# About the dense vector compression

**URL:** <https://discuss.elastic.co/t/about-the-dense-vector-compression/319398>\
**Category:** Elasticsearch\
**Tags:** vector-search\
**Created:** [November 21, 2022, 8:36am UTC](https://discuss.elastic.co/t/about-the-dense-vector-compression/319398 "2022-11-21T08:36:48Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![myazi](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/myazi/32/113566_2.png) [@myazi](https://discuss.elastic.co/u/myazi)\
**Post date:** [November 21, 2022, 8:36am UTC](https://discuss.elastic.co/t/about-the-dense-vector-compression/319398/1 "2022-11-21T08:36:48Z")

</div>

Does elasticsearch have plans to support product quantization and SQ8 quantization? we have close to 10 Billion semantic vector index. The dense vector is 768 dimensional.

---

<div class="post-metadata">

**Author:** ![BenTrent](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/bentrent/32/33915_2.png) [@BenTrent](https://discuss.elastic.co/u/BenTrent)\
**Post date:** [November 22, 2022, 12:00am UTC](https://discuss.elastic.co/t/about-the-dense-vector-compression/319398/2 "2022-11-22T00:00:18Z")

</div>

@myazi , Elasticsearch doesn't do any quantization. But, with version 8.6 (the next version to be released), we will support signed int8 quantized values for vectors. So, in 8.6.0, quantize your vectors to int8 and index them.

Be sure to set `element_type` to `byte` in the mapping.

Note: to be clear, 8.6.0 is not released at the time when I wrote this.

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [December 20, 2022, 12:00am UTC](https://discuss.elastic.co/t/about-the-dense-vector-compression/319398/3 "2022-12-20T00:00:55Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.