Humans and systems are generating and collecting vast amounts of data every day. From social media posts and videos to sensor data, IoT devices and of course the surge in AI applications, the volume of data is growing exponentially. A significant portion of today’s data is unstructured, such as images, videos, documents, and log files. Object storage excels at handling unstructured data by allowing flexible metadata and seamless scalability. Further, it provides a foundation for data analytics and AI applications, a scalable and cost-effective solution to store and manage this massive volume of data. It allows large datasets to be stored in a distributed manner, enabling efficient data processing and analysis. Overall, object storage is important for today’s data-driven organizations because it addresses the challenges of managing massive volumes of unstructured data, ensures data durability and availability, enables advanced analytics, and provides cost-effective storage solutions.

The Ampere® Altra® Family aarch64 processors (32 to 128 cores per socket) are complete system-on-chip (SOC) solutions ideally suited for large-scale, high-performance storage. In addition to incorporating many high-performance cores, Ampere’s innovative architecture delivers predictable high performance, linear scaling, and high energy efficiency. More importantly, the high I/O bandwidth provides direct connections to multiple PCIe devices like SSD drives, which are essential for heavy workload applications.

MinIO is fully compatible with Kubernetes, and Docker deployments, and bare metal is fully supported on a swath of operating systems, including most current versions of Linux. Its open-source, software-defined, object storage system is optimized for the private cloud. MinIO is the world's fastest-growing object storage company, with more than 1 billion Docker pulls and more than 46K GitHub Stars.

MinIO offers a host of enterprise features including inline erasure coding, bit-rot detection, state-of-the-art encryption, active-active replication, object locking, lifecycle management, and identity + access management.

Typical MinIO use cases

• Data backups
• Application data caching
• Image thumbnails
• Sensor data / IoT
• Transaction logs
• Metadata storage
• CDN caching

Ampere® Altra® Family Processors Deliver Disruptive Value for Large-Scale, High-Performance Storage Solutions

• Higher core countsas compared to competitors.
• Predictable and linear workload performance.
• Leading energy efficiency reduces power costs and carbon footprint.
• High I/O bandwidth: 128 lanes PCIe Gen4 which are essential for high-performance storage solutions.

In this brief around MinIO SDS, Ampere®, MinIO, Supermicro and Samsung collaborate to showcase MinIO performance data after running tests with an Ampere Altra Max and with Intel Xeon Gold 6330 processor on a single node under 100Gbps network bandwidth.

The setup consists of a single MinIO host and four MinIO clients in which we use to collect MinIO benchmark data for various test modes. For each test mode, we collect data for a variety of object sizes, from a small (around 4k) to much bigger object size of 128M. To perform the benchmark study, we use WARP, which is an open-source full-featured S3 performance assessment software built to conduct tests between WARP clients and object storage hosts.

We ran a series of tests using MinIO SDS to benchmark the highest storage performance on a single Ampere and Intel Ice Lake node . The Supermicro servers using in this brief were each equipped with either one Ampere® Altra® Max processor (128 cores) or two Intel(R) Xeon(R) Gold 6330 CPUs, with eight Samsung MZWLR1T9HCJR-00A07 NVMe SSDs, and two Mellanox ConnectX-5/6 network cards to achieve 100Gbps bandwidth.

Table 4 shows MinIO performance running on Ampere 1P Altra Max M128-30 vs. 2P Intel Ice Lake 6330 platform across various object size from 4k-128M:

Figures are averages of various object sizes: 4k - 128M

MinIO SDS powered by Ampere® can drive sustained peak loads into the drives, delivering industry-leading performance, more efficient, and availability based on a single 2U nodes with eight Samsung NVMe drives and 100Gbps network bandwidth configuration. Consider having higher network bandwidth could result in getting even higher benchmark data.

In the other words, network bandwidth is the bottleneck.

All data and information contained herein is for informational purposes only and Ampere reserves the right to change it without notice. This document may contain technical inaccuracies, omissions and typographical errors, and Ampere is under no obligation to update or correct this information. Ampere makes no representations or warranties of any kind, including but not limited to express or implied guarantees of noninfringement, merchantability, or fitness for a particular purpose, and assumes no liability of any kind. All information is provided “AS IS.” This document is not an offer or a binding commitment by Ampere. Use of the products contemplated herein requires the subsequent negotiation and execution of a definitive agreement or is subject to Ampere’s Terms and Conditions for the Sale of Goods.

Supermicro’s Ampere® platform is a 2U form factor chassis. See efficiency footnotes for details details

System configurations, components, software versions, and testing environments that differ from those used in Ampere’s tests may result in different measurements than those obtained by Ampere.

©2024 Ampere Computing. All Rights Reserved. Ampere, Ampere Computing, Altra and the ‘A’ logo are all registered trademarks or trademarks of Ampere Computing. Arm is a registered trademark of Arm Limited (or its subsidiaries). All other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.

Ampere Computing^® / 4655 Great America Parkway, Suite 601 / Santa Clara, CA 95054 / amperecomputing.com