AI is no longer a distant vision for the future—it’s the backbone of businesses seeking to compete in today’s data-driven economy. From fueling innovation in customer experiences to automating complex processes, organizations are racing to harness the capabilities of artificial intelligence. At Dell Technologies, we understand that AI’s game-changing potential hinges on not just cutting-edge algorithms but also the infrastructure that supports and powers them.
That’s why we’re proud to announce the latest updates to PowerScale, our unstructured data storage platform, designed to help businesses of all scales unlock new possibilities with AI. With this launch, Dell introduces faster networking with 200Gb Ethernet and InfiniBand options, higher storage density with 61TB SSD drives and 24TB HDDs, and enhanced searchability through the new metadataIQ feature.
Here’s a closer look at what the latest OneFS update offers and what it means for the future of AI-enabled enterprises:
Faster Networking for Accelerated AI Workflows
The effectiveness of AI processes depends not just on how powerful a system is, but on how quickly it can access and move data at scale. To eliminate bottlenecks and push the boundaries of AI performance, PowerScale introduces 200GbE and 200Gb InfiniBand connectivity. These upgrades deliver up to 220%¹ streaming write and 99%² streaming read boost over last year, enabling seamless data flow across every stage of the AI pipeline.
What does this mean for AI workloads?
- AI Pipeline Optimization: From data ingestion to model training and inference, ultra-fast, low-latency connectivity ensures that even massive datasets flow smoothly between storage and compute. Reducing data bottlenecks means businesses can now perform more frequent training checkpoints and accelerate model training for large-scale AI implementations like Generative AI applications.
- Enhanced GPU Performance: For organizations deploying GenAI environments, faster networking directly impacts the performance of GPUs by streamlining data delivery. With PowerScale, businesses can process larger datasets in less time, allowing them to achieve greater insights and make quicker, data-driven decisions.
Higher Storage Density for Efficiency and Cost Savings
Staying competitive in AI requires handling massive datasets while managing costs—a challenge that modern businesses know all too well. PowerScale now offers 61TB SSD drives with the all-flash PowerScale F710, a density milestone that delivers compelling benefits for enterprises. Additionally, PowerScale also now offers 24TB HDDs for the A300, A3000, H700, and H7000 nodes to meet TCO-optimized customer demands for demanding hybrid workloads.
Key advantages include:
- Reduced Data Center Costs: The ability to shrink storage footprints by up to 50% leads to substantial savings in energy consumption and colocation expenses. It’s a green initiative that aligns with both business efficiency and sustainability goals.
- Advancing Generative AI Models: The increased density supports larger Large Language Models (LLMs) like those used in Generative AI, including NVIDIA’s DGX SuperPOD. Organizations deploying these models can now do so with fewer nodes, making AI implementations more cost-efficient while yielding deeper insights. For the HDD based platforms, the 24TB drive offers better SSD cache to HDD ratio compared to the 20TB drives – which directly results in higher output and lower latency.
Streamlined Data Management with Enhanced Searchability
Data drives AI, but the sheer volume of unstructured data can make it a challenge to manage and extract meaningful insights. PowerScale’s new MetadataIQ capabilities revolutionize the way enterprises handle metadata, unlocking valuable context across distributed datasets.
With the ability to export metadata into Elasticsearch databases, PowerScale allows organizations to efficiently discover their data and gain insights from billions of files and petabytes of data.
Transformative capabilities include:
- Improved Data Accessibility: Enterprises can quickly discover valuable information hidden within unstructured sources such as PDFs, knowledge bases, and support databases. Faster access means more informed, real-time decision-making across departments.
- Efficient AI Training: The ability to tap into rich metadata across distributed clusters ensures that training datasets are curated, consistent, and contextual. This approach allows businesses to keep their RAG workflows efficient, by eliminating the need to do treewalks and processing only updated or new content – thereby maximizing compute utilization to save time, costs, and derive accuracy in Generative AI environments.
Scale to new AI horizons with PowerScale—are you ready?
Data continues to fuel the evolution of AI, and the right storage system is pivotal to maximizing its potential. With PowerScale’s faster networking, higher storage density, and enhanced searchability, Dell Technologies empowers businesses to turn data into action, insights, and innovation. Whether powering multi-billion-dollar media productions or running critical financial models, these solutions allow companies to focus on outcomes—not infrastructure.
More than just cutting-edge technology, Dell PowerScale is the foundation for innovation in businesses around the globe delivering performance, scalability, flexibility, and simplicity. From scaling large datasets with unmatched flexibility to enabling seamless transitions between on-premises and cloud environments, PowerScale is the driving force behind resilient, future-ready AI deployments.
Curious to see what PowerScale can do for your business? Learn more about our latest updates or visit our website to get started.
1 Based on preliminary internal analysis of streaming write figures comparing F600p running OneFS 9.5 versus F710 running the latest version of OneFS using 200GbE. Dec, 2024
2 Based on preliminary internal analysis comparing F600p running OneFS 9.5 versus F710 running the latest version of OneFS, using 200GbE. Dec, 2024
