Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
LeoFS.info presents a parallel file system / data storage solution designed for AI workloads and high-performance computing. The website highlights “Exceptional data storage solutions for AI workloads” and cites market figures such as a maximum single-cluster capacity of over 95PB and more than 5EB of cumulative deployed customer capacity. It is not positioned as general-purpose office SaaS, but rather for PB-scale file storage scenarios such as research, AI, media production, and satellite imagery.
Based on the available text, LeoFS is centered on a high-throughput, horizontally scalable parallel file system. It supports non-disruptive expansion, N+M file-level erasure coding, automated self-monitoring and self-healing, and emphasizes the absence of single points of failure across disks, nodes, or networks. On the performance side, it mentions RDMA, cross-protocol data access, space optimization, AI small-file aggregation, and cluster deployment on commodity hardware to reduce hardware vendor lock-in. The published examples include a 1.4PB four-node cluster, a 60PB on-premises customer cluster, and comparisons with solutions such as BeeGFS, EMC Isilon, and DDN.
Pricing is relatively straightforward: under a Software as a Service model, HDD/NVMe hybrid deployments cost $5,000 per node per year, while all-NVMe deployments cost $10,000 per node per year. First-time customers replacing Lustre, IBM GPFS, or BeeGFS can apply for a 1PB, 90-day trial. The website also mentions optional hardware bundles, commodity hardware, and transparent total cost of ownership estimates, but does not disclose payment methods, contract terms, or SLA details.
Its strengths are a clear focus on ultra-large-scale file storage and high-throughput AI/HPC workloads. A commodity-hardware-based approach can help control costs, while public pricing and case data make evaluation easier. 24/7 support and direct access to proprietary code developers are valuable for enterprise-scale deployments. The main limitations are that the public materials lack enterprise software details such as access control, auditing, encryption, compliance certifications, APIs/SDKs, and a management console, so technical due diligence and a PoC are necessary before procurement.
LeoFS is better suited to university research centers, AI infrastructure teams, HPC cluster operations, and imaging/media data platforms. It is not designed for lightweight team file collaboration. Access from China is not described in the available text, so it should be considered unknown; payment methods are also not disclosed. For domestic deployment in China, it can be compared with Lustre, BeeGFS, IBM GPFS/Spectrum Scale, DDN, Dell EMC Isilon, and local distributed storage solutions.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on leofs.info official site.
leofs.info is an Unknown VPS provider. TG4G tracks its product information, an overall rating of 6.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach leofs.info directly.