How does Vald hold up to Milvus in terms of performance and storage? We make a lot of use of Milvus and they can do crazy fast 10ms like returns over a large se

How does Vald hold up to Milvus in terms of performance and storage? We make a lot of use of Milvus and they can do crazy fast 10ms like returns over a large set of vectors without even making a single core sweat a lot of the time and for about 200m vectors we only use like 40gb in storage at dim 128 after quantization and index building.