Qdrant
-
|
In Qdrant, the maximum dimension of dense vector is 65536. I was wondering if there is any limitation of dimension when applying the HNSW indexing? Many populer vector DB like milvus, pg_vector, etc. has a limitation while applying the HNSW indexing. Though they support higher dimensional vectors. I looked into Qdrant documentation but couldn't find any information on this. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
|
There is no limit to the number of dimensions in HNSW. Or well, it's limited by the global limit of 65536. Do note that the larger your vectors get, the more expensive everything will get in terms of indexing time, search time and required resources. |
Beta Was this translation helpful? Give feedback.
There is no limit to the number of dimensions in HNSW. Or well, it's limited by the global limit of 65536.
Do note that the larger your vectors get, the more expensive everything will get in terms of indexing time, search time and required resources.