MinIO, the leader in high-performance storage for AI, is announcing new optimizations for Arm-based chipsets powering on its object store, expanding the software ecosystem for the power of Arm. Centering low-power, computationally dense chips, MinIO’s enhancements for Arm aim to better accommodate the AI-powered future.
One of the biggest challenges for modern AI is providing the adequate computation needed without incurring a slew of inefficiencies. Applying AI at scale requires significant power—and the capabilities of the Arm chipset may be the forerunning answer to this bottleneck.
“All the major cloud vendors went Arm—Amazon Graviton is an example [where]...they saw resounding success in the cloud,” said Anand Babu Periasamy, co-founder and CEO of MinIO. “That success is starting to percolate down into the private cloud, into the large enterprise, and we need to be ready for this.”
“Power is going to be your bottleneck in the modern infrastructure; the amount of power each rack can take is limiting [to] how much AI you can do on the data…in AI, it is going to be the deciding factor,” Periasamy continued. “Arm comes to rescue… [and it] is a necessary component for modern infrastructure.”
The latest Arm architecture offers a myriad of efficiency and scalability boons for storage workloads, including advanced functions like erasure coding, encryption, and hashing. Support for Arm and these cutting-edge functions from the software ecosystem, Periasamy explained, will be crucial for enterprise success in the AI era.
“The biggest resistance you find for Arm is not the ability of the ARM chip to deliver, it's more about availability of the software ecosystem. Great processor architectures get killed simply because of a lack of software ecosystem. Unless vendors like us support and embrace Arm, it's not going to happen,” he added.
Delivering this support, MinIO’s latest collaboration with Arm utilizes Scalable Vector Extension (SVE) version enhancements to improve the performance and efficiency of vector operations. SVE’s ability to run general purpose code—as well as perform complex storage and data processing functions—is fundamental in boosting Arm’s support for high-performance computing, AI/machine learning (ML), and a variety of other data-intensive applications, according to the company.
This release also features enhancements to MinIO’s existing implementation of its Reed Solomon erasure coding library, increasing throughput by 2x. Code erasure allows teams to reconstruct missing information communicated over a network through parity blocks, enabling enterprises to withstand server failure, drive failure, and network corruption.
In addition to code erasure, MinIO’s has enhanced its bit-rot detection through improvements to the Highway Hash algorithm. The deterioration of drives—an often silent, corruptive process—dictates that enterprises need a way to identify if any rot has occurred. MinIO has significantly improved the performance of this detection with a mathematical computation—the Highway Hash algorithm—through SVE support of its core hashing update function.
“Delivering performance gains while improving power efficiency is critical to building a sustainable infrastructure for the modern, intensive data processing workloads demanded by AI,” said Eddie Ramirez, vice president of marketing and ecosystem development, infrastructure line of business, Arm. “Performance efficiency is a key factor in why Arm has become pervasive across the data center and in the cloud, powering server SoCs and DPUs, and MinIO’s latest optimizations will further drive Arm-based AI innovation for the world’s leading data enterprises.”
To learn more about MinIO’s enhancements for Arm, please visit https://min.io/.