TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Brotli-G: A GPU compression/decompression standard for digital assets

88 pointsby josephscottover 2 years ago

7 comments

peter_d_shermanover 2 years ago
There&#x27;s an interesting side point here: There exist (and there will be more of these in the future) data compression algorithms which are in general, <i>too slow</i> for specific software use-cases. (related: <a href="https:&#x2F;&#x2F;superuser.com&#x2F;questions&#x2F;263335&#x2F;compression-software-with-slow-rates-and-very-high-compression" rel="nofollow">https:&#x2F;&#x2F;superuser.com&#x2F;questions&#x2F;263335&#x2F;compression-software-...</a>).<p>The thing is -- they typically run too slow for their intended applications on current, consumer-grade CPU&#x27;s...<p>But, could some of them be optimized to take advantage of GPUs (as Brotli is here) -- and would that then increase their performance to a level such that applications which previously could not use them because the algorithm previously took so long -- can now make use of them IF the software end-user has the proper GPU?<p>?<p>I think there&#x27;s a huge amount of possibilities here...<p>Especially when you get to compression algorithms that include somewhat esoteric stuff, like Fourier Transforms, Wavelet Transforms -- and other weird and esoteric math algorithms both known and yet-to-be-discovered...<p>In other words, we&#x27;ve went far beyond Huffman and Lempel-Ziv for compression when we&#x27;re in this territory...<p>(In fact, there should be a field of study... the confluence&#x2F;intersection of all GPUs and all known compression algorithms... yes, I know... something like that probably already exists(!)... but I&#x27;m just thinking aloud here! &lt;g&gt;)<p>In conclusion, I think there&#x27;s a huge amount of interesting future possibilities in this area...
评论 #33711962 未加载
评论 #33710779 未加载
评论 #33708499 未加载
corysamaover 2 years ago
There is a new interest in GPU-side general-purpose decompression due to Microsoft pushing DirectStorage.<p><a href="https:&#x2F;&#x2F;devblogs.microsoft.com&#x2F;directx&#x2F;directstorage-1-1-coming-soon&#x2F;" rel="nofollow">https:&#x2F;&#x2F;devblogs.microsoft.com&#x2F;directx&#x2F;directstorage-1-1-com...</a>
评论 #33711489 未加载
telendramover 2 years ago
Interesting, nVidia selected to go with zstd instead: <a href="https:&#x2F;&#x2F;developer.nvidia.com&#x2F;nvcomp" rel="nofollow">https:&#x2F;&#x2F;developer.nvidia.com&#x2F;nvcomp</a>
rektideover 2 years ago
Im interested to see what image format specific compressors go gpu. JpegXL, AVIF, WebP... who wants to show up &amp; throw down? Or even just fastpng?<p>Meanwhile we dont really hear or regard many of the gpu-oriented compression techs. TIL Basis&#x2F;KTX2 is itself zstd compressed (formerly LZ apparently?). <a href="https:&#x2F;&#x2F;github.com&#x2F;BinomialLLC&#x2F;basis_universal" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;BinomialLLC&#x2F;basis_universal</a>
darrinmover 2 years ago
Benchmarks?
pstuartover 2 years ago
It requires the source asset to be compressed using brotli-g as well, so it&#x27;s dependent on the hosting servers to be of any value.
评论 #33710662 未加载
评论 #33715583 未加载
ttoinouover 2 years ago
Anyone knows if there is a way to tune this algorithm to be lossy ?
评论 #33721753 未加载