---
title: "NVIDIA Vera Storage Benchmarks: Faster Encryption, Compression, Integrity Checking, and Recovery for AI-Native Storage"
description: "Agentic AI architectures require storage systems to secure, compress, and validate massive streams of persistent data and key-value cache entries across concurrent reasoning loops. These tasks run primarily on CPUs and frequently create processing bottlenecks that prevent modern networks and flash storage from operating at full capacity. NVIDIA evaluated the Vera CPU within the BlueField-4 STX Storage Processor using a purpose-built in-memory microbenchmark suite running standard libraries like OpenSSL, Zstandard, and LZ4. The results demonstrated that Vera outperformed x86 processors across core storage primitives including encryption, decompression, Reed-Solomon recovery, and CRC32C integrity checking. For a sequential multi-stage pipeline executing compression followed by encryption, Vera achieved up to 3.21x higher throughput than the comparison x86 CPU."
---

# NVIDIA Vera Storage Benchmarks: Faster Encryption, Compression, Integrity Checking, and Recovery for AI-Native Storage

[NVIDIA Developer Blog](https://yomu.fyi/company/nvidia-developer-blog) · Elizabeth Goodman · Aug 3, 2026

**Type:** Benchmark

## Summary

Agentic AI architectures require storage systems to secure, compress, and validate massive streams of persistent data and key-value cache entries across concurrent reasoning loops. These tasks run primarily on CPUs and frequently create processing bottlenecks that prevent modern networks and flash storage from operating at full capacity. NVIDIA evaluated the Vera CPU within the BlueField-4 STX Storage Processor using a purpose-built in-memory microbenchmark suite running standard libraries like OpenSSL, Zstandard, and LZ4. The results demonstrated that Vera outperformed x86 processors across core storage primitives including encryption, decompression, Reed-Solomon recovery, and CRC32C integrity checking. For a sequential multi-stage pipeline executing compression followed by encryption, Vera achieved up to 3.21x higher throughput than the comparison x86 CPU.

## Context

In agentic AI workflows, data management tasks, KV cache reuse, and storage services run on CPUs while inference runs on GPUs. Storage primitives such as encryption, decryption, checksum calculation, compression, and redundancy reconstruction sit directly in the data path. Scaling these operations using conventional x86 CPUs requires substantial cores, power, and cooling, making the storage CPU a potential bottleneck for high-speed SSDs and networks under large context volumes and high agent concurrency.

## Approach / What changed

NVIDIA benchmarked the Vera CPU in the BlueField-4 STX Storage Processor, which features 88 Olympus Armv9.2 cores, 176 Spatial Multithreading threads, a Scalable Coherency Fabric with 3.4 TB/s bisection bandwidth, a 164 MB unified L3 cache, and 1.2 TB/s aggregate SOCAMM2 LPDDR5X memory bandwidth. A custom in-memory testing framework isolated and evaluated storage primitives and a multi-stage write pipeline against x86 implementations.

## Takeaways

- NVIDIA Vera integrates 88 Olympus Armv9.2 cores with 176 Spatial Multithreading threads, delivering up to 3.4 TB/s fabric bisection bandwidth and 1.2 TB/s aggregate SOCAMM2 LPDDR5X memory bandwidth.
- In a memory-resident multi-stage storage write benchmark applying compression followed by encryption, Vera delivered up to 3.21x higher pipeline throughput than the baseline x86 processor.
- Standalone Vera achieved up to 1.8x higher per-core performance in agentic tools while delivering higher measured performance per watt across encryption, CRC32C validation, and Reed-Solomon recovery.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [Machine Learning](https://yomu.fyi/topic/machine-learning), [Performance](https://yomu.fyi/topic/performance), [Scalability](https://yomu.fyi/topic/scalability)

- Source: [NVIDIA Developer Blog](https://developer.nvidia.com/blog/nvidia-vera-storage-benchmarks-faster-encryption-compression-integrity-checking-and-recovery-for-ai-native-storage)
- Source URL: https://developer.nvidia.com/blog/nvidia-vera-storage-benchmarks-faster-encryption-compression-integrity-checking-and-recovery-for-ai-native-storage
- Ingested by Yomu: 2026-08-28T00:01:12.012Z

[Read original post](https://developer.nvidia.com/blog/nvidia-vera-storage-benchmarks-faster-encryption-compression-integrity-checking-and-recovery-for-ai-native-storage)