GPU Accelerated Indexing

5/11/2026

Introduces TurboQuant, a new quantization method developed by Google Research, which offers twice the compression ratio of scalar quantization with similar recall and speed, and improved recall compared to binary quantization at equivalent storage budgets. Also adds memory monitoring per collection, the ability to add/remove named vectors without collection recreation, audit logging improvements with a query API and tracing IDs, per-collection API metrics, and new strict mode guardrails for memory and search batch size.

4/28/2026

This post announces the availability of GPU indexing on Qdrant Cloud, detailing its performance benefits (up to 4x faster HNSW builds) and hardware requirements (NVIDIA T4, 16 GiB RAM per node). It also introduces Multi-AZ replication for improved uptime (99.95% SLA) and audit logging for API operations, which can be integrated with SIEM solutions. These features are available on specific Qdrant Cloud tiers and configurations.

1/23/2025

This post details the implementation of GPU-accelerated HNSW indexing in Qdrant 1.13, highlighting its architecture-free support for NVIDIA, AMD, and Intel GPUs via Vulkan API. It emphasizes the in-house development, vendor-agnostic approach, and performance benefits (up to 10x faster indexing). Benchmarks on common GPUs and cost comparisons are provided. The post also introduces Strict Mode for operational control, HNSW graph compression using Delta Encoding (reducing memory by up to 30%), filtering by named vectors, and a new custom storage engine.