NVIDIA Releases Vera Storage Benchmark: CRC Verification Performance Improved 3.67x, Multiple Storage Tasks Surpass x86

Odaily News NVIDIA has released Vera CPU storage benchmark results, showing that its NVIDIA BlueField-4 STX storage processor can significantly improve encryption, compression, integrity verification, and data recovery performance in AI-native storage, helping enterprises address the growing data processing demands of the Agentic AI era.

NVIDIA stated that as AI agents execute knowledge retrieval, call tools, manage long-term memory, and process larger context windows, storage systems are no longer just simple data read/write components but have become a critical link in the AI inference pipeline. Large volumes of data need to undergo encryption, compression, verification, and recovery along the storage path, tasks typically handled by CPUs, which may become a performance bottleneck for AI infrastructure. Benchmark results show that the BlueField-4 STX storage processor equipped with the Vera CPU achieved performance improvements across multiple storage tasks compared to a reference x86 CPU: AES-128 encryption performance improved by up to 1.43x

AES-128 decryption performance improved by up to 1.29x

Reed-Solomon data recovery performance improved by up to 3.26x

CRC32C integrity verification performance improved by up to 3.67x

Compression performance improved by up to 3.29x

Decompression performance improved by up to 1.72x

Compression + encryption multi-stage storage pipeline performance improved by up to 3.21x

NVIDIA noted that traditional CPU-based storage scaling typically requires adding cores, power consumption, and cooling costs, while AI-native storage demands low latency under higher concurrency and larger data volumes. By increasing processing capability per CPU resource, Vera helps storage systems support more AI agent workloads without significantly increasing infrastructure costs.