Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Performance evaluation of image processing algorithms on the GPU.

Daniel Castaño-Díez1, Dominik Moser, Andreas Schoenegger

  • 1Computational and Structural Biology, European Molecular Biology Laboratory, Meyerhofstr. 1, 69117 Heidelberg, Germany. castano@embl.de

Journal of Structural Biology
|August 12, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

Parallel Processing01:20

Parallel Processing

The brain processes sensory information rapidly due to parallel processing, which involves sending data across multiple neural pathways at the same time. This method allows the brain to manage various sensory qualities, such as shapes, colors, movements, and locations, all concurrently. For instance, when observing a forest landscape, the brain simultaneously processes the movement of leaves, the shapes of trees, the depth between them, and the various shades of green. This enables a quick and...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Tomogram exploration through template matching and deep learning.

Current opinion in structural biology·2026
Same author

Cargo-Adaptor Cooperation Programs Retromer Coat Architecture.

bioRxiv : the preprint server for biology·2026
Same author

Molecular mechanism of transition-state inhibitors of bacterial antibiotic efflux pumps.

npj antimicrobials and resistance·2026
Same author

Enhancing Charge Transport and Extraction in Ta<sub>3</sub>N<sub>5</sub> Photoanodes Via Ti Doping and TiN Contact Layers.

Small (Weinheim an der Bergstrasse, Germany)·2026
Same author

copick: An open dataset interface and toolkit for collaborative annotation and analysis of cryo-electron tomography data.

Protein science : a publication of the Protein Society·2026
Same author

Sex differences in patients hospitalized for repair of intact abdominal aortic aneurysms in 11 high-income countries: A cross-sectional cohort study.

Journal of vascular surgery·2026
Same journal

MLAC: MicroED-assisted ligand structure analysis in complexes and its application to hERG-ligand complexes.

Journal of structural biology·2026
Same journal

Ultrastructural evidence of autophagy-related processes and mitochondrial remodeling in the myxozoan parasite Henneguya piaractus.

Journal of structural biology·2026
Same journal

Architecture and dynamics of a supramolecular oxygen transport system in human homogentisate 1,2-Dioxygenase.

Journal of structural biology·2026
Same journal

Connecting pathways between mineralized fibrocartilage and bone at the Achilles tendon insertion.

Journal of structural biology·2026
Same journal

Structural and functional characterization of thermostable EstS1 esterase for BHET degradation.

Journal of structural biology·2026
Same journal

Following the white rabbit: multiscale 2D3D correlative imaging of bone structure.

Journal of structural biology·2026
See all related articles

Graphics processing units (GPUs) now act as powerful co-processors, accelerating computationally intensive applications like 3D image processing. Utilizing the CUDA platform, GPUs achieve 10-20x speed-ups over traditional CPUs for common algorithms.

Area of Science:

  • Computer Science
  • Image Processing
  • Scientific Computing

Background:

  • Graphics Processing Units (GPUs) have evolved from visualization tools to powerful co-processors.
  • Elaborate interfaces enable GPUs to process data and handle computationally intensive tasks.
  • GPU performance is optimal for algorithms with high data parallelism and arithmetic intensity.

Purpose of the Study:

  • To evaluate the performance of Graphics Processing Units (GPUs) for common three-dimensional image processing algorithms.
  • To assess the speed-up factors achievable by porting C code to GPUs using the CUDA platform.
  • To analyze the performance gains across various algorithms including spatial transformations, Fourier operations, pattern recognition, reconstruction, and classification.

Main Methods:

Related Experiment Videos

  • Implementation of common 3D image processing algorithms on a new software platform, CUDA.
  • Direct porting of C code to the GPU architecture.
  • Performance evaluation by comparing GPU execution times against a state-of-the-art conventional central processing unit (CPU).

    • Main Results:

    • Direct porting of C code to the GPU achieved typical acceleration factors of 10-20 times compared to conventional processors.
    • Speed-up varied depending on the specific algorithm's characteristics.
    • The performance gains were realized on standard workstations without additional hardware costs.

    Conclusions:

    • GPUs, when programmed with platforms like CUDA, offer significant computational acceleration for 3D image processing tasks.
    • The direct translation of C code to GPU architecture is an effective method for achieving substantial speed-ups.
    • The use of GPUs for computationally intensive scientific applications is a cost-effective approach, leveraging existing hardware.