MB-59670: GPU-Accelerated Vector Search

[CascadingRadium]

• Adds the faiss_gpu_free_memory API that returns the amount of memory available in a GPU in bytes.
• Add pagination support for residual calculation in IVF SQ indexes, which allows long term indexing workloads to stay within the memory limit, avoiding GPU memory spikes and potential GPU OOM kills.

[Copilot]

Pull Request Overview

This PR adds a new C API function to retrieve the amount of free GPU memory in bytes. The implementation wraps the existing C++ faiss::gpu::getFreeMemory function to expose it to C callers.

• Adds faiss_get_free_memory function that takes a device ID and returns the free memory via an output parameter
• Follows the established pattern for C API wrappers in the codebase
• Uses the CATCH_AND_HANDLE macro for consistent error handling

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
c_api/gpu/DeviceUtils_c.h	Declares the new faiss_get_free_memory function with proper documentation
c_api/gpu/DeviceUtils_c.cpp	Implements the function by wrapping faiss::gpu::getFreeMemory with appropriate error handling

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 3 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.