WEKA Announces New NeuralMesh Architecture Built for NVIDIA BlueField-4
WASHINGTON and CAMPBELL, Calif., October 28, 2025
From GTC Washington, D.C. 2025: WEKA today announced that it is developing the next generation of NeuralMesh™ by WEKA®, its intelligent storage system, for the newly announced NVIDIA BlueField-4 data processing unit (DPU), marking a transformational shift in how AI infrastructure will be architected and deployed. This groundbreaking approach eliminates the need for standalone CPU servers, instead leveraging the unprecedented 800 Gb/s networking bandwidth and 6x compute improvement of NVIDIA BlueField-4 to deliver what WEKA CEO Liran Zvibel hails as “the perfect foundation for the future of AI data infrastructure.”
Reimagining AI Infrastructure Architecture
Rather than deploying NeuralMesh on traditional servers with separate CPUs, the new generation of WEKA's NeuralMesh storage software will run directly on NVIDIA BlueField-4, fundamentally simplifying AI infrastructure deployment while dramatically improving economics, performance, and power efficiency. This architectural shift represents WEKA's most substantial alignment yet with NVIDIA's infrastructure roadmap and vision for the AI data center.
The Perfect Complement to Modern AI Infrastructure
NVIDIA BlueField-4 brings programmable infrastructure, acceleration, zero-trust security, and massive networking capabilities that align seamlessly with NeuralMesh's microservices-based mesh architecture. When paired with NVIDIA's latest accelerated compute platforms, BlueField-4 and NeuralMesh create a unified, efficient foundation for gigascale AI factories.
“The rapid evolution of AI demands intelligent data infrastructure that can adapt and scale with unprecedented efficiency,” said Liran Zvibel, co-founder and CEO at WEKA. “WEKA's NeuralMesh architecture, built for NVIDIA BlueField-4, is a tremendous leap forward, providing the foundational storage for next-generation AI factories.”
“The demands of AI reasoning and other inference workloads require a new foundation for data infrastructure – one that unifies compute, network, and storage for unprecedented efficiency and scale,” said Justin Boitano, vice president, Enterprise Products, NVIDIA. “With NVIDIA BlueField-4 integration, WEKA’s NeuralMesh storage system supports this architectural transformation to power efficient data processing in gigascale AI factories.”
Benefits of the BlueField-4-Powered Architecture
WEKA's next-generation NeuralMesh storage system, built on BlueField-4, will deliver:
- Exceptional Power Efficiency: Delivering over 100x tokens/watt improvement for agentic AI workflows
- Simplified Infrastructure: Eliminates the need for dedicated storage servers with standalone CPUs, reducing complexity and footprint.
- Superior Economics: Dramatically lowers infrastructure costs by consolidating compute, networking, and storage functions.
- Enhanced Performance: Leverages BlueField-4's 800 Gb/s networking and 6x compute improvement for unmatched data movement speeds.
- Seamless Integration: Aligns with NVIDIA's infrastructure roadmap and AI factory architectures.
- Zero-Trust Security: Built-in security acceleration and tenant isolation at line rate through the NVIDIA DOCA™ software framework.
A Collaboration for the Age of Reasoning
WEKA and NVIDIA have collaborated closely for close to a decade, spanning NVIDIA AI Enterprise software, WEKA’s achievement of multiple NVIDIA certifications, and collaboration on the introduction of new NVIDIA technologies. NeuralMesh storage software built on BlueField-4 represents WEKA’s deepest integration of NVIDIA technologies yet.
“Power efficiency is key to bringing gigascale AI factories online. NVIDIA BlueField-4 and WEKA's NeuralMesh storage software will directly address this challenge, delivering breakthrough tokens-per-watt efficiency without compromising performance, scale, or operational simplicity. We view NeuralMesh’s integration with BlueField-4 as a strategic growth engine that can simplify and accelerate Enterprise AI factory deployments for our customers,” said Nilesh Patel, chief strategy officer at WEKA.
Availability and Next Steps
WEKA is actively developing its next-generation NeuralMesh software release for NVIDIA BlueField-4, with detailed roadmap information to be shared with customers and partners in early 2026. Organizations interested in learning more about WEKA's NeuralMesh architecture can contact WEKA directly or visit www.weka.io.
About WEKA
WEKA is the AI data and memory infrastructure company transforming the economics of agentic AI. Its NeuralMesh™ platform unifies high-performance data storage with extended GPU memory, giving enterprises, AI cloud providers, and AI builders a single foundation for training, inference, and agentic workloads. With Augmented Memory Grid, NeuralMesh extends GPU memory capacity by 1000x, accelerates time to first token by up to 20x, and delivers 10x more concurrent users from the same GPU footprint, proven in production benchmarks. Trusted by 30% of the Fortune 50, WEKA enables organizations to scale AI faster, optimize GPU utilization, and reduce the cost of every token served. Learn more at www.weka.io or connect with us on LinkedIn and X.
WEKA and the W logo are registered trademarks of WekaIO, Inc. Other trade names herein may be trademarks of their respective owners.
Media Contact:
WEKA Communications
pr@weka.io
What's Next
WEKA and Oracle Cloud Infrastructure Validate 10x Throughput Gains for Long-Context AI Inference
WEKA Partners With Glocomp to Deliver End-to-End AI Infrastructure Solutions Across Malaysia
WEKA Breaks The AI Memory Barrier With Augmented Memory Grid on NeuralMesh
WEKA Unveils Next-Gen WEKApod Appliances to Redefine AI Storage Economics
Scale Production AI Faster with NeuralMesh
Your models aren't slow. Your data is. Fix AI bottlenecks with high-throughput infrastructure.