FMS 2026: MaxLinear Showcases Panther for AI Storage Efficiency and AI Inference Performance

MaxLinear, Inc., a provider of high‑performance connectivity and data movement solutions, announced it will showcase its Panther Storage Accelerator platform at Future of Memory and Storage (FMS) 2026, held August 4–6 in Santa Clara, California.As AI evolves from model-centric deployments to agentic AI architectures, infrastructure requirements are shifting from model performance to agent scalability. AI clusters are increasingly required to support millions of concurrent agents, driving exponential growth in memory capacity and bandwidth requirements. As agent populations grow, HBM and DDR resources become constrained, increasing memory pressure and reducing GPU and CPU efficiency. Panther addresses these challenges through hardware-accelerated compression, delivering up to a 3× compression ratio on AI data. This increased efficiency can enable up to 2.2× greater effective KV cache capacity, helping systems support up to 2.2× more AI agents within existing infrastructure.

“As AI adoption accelerates, organizations are rethinking how they store, move, and manage data at scale,” said Vikas Choudhary, SVP and GM, connectivity and storage, MaxLinear. “Panther helps address these challenges by enabling more efficient use of storage and compute resources, allowing customers to increase effective storage capacity for AI data while improving overall infrastructure efficiency. We’re excited to showcase these capabilities at FMS and demonstrate how Panther can help organizations scale next-generation AI workloads more efficiently.”

Panther combines hardware-accelerated compression, encryption, erasure coding, data integrity services, and high-speed data movement in a single accelerator platform. By offloading these data-intensive functions from host CPUs, Panther helps organizations increase effective storage capacity, improve infrastructure utilization, reduce power consumption, and accelerate AI and storage workloads while lowering overall infrastructure costs.

Panther Technology Showcases at FMS 2026

• Live AI Workload Compression Demo — See Panther perform hardware-accelerated compression and decompression on AI datasets, reducing CPU overhead while maintaining the throughput required for modern AI environments
• Live Peer-to-Peer DMA Demo — Experience direct PCIe endpoint-to-endpoint data movement enabled by Panther’s peer-to-peer DMA architecture, reducing host CPU involvement and memory bandwidth consumption
• OpenZFS Acceleration Showcase — Learn how Panther integrates with OpenZFS through the ZIA and DPUSM frameworks to enhance storage performance and infrastructure efficiency
• SPDK Compression Offload Showcase — Explore Panther integration with the SPDK framework to accelerate compression and optimize storage and AI data pipelines

Built for AI Storage Infrastructure
Designed for AI and data-intensive workloads, Panther helps organizations maximize storage and compute resources while improving infrastructure efficiency. FMS showcases will highlight how Panther can enable up to 2.5× greater effective storage capacity for AI data while freeing CPU resources for higher-value workloads across enterprise, cloud, hyperscale, and high-performance computing environments.

Visit MaxLinear at FMS 2026
Attendees are invited to visit MaxLinear at Booth #513 during Future of Memory and Storage 2026 to experience Panther through live demonstrations and technology showcases focused on AI data movement, compression acceleration, OpenZFS integration, and peer-to-peer DMA.