https://3demmethods.i2pc.es/index.php?action=history&feed=atom&title=2022Wang_Thunder2022Wang Thunder - Revision history2026-06-13T12:17:06ZRevision history for this page on the wikiMediaWiki 1.44.2https://3demmethods.i2pc.es/index.php?title=2022Wang_Thunder&diff=4370&oldid=prevWikiSysop: Created page with "== Citation == Wang, Zhao / Ruan, Huabin / Yang, Guangwen / Li, Xueming. Parallelizing the cryo-EM structure determination in THUNDER using GPU cluster. 2022. Engineering Rep..."2023-07-19T06:29:06Z<p>Created page with "== Citation == Wang, Zhao / Ruan, Huabin / Yang, Guangwen / Li, Xueming. Parallelizing the cryo-EM structure determination in THUNDER using GPU cluster. 2022. Engineering Rep..."</p>
<p><b>New page</b></p><div>== Citation ==<br />
<br />
Wang, Zhao / Ruan, Huabin / Yang, Guangwen / Li, Xueming. Parallelizing the cryo-EM structure determination in THUNDER using GPU cluster. 2022. Engineering Reports. p. e12601 <br />
<br />
== Abstract ==<br />
<br />
Electron cryo-microscopy (cryo-EM) is a powerful tool utilized by biologists<br />
for understanding the mysteries of life. However, obtaining high-resolution<br />
3D reconstructions from innumerable noisy images of macromolecules is an<br />
extremely complicated task, involving massive image analysis and calculation<br />
on a computing cluster. Although extensive efforts have been made for<br />
improving the computational efficiency, methods for completely utilizing the<br />
computing resources are still challenging for modern cryo-EM programs. Here,<br />
we designed a new computing approach specialized for GPU to optimize and<br />
maximize the computing power of a single GPU, multiple GPU, and the GPU<br />
cluster, highlighted by a well-designed cache structure and mixed computing<br />
precision of single-precision and double-precision. Our approaches achieved<br />
remarkable improvement in performance and linear scalability. At an identical<br />
cost of the hardware, three-fold more speed-up was achieved. The average parallel<br />
efficiency can increase up to 84% when multiple GPU configurations are<br />
parallelized.<br />
<br />
== Keywords ==<br />
<br />
== Links ==<br />
<br />
https://onlinelibrary.wiley.com/doi/full/10.1002/eng2.12601<br />
<br />
== Related software ==<br />
<br />
== Related methods ==<br />
<br />
== Comments ==</div>WikiSysop