iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
NFS can be a practical shared-storage starting point for a small AI deployment, but performance depends on the whole path from dataset layout and storage media to server, network, and GPU client. Size that path for simultaneous reads, measure it with the real workload, and consider RDMA or GPUDirect Storage only when the hardware and software stack support them. There is no single NFS server build or throughput figure that applies to every AI workload.
Is NFS good for AI workloads?
For a smaller GPU server configuration, NFS offers a familiar way to share datasets without beginning with a specialized parallel filesystem. NVIDIA’s DGX storage guidance describes it as a good starting point when storage and network bandwidth are properly sized. That is a conditional recommendation, not a promise that any NFS server will keep every GPU fed.
The same guidance gives a concrete reference: NFS over 10Gb/s Ethernet can provide up to 1.25 GB/s of I/O throughput for large block sizes. This is an upper configuration reference in NVIDIA’s guidance, not a guaranteed result or an independently reproduced benchmark. Real throughput depends on the server, network path, clients, workload, and protocol configuration.
NFS becomes less attractive when a single server or its network interfaces cannot meet aggregate demand, or when many clients create high concurrent load. NVIDIA identifies requirements above 10 GB/s and deployments with hundreds or thousands of nodes as cases where other technologies may be more efficient and scale better. Treat those as decision prompts from that guidance, not universal cutoffs: resilience, operations, workload behavior, and measured capacity matter too.
#1 Best Overall
- Entry-level NAS Personal Storage：UGREEN NAS DH2300 is your first and best NAS made easy. It is designed for beginners who want a simple, private way to store videos, photos and personal files, which is intuitive for users moving from cloud storage or external drives and move away from scattered date across devices. This entry-level NAS 2-bay perfect for personal entertainment, photo storage, and easy data backup (doesn't support Docker or virtual machines).
- Set Your Devices Free, Expand Your Digital World: This unified storage hub supports massive capacity up to 64TB.*Storage drives not included. Stop Deleting, Start Storing. You can store 22 million 3MB images, or 2 million 30MB songs, or 43K 1.5GB movies or 67 million 1MB documents! UGREEN NAS is a better way to free up storage across all your devices such as phones, computers, tablets and also does automatic backups across devices regardless of the operating system—Window, iOS, Android or macOS.
- The Smarter Long-term Way to Store: Unlike cloud storage with recurring monthly fees, a UGREEN NAS enclosure requires only a one-time purchase for long-term use. For example, you only need to pay $459.98 for a NAS, while for cloud storage, you need to pay $719.88 per year, $2,159.64 for 3 years, $3,599.40 for 5 years. You will save $6,738.82 over 10 years with UGREEN NAS! *NAS cost based on DH2300 + 12TB HDD; cloud cost based on 12TB plan (e.g. $59.99/month).
- Blazing Speed, Minimal Power: Equipped with a high-performance processor, 1GbE port, and 4GB RAM on Board, this NAS handles multiple tasks with ease. File transfers reach up to 125MB/s—a 1GB file takes only 8 seconds. Don't let slow clouds hold you back; they often need over 100 seconds for the same task. The difference is clear.
- Let AI Better Organize Your Memories: UGREEN NAS uses AI to tag faces, locations, texts, and objects—so you can effortlessly find any photo by searching for who or what's in it in seconds. It also automatically finds and deletes similar or duplicate photo, backs up live photos and allows you to share them with your friends or family with just one tap. Everything stays effortlessly organized, powered by intelligent tagging and recognition.
How do I size the storage server and network?
Start with the demand at the GPUs rather than the advertised speed of one network link. Estimate how much data each GPU must read, how many GPUs and clients will read concurrently, whether access is mostly large sequential I/O or small-file and metadata-heavy activity, and how much data is already cached. Then compare that demand with every shared part of the path.
Map the complete data path
- Dataset and application: Identify file sizes, access pattern, read concurrency, and whether the training pipeline repeatedly revisits data.
- Storage media: Check whether the storage devices can sustain the requested reads, including when multiple clients access them at once.
- Server: Account for storage interfaces, server network interfaces, CPU and memory, and the ability of the NFS service to handle concurrent requests.
- Network: Check the client link, server link, switches, and every hop between them. Aggregate client demand must fit the capacity of shared links, not just the speed of an individual port.
- Client and GPU path: Check client configuration and, where applicable, the NIC-to-GPU topology and supported transfer path.
A 10Gb/s Ethernet NFS reference of up to 1.25 GB/s applies to large blocks in NVIDIA’s DGX storage guidance; it does not tell you how many GPUs a given server can serve. If measured throughput is substantially below the reference, NVIDIA advises checking for a constrained network hop, server misconfiguration, or a smaller MTU somewhere along the path. Validate the end-to-end path rather than changing one setting in isolation.
Rank #2
- 【Advanced Home Data & Media Hub】For advanced home users who need phone backup, file storage, and centralized data management. Centralize family photos, 4K videos, movies, computer backups, and personal files in one place while running multiple apps for home entertainment and everyday data management. Suitable for households with growing digital libraries and multiple NAS use cases.
