e.l.f. Beauty stock plummets 20% as revenue and guidance fall short of expectations
SAN FRANCISCO - Elastic (NYSE:ESTC) has released DiskBBQ, a new vector search algorithm for Elasticsearch that aims to provide more efficient vector search at scale compared to commonly used industry techniques. The cloud search company, which generated $1.55 billion in revenue with 17.42% growth over the last twelve months, continues to innovate despite not yet achieving profitability. According to InvestingPro data, Elastic holds more cash than debt on its balance sheet, positioning it well for continued R&D investments.
Available in Elasticsearch 9.2, DiskBBQ is designed to eliminate the need to keep entire vector indexes in memory, which the company claims delivers more predictable performance at lower cost than traditional approaches like Hierarchical Navigable Small Worlds (HNSW).
The new algorithm uses Better Binary Quantization (BBQ) to compress vectors and cluster them into compact partitions for selective disk reads, reducing RAM requirements. According to the company, this approach helps avoid spikes in data retrieval time while improving system performance for data ingestion and organization.
"DiskBBQ is a smarter, more scalable approach to high-performance vector search on very large datasets that accelerates both indexing and retrieval," said Ajay Nair, general manager of Platform at Elastic.
In benchmark testing conducted by Elastic, DiskBBQ reportedly maintained query latencies of approximately 15 milliseconds while operating with as little as 100 MB of total memory, in scenarios where the company claims traditional HNSW indexing could not function. The company states that as available memory increased, DiskBBQ’s performance scaled without the sharp latency issues that can affect in-memory approaches.
The algorithm is designed to offload data to disk and read only relevant vector clusters at query time, potentially allowing Elasticsearch to scale to larger datasets limited primarily by CPU and disk rather than memory.
DiskBBQ is currently available in technical preview in Elasticsearch Serverless, according to the press release statement.
In other recent news, Elastic announced the release of Streams, an AI-powered solution aimed at enhancing the efficiency of log data analysis for site reliability engineers. This new tool automatically processes raw logs to highlight critical errors and anomalies, as stated in a company press release. Additionally, Elastic introduced Agent Builder, which allows developers to create custom AI agents using company data, enhancing the process of context engineering. Stifel reiterated its Buy rating on Elastic, citing growth opportunities in the generative AI sector and maintaining a price target of $134.00. TD Cowen, however, maintained a Hold rating with a $105.00 price target, noting concerns about AI growth despite Elastic’s optimistic sales projections. Meanwhile, William Blair reaffirmed an Outperform rating, emphasizing Elastic’s strong position in the AI and search markets, as well as its potential to expand in observability and security sectors. These developments reflect Elastic’s strategic focus on AI capabilities and its positioning for future growth.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.