Intel's SVS library offers significant performance advantages over HNSWlib for vector search tasks. With Intel Xeon 6972P processors, SVS can handle large-scale datasets (45 million vectors) and high concurrency (over 200,000 users) more efficiently than HNSWlib. This superior performance makes SVS an ideal choice for demanding AI and machine learning applications.
In the deep learning era, high-dimensional vectors have become the quintessential data representation for all forms of unstructured data. Searching for semantically similar vectors is a critical task for many modern applications. They include semantic search during web search, recommender systems, and emerging applications, notably retrieval augmented generation (RAG) for factual grounding, long term memory and hallucination mitigation in AI systems.
Intel Labs has developed an optimized Scalable Vector Search (SVS) performance library for vector similarity search.
As a software developer, you can take advantage of the SVS library through GitHub by visiting
[ Ссылка ]
Using the latest generation of the Intel® Xeon® 6972P Processor (96C), this video compares the Intel SVS library to the widely used open source vector search library HNSWlib, showing a performance benefit of over 8x in latency over HNSWlib when using 45 million vector embeddings with 1536 dimensions. On HNSWlib, vector search starts to backlog when there are more than 25 thousand concurrent users, slowing down the chatbot response or making it unresponsive. Intel® Scalable® Vector Search (SVS) can run over 200,000 concurrent users with the same configuration.
We also compare the similarity search performance using Intel SVS on both AMD* EPYC* 9654 Processor (96c) and the Intel® Xeon® 6972P Processor (96c). We offer both community and Intel-optimized edition of Intel SVS. Platforms can take advantage using this library, and Intel hardware will take advantage of the Intel-optimized features such as vector compression. This example shows the superiority of Intel hardware + Intel optimized software on this workload. The Intel® Xeon® 6 platform similarity search performance is 6.8x faster than AMD Genoa at the same core count. One would need various Genoa servers to deliver the same vector search capabilities that can be completed by a single Xeon 6 system.
Performance claims used can be found here:
[ Ссылка ]
Intel Innovation: [ Ссылка ]
Intel Events: [ Ссылка ]
About Intel Software:
Intel® Developer Zone is committed to empowering and assisting software developers in creating applications for Intel hardware and software products. The Intel Software YouTube channel is an excellent resource for those seeking to enhance their knowledge. Our channel provides the latest news, helpful tips, and engaging product demos from Intel and our numerous industry partners. Our videos cover various topics; you can explore them further by following the links.
Connect with Intel Software:
INTEL SOFTWARE WEBSITE: [ Ссылка ]
INTEL SOFTWARE on FACEBOOK: [ Ссылка ]
INTEL SOFTWARE on TWITTER: [ Ссылка ]
INTEL SOFTWARE GITHUB: [ Ссылка ]
INTEL DEVELOPER ZONE LINKEDIN: [ Ссылка ]
INTEL DEVELOPER ZONE INSTAGRAM: [ Ссылка ]
INTEL GAME DEV TWITCH: [ Ссылка ]
#intelsoftware #innovation
Intel® Scalable Vector Search on the Intel® Xeon® 6 Processor | Demo | Innovation Selects
