In today’s AI-driven world, organizations face a growing challenge: managing and processing massive amounts of data efficiently while keeping power consumption in check. Lenovo and Cloudian have stepped up to address this challenge with a HyperStore-powered AI data lake platform, now generally available.
Breaking Performance Barriers with Power Efficiency
The solution combines Cloudian’s proven data platform software with Lenovo ThinkSystem SR635 V3 all-flash servers, powered by AMD EPYC 9454P processors. The results are impressive: testing showed the six-node cluster achieving 28.7 GB/s reads and 18.4 GB/s writes. But what’s truly remarkable is the 74% improvement in power efficiency compared to traditional HDD-based systems.
This breakthrough in efficiency couldn’t come at a better time. With Morgan Stanley forecasting a 70% annual increase in power consumption for generative AI through 2027, organizations need solutions that can deliver high performance without sending energy costs through the roof.
Purpose-Built for AI and Analytics
The platform has been validated with popular AI and data analytics tools including PyTorch, TensorFlow, Kafka, and Druid, making it ideal for data-intensive industries such as:
- Media and entertainment
- Financial services
- Life sciences
- Healthcare
- Research institutions
Feature Rich
This AI ready data lake offers capabilities that make it ideal for your most demanding storage workloads:
- S3-Compatible and Hybrid/Multi-Cloud Ready: Cloudian’s native S3 API implementation offers the industry’s highest level of S3 interoperability, letting you capitalize on the rapidly growing ecosystem of S3-enabled applications. It’s also compatible with public cloud platforms including AWS, Google, and Microsoft. Integrated data management tools make it easy to integrate your on prem and cloud storage environments.
- Uncompromising Data Durability: Get up to 14 nines data durability and configurable data protection. Only Lenovo Object Storage offers granular management that lets you customize data protection for various data types within a cluster. Select from multiple levels of data replication and erasure coding.
- Object Lock for Immutable: Data Object Lock prevents object version deletion during a user-defined retention period. Immutable S3 objects are protected using object- or bucket-level configuration of WORM and retention attributes. The retention policy is defined using the S3 API or bucket-level defaults. Objects are locked for the duration of the retention period, and legal hold scenarios are also supported. This functionality provides data protection from accidental or malicious deletion as well as ransomware. Regulatory compliance with SEC Rule 17a-4(f), FINRA Rule 4511, CFTC Regulation 1.31 is validated in an independent assessment.
- Integrated Multi-Site Capability: Extend your storage with the industry’s most flexible geo-distribution architecture. Replicate data across regions to provide local access and resilience in the event of a data center failure. Configurable consistency levels let you select either synchronous or asynchronous replication. Or employ distributed erasure coding to stripe data across sites as a highly efficient disaster recovery option.
- Scalable Performance: Process multiple requests in parallel, across all nodes in Lenovo Object Storage’s shared-nothing cluster. This eliminates bottlenecks and allows performance to grow as you add nodes. For large file transfers, multi-part upload lets you transfer files seamlessly between sites or to the public cloud.
- Security: Lenovo Object Storage supports KMIP standardized key management as a Server-Side Encryption (SSE) option. SSE supports AES-256 and FIPS-validated crypto-graphic modules, allowing enterprises and service providers to easily and securely encrypt data stored at rest. SSL encryption ensures data confidentiality for data in transit (HTTPS). And with S3-compatible ACLs, SAML, MFA, IAM, and secure trust policies, system administrators can better manage access to buckets and objects.
- Hybrid Storage Policies: Objects are automatically assigned the most appropriate storage policy associated with the bucket; where smaller-size objects utilize a replica factor (RF) policy for optimal performance and larger objects utilize an erasure coding (EC) policy for best capacity efficiency.
- Multi-Tenancy: Establish multiple user spaces with Lenovo Object Storage’s multi-tenancy capabilities. Provide individual management domains and namespaces within a shared cluster. Cloudian offers QoS controls and billing features that let you manage bandwidth and chargebacks.
- Intuitive Management: Manage Cloudian via the intuitive GUI or automate tasks using the RESTful API. Cluster management tasks are simple as well. Add or remove nodes and update system software without disruption.
- Variable-size Metadata Tags: Record metadata along with user data to facilitate data management and data search. Elastic Search integration provides near real-time search capability. Variable metadata tag sizes give you flexibility not found in other object storage systems.
A Powerful Partnership
“Lenovo’s industry-leading servers with AMD EPYC processors perfectly complement Cloudian’s high-performance data platform software,” notes Michael Tso, CEO and co-founder at Cloudian. “Together, they deliver the limitlessly scalable, performant, and efficient foundation that AI and data analytics workloads require.”
The partnership leverages each company’s strengths: Lenovo’s expertise in high-performance computing infrastructure, AMD’s cutting-edge processor technology, and Cloudian’s proven track record in enterprise-scale data management. With over 800 enterprise deployments worldwide, the platform offers organizations a secure way to develop proprietary AI models while maintaining complete control over their data.
Available Now
The combined Lenovo/AMD/Cloudian solution is available immediately through Lenovo and authorized resellers, offering organizations a clear path to accelerate their AI initiatives while maintaining efficiency and security. For enterprises looking to innovate or drive research and discovery with AI, ML, and HPC, this solution represents a significant step forward in data management technology.
To learn more, contact Cloudian here, or your Lenovo representative, or visit www.lenovo.com.
Read about Cloudian at www.cloudian.com