{"slug": "truenas-head-honcho-talks", "title": "TrueNAS head honcho talks", "summary": "TrueNAS CEO Brett Davis said the company is expanding tiering capabilities to help customers optimize storage costs amid memory and SSD price increases driven by AI demand. Davis noted that TrueNAS is collaborating on open pNFS solutions for HPC and GPU clusters, and that AI workloads requiring RDMA, higher bandwidths, and Kubernetes integration are growing dramatically among customers.", "body_md": "# TrueNAS head honcho talks\n\n[TrueNAS](https://www.blocksandfiles.com/file/2024/04/08/ixsystems-no-one-is-being-marooned-by-debian-focus/1609738), the trading name for iXsystems, and the product name of the most widely used and open-source network-attached storage (NAS) worldwide, is headed up by CEO Brett Davis. It claims more than 1 million users, deployments by more than 60 percent of the Fortune 500, and more than 15 million cumulative downloads since it was started in 2002.\n\nOur understanding is that TrueNAS focuses on standard NFS sharing with good performance via tuning, such as using NFS over RDMA on enterprise hardware. There is no built-in parallel NFS ([pNFS](https://www.blocksandfiles.com/glossary/2022/05/05/pnfs/1586765)) support and neither is there a partner relationship with Nvidia that enterprise market file system suppliers, such as IBM, NetApp and WEKA have.\n\nWe see the huge wave of AI rushing across the file landscape but not in all directions equally. It is bringing a need for faster file access, plus supply shortages and prices increases for memory and SSDs; so-called Memflation, in its wake. We asked Brett Davis some questions to find out the company’s attitude towards this, parallel NFS, AI, and the future for TrueNAS.\n\n**Blocks & Files: How can TrueNAS help with Memflation through tiering from SSDs to HDDs and, perhaps, onward to the public cloud or tape?**\n\n**Brett Davis: **Our hybrid-flash (disk+flash) solutions have increased significantly in popularity in 2026. TrueNAS is expanding its tiering capabilities to help customers optimize storage costs across flash and disk. Unlike traditional storage vendors where you pay for features upfront and see diminishing value over time, TrueNAS delivers continuous improvements that maximize the value of your existing hardware investment. We're shipping major enhancements throughout this year and next that will unlock more performance and efficiency from the systems customers already own, providing seamless upgrade paths as the platform evolves.\n\nIn terms of public cloud, we offer customers the capability to offload data to it from TrueNAS, when needed, but it's not really worthwhile as a safe harbor from Memflation. Cloud costs were already far higher, and they continue to rise as well. On premises TrueNAS was far more cost-efficient pre-Memflation crisis, and we don't see that changing. As for tape, it certainly becomes more interesting in this market.\n\n**Blocks & Files: What is your view of pNFS and its relevance to TrueNAS customers?**\n\n**Brett Davis: **pNFS is a better solution for HPC and GPU clusters than Lustre or proprietary file systems like [VAST](https://www.blocksandfiles.com/ai-ml/2026/06/16/vast-data-and-the-evolving-neoclouds/5256380). TrueNAS is collaborating on open pNFS solutions with open clients, storage, and metadata.\n\n**Blocks & Files: How is AI training and, separately, AI inferencing affecting TrueNAS customers? Does the product roadmap now contain developments pertaining to AI training and inferencing?**\n\n**Brett Davis: **Our customers' AI workloads are growing dramatically. Their workloads require RDMA, much higher bandwidths and Kubernetes integration. TrueNAS has been delivering these to our customers. For example, a major AI inference chip company* was based on TrueNAS storage. In the future, these workloads will also need tiered storage for cost reduction.\n\n**Blocks & Files: How do you see the future of TrueNAS?**\n\n**Brett Davis: **We're seeing growth in AI, M&E, backup and a dramatic growth in new virtualization infrastructures due to VMware/Broadcom. Our newer platforms (H-Series, V-Series) enable NVMe or HDDs in any bay, for the ultimate flexibility. Customers consistently praise the flexibility and value that TrueNAS provides, with the same software running all-flash, hybrid-flash, and disk storage infrastructure. But I see the future of TrueNAS as much bigger than storage software or storage appliances alone.\n\nThe market is changing. Organizations are generating more data than ever, and AI is making more of that data valuable again. The cost of traditional enterprise storage was already becoming harder to justify in a normal market. Customers didn't want to be locked into rigid architectures, proprietary ecosystems, or cloud models that become more expensive as they scale. The Memflation crisis has only made that problem acutely obvious to more organizations.\n\nTrueNAS is well positioned for this world because we sit at the intersection of three things customers increasingly care about: enterprise reliability, open innovation, and better storage economics.\n\nOur goal is to become the most trusted name in data storage by giving customers a platform that is powerful enough for serious enterprise workloads, flexible enough to adapt across flash, hybrid, backup, virtualization, media, research, and AI use cases, and efficient enough to challenge the cost structure of legacy storage vendors.\n\nThe community remains a major part of that future. TrueNAS has the largest storage community in the world, and that gives us a different kind of product engine. We are not building in isolation. We have a global base of users constantly testing, deploying, learning, and pushing the platform forward. The opportunity now is to preserve what made TrueNAS trusted by the community while continuing to mature the enterprise experience around support, hardware, security, scale, lifecycle management, and customer success.\n\n##### Comment\n\nThe major AI inference chip company could be Cerebras but TrueNAS is not mentioned anywhere in the Cerebras Training [documentation](https://training-docs.cerebras.ai/rel-2.5.0/getting-started/pre-train-your-first-model#model-config). We can find no public evidence shows that AI Inference suppliers Groq, SambaNova, or Tenstorrent officially use or reference TrueNAS.", "url": "https://wpnews.pro/news/truenas-head-honcho-talks", "canonical_source": "https://www.blocksandfiles.com/file/2026/06/26/truenas-head-honcho-talks/5262880", "published_at": "2026-06-26 13:27:19+00:00", "updated_at": "2026-06-26 13:34:12.587625+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-research", "ai-products"], "entities": ["TrueNAS", "iXsystems", "Brett Davis", "Nvidia", "IBM", "NetApp", "WEKA", "VAST"], "alternates": {"html": "https://wpnews.pro/news/truenas-head-honcho-talks", "markdown": "https://wpnews.pro/news/truenas-head-honcho-talks.md", "text": "https://wpnews.pro/news/truenas-head-honcho-talks.txt", "jsonld": "https://wpnews.pro/news/truenas-head-honcho-talks.jsonld"}}