- 【Built for Creators, Media Servers & Advanced Apps】Powered by the Intel N100 Quad-Core CPU, 8GB DDR5 RAM, 2.5GbE networking, and dual M.2 NVMe slots, DXP2800 handles large files and heavier workloads with ease. Run Docker, virtual machines, and media server applications compatible with Plex—ideal for content creators, tech enthusiasts, and advanced home users managing 4K videos, RAW photos, personal media libraries, and multiple NAS apps.
- 【Up to 80TB for Growing Digital Libraries】 Supports up to 80TB of storage using two HDD bays and two M.2 NVMe SSD slots for family photos, movies, RAW photos, 4K videos, work files, and device backups. AI photo management supports recognition of people, objects, scenes, and locations, album organization, and duplicate photo detection. HDDs and SSDs are not included.
- 【AI-powered Home Surveillance】Turn DXP2800 into a centralized home surveillance hub by connecting compatible network cameras and storing recordings locally on your NAS. AI-powered features include Face Recognition, People Detection, and Pet Detection, helping advanced home users review important events more efficiently while managing home surveillance and personal data in one place.
- 【One data Center Across Your Devices】Keep files from desktops, laptops, phones, tablets, and other devices together instead of scattered across cloud accounts and external drives. Access, back up, organize, and share data across Windows, macOS, Android, iOS, web browsers, and compatible smart TVs—ideal for creators and advanced home users working across multiple devices.
Use workload estimates cautiously
An older NVIDIA DGX best-practices document, DG-08868-001_v06, gives a planning estimate of 150–200 MB/s per GPU for 1080p files and says 4K or uncompressed data may require more. It also states that uncompressed images are 5–10 times larger than compressed images. These are source-specific planning figures from an older document, not universal requirements: image dimensions, encoding, preprocessing, batching, and application behavior change the demand.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
How can dataset layout affect NFS performance?
A workload that opens and reads many small files can spend substantial time on filesystem metadata and request overhead rather than transferring large blocks. NVIDIA’s DGX best-practices guidance warns that many small-file reads and writes can reduce performance on network and local filesystems.
Rank #3
- Your Personal Streaming Server - Build your own Netflix-style media library and stream 4K movies, shows and photos to any device without monthly fees
- Create Your Own Cloud - Store your entire photo, video and music collection; access from anywhere with fast 282 MB/s transfer speeds
- Creator-Grade Backup Solution - Protect your irreplaceable content with automated backups to cloud services, external drives and remote NAS
- Multi-Layered Data Protection - Combine RAID redundancy, automated backups and snapshot technology to prevent data loss from any cause
- Smart Home Surveillance - Support up to 30 IP cameras with AI detection, instant alerts and secure remote monitoring
Packaging samples into a database or archive can reduce metadata activity. The same guidance names HDF5, LMDB, and TFRecord as formats used for this purpose, but a format change is not free: memory overhead, access patterns, framework compatibility, and fast mmap() performance may matter. Benchmark the actual training pipeline before converting a dataset, and include startup, preprocessing, and random-access behavior—not just a sequential read test.
Can GPUDirect Storage use NFS?
Yes, in configurations supported by NVIDIA’s GPUDirect Storage (GDS) software and hardware stack. NVIDIA’s GDS guidance describes NFS using RDMA transport in its documented configuration. The intended benefit is a more direct data path toward GPU memory, reducing the CPU bounce-buffer path; it does not make storage, the network, or the GPU-side path unlimited.
Rank #4
- Entry-level NAS Home Storage: The UGREEN NAS DH4300 Plus is an entry-level 4-bay NAS that's ideal for home media and vast private storage you can access from anywhere and also supports Docker but not virtual machines. You can record, store, share happy moment with your families and friends, which is intuitive for users moving from cloud storage, or external drives to create your own private cloud, access files from any device.
- Smart Photo Backup & AI Album: Automatically back up photos and videos from your phone in real time and keep growing family memories organized with AI-powered photo albums. Semantic search, custom learning, and recognition of people, objects, pets, and similar photos help you quickly find the moments you want. Duplicate photo removal also helps keep your library organized—ideal for families and users with large photo collections.
- User-Friendly App & Easy Setup: Connect quickly via NFC, set up simply and share files fast on Windows, macOS, Android, iOS, web browsers, and smart TVs. You can access data remotely from any of your mixed devices. What's more, UGREEN NAS enclosure comes with beginner-friendly user manual and video instructions to ensure you can easily take full advantage of its features.
- More Cost-effective Storage Solution: Unlike cloud storage with recurring monthly fees, A UGREEN NAS enclosure requires only a one-time purchase for long-term use. For example, you only need to pay $629.99 for a NAS, while for cloud storage, you need to pay $719.88 per year, $1,439.76 for 2 years, $2,159.64 for 3 years, $7,198.80 for 10 years. You will save $6,568.81 over 10 years with UGREEN NAS! *NAS cost based on DH4300 Plus + 12TB HDD; cloud cost based on 12TB plan (e.g. $59.99/month).
- Your Data, You Control：No third-party clouds, no hidden access, UGREEN NAS provides a more secure and private data storage solution. It stores data locally on your private hard drives and does automatic backups. Thus, you can keep full control over it. The advanced encryption is TRUSTe certified in the United States and is awarded the first (and only) ETSI EN 303 645 certification mark for NAS products by TÜV SÜD Group.
Support depends on the system, hardware topology, software, and workload. NVIDIA’s documentation describes RDMA traffic across InfiniBand interfaces and DMA transfers between network cards and GPUs, and emphasizes that PCIe topology affects configuration. Bandwidth, protocol overhead, MTU, NFS attributes, software stack, and workload characteristics can all affect results. NVIDIA AI Enterprise documentation also describes GDS for supported configurations with local or remote storage, including NFS.
Recommended Free Tools
RDMA and GDS are therefore options to validate, not checkboxes that guarantee faster training. Confirm the platform’s supported configuration and topology against current vendor guidance. If the GPU is waiting on preprocessing, metadata, or another constrained component, shortening one part of the data path may not improve end-to-end training throughput.
Best Value
- Value NAS with RAID for centralized storage and backup for all your devices. Check out the LS 700 for enhanced features, cloud capabilities, macOS 26, and up to 7x faster performance than the LS 200.
- Connect the LinkStation to your router and enjoy shared network storage for your devices. The NAS is compatible with Windows and macOS*, and Buffalo's US-based support is on-hand 24/7 for installation walkthroughs. *Only for macOS 15 (Sequoia) and earlier. For macOS 26, check out our LS 700 series.
- Subscription-Free Personal Cloud – Store, back up, and manage all your videos, music, and photos and access them anytime without paying any monthly fees.
- Storage Purpose-Built for Data Security – A NAS designed to keep your data safe, the LS200 features a closed system to reduce vulnerabilities from 3rd party apps and SSL encryption for secure file transfers.
- Back Up Multiple Computers & Devices – NAS Navigator management utility and PC backup software included. NAS Navigator 2 for macOS 15 and earlier. You can set up automated backups of data on your computers.
What does setting up NFS over RDMA involve?
Linux kernel NFS documentation covers client and server support for NFS/RDMA, including exports, starting the RDMA server transport, mounting with the RDMA option, and checking the negotiated mount protocol. The setup is dependent on the distribution, kernel, NFS utilities, RDMA stack, and vendor support.
A NetApp- and Open Grid Computing-credited NFS/RDMA setup document dated May 29, 2008 warns that it may be obsolete. Its historical performance results are not a current benchmark. Use current documentation for your Linux distribution, kernel, NFS utilities, and hardware rather than copying old commands or assuming a legacy example applies. Verify that both server and client are configured for the intended transport and that the mounted connection actually uses it.
How should I validate an AI storage path?
- Characterize the training workload. Record data volume, file-size distribution, access pattern, client and GPU count, and whether data is cached. Include the application’s real concurrency.
- Establish a baseline. Measure the workload as it runs, including throughput, latency, I/O sizes, GPU activity, and storage and network utilization. A high aggregate rate alone can hide latency or starvation at individual clients.
- Find the limiting resource. Compare observed demand and utilization across storage media, server interfaces, network hops, NFS configuration, and the client-to-GPU path. Check for mismatched link capacity, topology, MTU, or protocol settings.
- Change one layer at a time. For example, test a dataset layout change separately from a network or NFS transport change. Keep workload and concurrency comparable so the result is interpretable.
- Re-test at target scale. A single-client result does not establish aggregate performance. Measure with the expected number of concurrent clients and confirm that per-client behavior remains acceptable.
When should I move from NFS to a parallel filesystem?
Consider another architecture when measured aggregate demand exceeds what a single NFS server and its network path can serve, when concurrent clients create unacceptable contention, or when resilience and operational needs are not met by the current design. A distributed or parallel filesystem can distribute work across multiple servers, but it adds its own compatibility, deployment, and operations requirements; it is not automatically faster for every workload.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problems| Approach | Where it can fit | What to validate |
|---|---|---|
| NFS on a single server | Smaller deployments where a straightforward shared filesystem meets measured aggregate demand. | Storage and server capacity, network path, concurrency, and small-file behavior. |
| NFS with RDMA or GDS in a supported configuration | Deployments where the system supports the relevant transport and GPU data path, and measurements identify data movement as a constraint. | Supported hardware and software, NIC/GPU topology, protocol configuration, workload, and end-to-end results. |
| Distributed or parallel filesystem | Deployments whose aggregate load, client count, or resilience requirements call for a different scale-out model. | Aggregate and per-client throughput, latency, fault tolerance, compatibility, operational effort, and cost. |
NVIDIA’s DGX storage guidance names EXAScaler, GRIDScaler, Ceph, Lustre, MapR-FS, GPFS, Weka.io, and Gluster as examples of alternative distributed filesystem technologies. The names are examples in that guidance, not a current endorsement or a statement that each is suitable for a particular deployment. Verify present product status, capabilities, support, and compatibility before evaluating any option.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
