulation studies are being performed to define the future of the…

This paper presents ARCS (Autoregressive Circuit Synthesis), a system for amortized analog circuit generation. ARCS produces complete, SPICE-simulatable designs (topology and component values) in milliseconds rather than the minutes…

As Graphics Processing Units (GPUs) have gained in capability and GPU development environments have matured, developers are increasingly turning to the GPU to off-load the main host CPU of numerically-intensive, parallelizable computations.…