During GTC 2024, WekaIO unveiled the WEKApod, an advanced data platform appliance optimized for NVIDIA DGX SuperPOD with NVIDIA DGX H100 systems. The appliance merges WEKA’s specialized AI data platform software with leading-edge storage technology to create a turnkey environment for AI workloads.
During GTC 2024, WekaIO unveiled the WEKApod, an advanced data platform appliance optimized for NVIDIA DGX SuperPOD with NVIDIA DGX H100 systems. The appliance merges WEKA’s specialized AI data platform software with leading-edge storage technology to create a turnkey environment for AI workloads. Users can integrate WEKApod into their NVIDIA DGX SuperPOD ecosystems, allowing for swift deployment of computational, storage, and network resources, thus accelerating AI project deployment.
The WEKApod Data Platform Appliance sets a high bar in data management for NVIDIA DGX SuperPOD deployments. It is capable of delivering up to 18.3 million IOPs and efficient write performance vital for AI model development. This supports enhanced training productivity, scalability, resilience, and real-time application support.
WEKApod packages WEKA’s Data Platform software, a full range of features in an appliance suitable for AI-focused enterprise environments, generative AI, and GPU cloud infrastructures.
Key Advantages of WEKApod include:
- Leading-Edge AI Storage Performance: WEKA’s AI-native architecture provides top-tier AI storage, ensuring minimal latency for all file sizes and substantial write throughput to maintain critical operations continuity.
- Energy-Efficient Sustainability: The platform offers enhanced performance density, reducing energy consumption through optimal use of space, improved cooling, power efficiency, and reduced idle power use, with potential carbon footprint reductions of up to 260 tons of CO2 equivalent per petabyte stored.
- Optimized Storage Access: It ensures high efficiency in storage access, improving performance for both training and inference tasks, minimizing idle times, and lowering overall energy use. This contributes to sustainability efforts by lowering both energy costs and carbon footprints in AI initiatives.
WEKApod offers versatile deployment options for on-premises and hyperscaler, hybrid, and GPU cloud services. Since 2013, WekaIO has been at the forefront of the software data platform space for AI and high-performance computing and was the first to offer subscription-based software that operates on leading server hardware and across major public cloud providers. WEKApod represents an extension of this offering, enabling on-premises and large-scale AI and GPU cloud deployments of the WEKA Data Platform software.
Engage with StorageReview
Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed
