# NVMe Published articles for NVMe. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Lenovo ThinkAgile VX850 V4 Takes On In-Memory Virtualization With New Deploy Tiers and Four-Hypervisor Express Bundles DevFeed: [Lenovo ThinkAgile VX850 V4 Takes On In-Memory Virtualization With New Deploy Tiers and Four-Hypervisor Express Bundles]() Original publisher: [Read original article]() Author: Harold Fritts Published: 2026-09-17T16:43:22Z Content type: news Language: en Sources: [StorageReview.com]() Topics: [virtualization](), [Hardware](), [Deployment](), [hyper-v](), [Windows](), [Microsoft](), [Azure](), [nutanix](), [intel]() Tags: [azure](), [capacity](), [deployment](), [enterprise](), [hardware](), [hyper-v](), [intel](), [microsoft](), [nutanix](), [nvme](), [pcie](), [server](), [virtualization]() ### AI overview Lenovo added the four-socket ThinkAgile VX850 V4 to its virtualization lineup for demanding in-memory workloads. The launch also introduces three Infrastructure Deployment Services tiers and pre-validated Top Choice Express bundles covering Microsoft, Red Hat, SUSE, and Nutanix. Lenovo has not yet published the VX850 V4's own product specifications. ### Source excerpt Lenovo has added the ThinkAgile VX850 V4 to its virtualization lineup, a four-socket node it describes as purpose-built for highly virtualized and demanding in-memory workloads, and paired the launch with a three-tier deployment service framework and a set of pre-validated Top Choice Express bundles spanning Microsoft, Red Hat, SUSE, and Nutanix. The announcement follows Lenovo's The post Lenovo ThinkAgile VX850 V4 Takes On In-Memory Virtualization With New Deploy Tiers and Four-Hypervisor Express Bundles appeared first on StorageReview.com. ## TrueNAS Releases a Native Proxmox VE Plugin for Managing VM Storage DevFeed: [TrueNAS Releases a Native Proxmox VE Plugin for Managing VM Storage]() Original publisher: [Read original article]() Author: Brandon Lee Published: 2026-09-16T13:40:55Z Content type: article Language: en Sources: [Virtualization Howto]() Topics: [truenas](), [Proxmox](), [virtualization](), [NVMe]() Tags: [bug](), [home-lab](), [home-server](), [nvme](), [proxmox](), [release](), [testing](), [truenas](), [virtualization](), [zfs]() ### AI overview TrueNAS released an official Proxmox VE storage plugin that lets administrators provision and manage TrueNAS-backed VM disks through Proxmox workflows. The plugin supports iSCSI, NVMe over TCP, and OpenZFS snapshots, but the article describes it as an early-adopter release that is not yet recommended for production workloads. ### Source excerpt It really feels like Proxmox is on fire lately with announcements and momentum behind the solution. We just came off the heels of Proxmox announcing they were bringing on 24x7... The post TrueNAS Just Fixed One of Proxmox's Biggest Storage Headaches appeared first on Virtualization Howto. ## AAEON MIX-PTLWV1 Panther Lake industrial mini-ITX motherboard offers quad DisplayPort, dual 10GbE, and more DevFeed: [AAEON MIX-PTLWV1 Panther Lake industrial mini-ITX motherboard offers quad DisplayPort, dual 10GbE, and more]() Original publisher: [Read original article]() Author: Jean-Luc Aufranc (CNXSoft) Published: 2026-09-16T12:00:59Z Content type: news Language: en Sources: [CNX Software - Embedded Systems News]() Topics: [Embedded Systems](), [intel](), [Intel Core](), [ddr5](), [M.2 NVMe](), [pcie](), [Graphics](), [Specifications]() Tags: [10gbe](), [2-5gbe](), [4g-lte](), [5g](), [aaeon](), [artificial-intelligence-ai](), [automotive](), [ddr5](), [graphics](), [hardware](), [industrial](), [intel-core](), [linux](), [m-2-nvme](), [mini-itx](), [motherboard](), [nvme](), [panther-lake](), [pcie](), [single-board-computer](), [smart-city](), [specifications](), [ubuntu](), [windows-11]() ### AI overview CNX Software reports on the AAEON MIX-PTLWV1, an industrial mini-ITX motherboard built around Intel Core Ultra Series 3 Panther Lake processors. It supports up to 128GB of DDR5 memory, NVMe storage, four DisplayPort outputs, optional wireless and cellular modules, multiple Ethernet ports including up to two 10GbE connections, serial interfaces, and PCIe expansion. ### Source excerpt AAEON MIX-PTLWV1 is an industrial mini-ITX motherboard based on Intel Core Ultra Series 3 "Panther Lake" processors with up to two 10GbE and two 2.5GbE RJ45 jacks, and four 4K-capable DisplayPort video outputs. The motherboard also supports up to 128GB DDR5 C/SO-DIMM memory, M.2 NVMe SSD storage, optional WiFi and Bluetooth, and/or 4G LTE/5G cellular via M.2 modules, a PCIe Gen4 x8 slot for expansion, six serial interfaces, 12V-24V wide DC input, and more. AAEON MIX-PTLWV1 specifications: SoC - Intel Panther Lake processor (PTL-H404/PTL-H484 families) with Intel Xe LPG graphics; TDP: 15 to 25W System Memory - Up to 128GB via 2x 262-pin DDR5 C/SO-DIMM sockets up to 6400MHz (H404 SKU) or 7200MHz (H484 SKU) Storage - M.2 2280 M-Key (PCIe Gen4 x4) socket for NVMe SSD Display Interfaces 4x DisplayPort 1.4 up to 4096 x 2160 @ 60Hz 40-pin LVDS/eDP co-layout connector for Default: 18/24 bit dual-channel LVDS up [...] The post AAEON MIX-PTLWV1 Panther Lake industrial mini-ITX motherboard offers quad DisplayPort, dual 10GbE, and more appeared first on CNX Software - Embedded Systems News. ## Ubuntu 26.10 Set To Deliver Better Performance For Intel Core 3 Wildcat Lake DevFeed: [Ubuntu 26.10 Set To Deliver Better Performance For Intel Core 3 Wildcat Lake]() Original publisher: [Read original article]() Author: Michael Larabel Published: 2026-09-15T15:34:07Z Content type: comparison Language: en Sources: [Phoronix]() Topics: [Ubuntu](), [Intel Core](), [intel](), [Hardware](), [Linux](), [SOC]() Tags: [comparison](), [core](), [desktop-linux](), [graphics](), [hardware](), [intel-core](), [kernel](), [linux-benchmarking](), [linux-hardware-benchmarks](), [linux-hardware-reviews](), [linux-how-to](), [linux-performance](), [linux-server-benchmarks](), [nvme](), [open-source-graphics](), [performance](), [phoronix](), [phoronix-test-suite](), [release](), [soc](), [software](), [ubuntu](), [ubuntu-benchmarks](), [ubuntu-hardware](), [updates](), [upgrade](), [upgrades](), [wildcat-lake]() ### AI overview The article compares Ubuntu 26.04 LTS with Ubuntu 26.10 in development on a CHUWI UniBook equipped with an Intel Core 3 304 SoC, 8GB of RAM, and a 256GB NVMe SSD. It reports that Ubuntu 26.10's newer Linux kernel, Mesa graphics stack, and other software upgrades are expected to improve performance on Intel Wildcat Lake hardware. ### Source excerpt While Ubuntu 26.04 LTS is working fine out-of-the-box on new Intel Core 3 "Wildcat Lake" laptops like the CHUWI UniBook, next month's release of Ubuntu 26.10 will help deliver better performance out of these low-cost laptop options. ## Lightbits Inferra KV Cache Engine Claims 16x Session Density and 10M-Token Contexts DevFeed: [Lightbits Inferra KV Cache Engine Claims 16x Session Density and 10M-Token Contexts]() Original publisher: [Read original article]() Author: Harold Fritts Published: 2026-09-14T16:23:21Z Content type: news Language: en Sources: [StorageReview.com]() Topics: [Inference](), [Inference Performance](), [Orchestration](), [GPU](), [Multi-tenancy](), [sglang](), [TensorRT](), [vllm](), [Security]() Tags: [ai](), [cache](), [concurrent](), [enterprise](), [gpu](), [inference](), [latency](), [nvme](), [orchestration](), [security]() ### AI overview Lightbits Labs is introducing Inferra, a KV cache orchestration engine for AI inference. It virtualizes GPU memory across DRAM and NVMe storage, preserving attention states for long-context and multi-session workloads. Lightbits claims up to 16 times more concurrent sessions, more than 100 times lower latency than recomputation, and context windows of up to 10 million tokens. Inferra supports vLLM, TensorRT, and SGLang and includes tiering, predictive prefetching, tenant isolation, and encrypted data transfer. ### Source excerpt Lightbits Labs, the company that invented NVMe over TCP, is moving into inference software with Inferra, a KV cache orchestration engine that makes its public debut tomorrow, September 15, at the AI Infra Summit in Santa Clara. The software virtualizes GPU memory across DRAM and NVMe storage tiers and turns the KV cache into a The post Lightbits Inferra KV Cache Engine Claims 16x Session Density and 10M-Token Contexts appeared first on StorageReview.com. ## Linux server performance: Is disk I/O slowing your application? DevFeed: [Linux server performance: Is disk I/O slowing your application?]() Original publisher: [Read original article]() Author: Hayden James Published: 2026-09-12T16:11:54Z Content type: tutorial Language: en Sources: [Hayden James]() Topics: [Linux](), [IO](), [Server](), [Command-line interface](), [Processes](), [NVMe]() Tags: [blog](), [command-line](), [cpu](), [linux](), [load](), [mysql](), [nvme](), [performance](), [processes](), [request](), [server](), [ssd](), [storage](), [sysadmins](), [terminal]() ### AI overview This tutorial explains how disk I/O bottlenecks can cause high Linux load averages even when CPU usage is low. It shows how to use top, atop, and iotop on a LEMP web server, including per-core I/O wait measurements, to identify storage-related application slowdowns. ### Source excerpt If your Linux server is bogged down by disk I/O, your first step may often be to use the top command in the terminal to check load averages. Continue reading... ## CachyOS vs. Windows 11 vs. Ubuntu 26.04 LTS On Intel Wildcat Lake + 8GB RAM DevFeed: [CachyOS vs. Windows 11 vs. Ubuntu 26.04 LTS On Intel Wildcat Lake + 8GB RAM]() Original publisher: [Read original article]() Author: Michael Larabel Published: 2026-09-11T15:08:55Z Content type: comparison Language: en Sources: [Phoronix]() Topics: [benchmarking](), [Linux](), [Ubuntu](), [Windows](), [cpu](), [Hardware](), [intel](), [Linux performance]() Tags: [benchmarking](), [desktop-linux](), [hardware](), [intel](), [intel-core](), [laptop](), [linux](), [linux-benchmarking](), [linux-hardware-benchmarks](), [linux-hardware-reviews](), [linux-how-to](), [linux-performance](), [linux-server-benchmarks](), [lts](), [memory](), [microsoft](), [nvme](), [open-source-graphics](), [performance](), [phoronix](), [phoronix-test-suite](), [release](), [review](), [ssd](), [testing](), [ubuntu](), [ubuntu-benchmarks](), [ubuntu-hardware](), [windows](), [windows-11]() ### AI overview This comparison measures Ubuntu 26.04 LTS, CachyOS, and Windows 11 performance on a $449 CHUWI UniBook laptop with an Intel Core 3 Wildcat Lake processor, 8GB of memory, and a 256GB NVMe SSD across various workloads. ### Source excerpt Often times when testing different Linux distributions or comparing Windows vs. Linux it's on leading flagship desktop or server hardware, but today we are looking at the Ubuntu vs. CachyOS vs. Windows performance at the opposite end of the spectrum. With the new CHUWI UniBook $449 laptop powered by Intel Core 3 Wildcat Lake and with 8GB of system memory, here is a look at how those three operating systems compare across a variety of workloads. ## HP ZGX Fury Is Now Orderable: GB300 Superchip, 748GB Unified Memory, and a Red Hat AI Factory Plan for the Edge DevFeed: [HP ZGX Fury Is Now Orderable: GB300 Superchip, 748GB Unified Memory, and a Red Hat AI Factory Plan for the Edge]() Original publisher: [Read original article]() Author: Brian Beeler Published: 2026-09-09T19:35:12Z Content type: news Language: en Sources: [StorageReview.com]() Topics: [Artificial Intelligence](), [Inference](), [Blackwell](), [GPU](), [Nvidia](), [DGX Station](), [Grace CPU](), [Hardware](), [Fine-tuning](), [Availability](), [NVLink](), [NCCL]() Tags: [10gbe](), [ai](), [ai-factory](), [availability](), [blackwell](), [connectx](), [consumer](), [dgx-station](), [fine-tuning](), [gpu](), [grace-cpu](), [hardware](), [inference](), [nvidia](), [nvlink](), [nvme](), [red-hat](), [usb](), [workstation]() ### AI overview HP's ZGX Fury AI station is available to order with a GB300 Grace Blackwell Ultra Desktop Superchip, 748GB of unified memory, and up to 20 petaFLOPS of FP4 compute. HP positions it as a shared inference system for departments, factory floors, and branch offices, supported by a collaboration with Red Hat and NVIDIA to run Red Hat AI Factory with NVIDIA. ### Source excerpt HP's ZGX Fury AI station is now available to order, and HP paired the availability news with a collaboration with Red Hat and NVIDIA to put Red Hat AI Factory with NVIDIA on top of it. The ZGX Fury is HP's take on NVIDIA's DGX Station design, built around the GB300 Grace Blackwell Ultra Desktop The post HP ZGX Fury Is Now Orderable: GB300 Superchip, 748GB Unified Memory, and a Red Hat AI Factory Plan for the Edge appeared first on StorageReview.com. ## QNAP TVS-hx77AX Brings NFS over RDMA and U.2 NVMe to Desktop ZFS NAS DevFeed: [QNAP TVS-hx77AX Brings NFS over RDMA and U.2 NVMe to Desktop ZFS NAS]() Original publisher: [Read original article]() Author: Lyle Smith Published: 2026-09-09T17:18:09Z Content type: news Language: en Sources: [StorageReview.com]() Topics: [systems](), [networking](), [virtualization](), [cpu](), [Network](), [GPU](), [Artificial Intelligence]() Tags: [10gbase-t](), [2-5gbe](), [amd](), [cpu](), [enterprise](), [gpu](), [medium-nas](), [memory](), [nas](), [networking](), [nvme](), [pcie](), [qnap](), [virtualization]() ### AI overview QNAP introduced the TVS-hx77AX series of desktop ZFS NAS systems, combining AMD Ryzen 7000 processors, DDR5 memory, U.2 NVMe storage, and NFS over RDMA. The 16-, 12-, and 8-bay models target video collaboration, virtualization, and small-scale AI workloads. ### Source excerpt QNAP has introduced the TVS-hx77AX series, a new line of desktop ZFS NAS systems combining AMD Ryzen 7000 series processors with DDR5 memory, U.2 NVMe storage, and NFS over RDMA support. The family includes the 16-bay TVS-h1677AX, 12-bay TVS-h1277AX, and 8-bay TVS-h877AX, and QNAP is aiming all three at video collaboration, virtualization, and small-scale AI The post QNAP TVS-hx77AX Brings NFS over RDMA and U.2 NVMe to Desktop ZFS NAS appeared first on StorageReview.com. ## AWS Launches Graviton5-Based EC2 R9g and R9gd Memory-Optimized Instances DevFeed: [AWS Launches Graviton5-Based EC2 R9g and R9gd Memory-Optimized Instances]() Original publisher: [Read original article]() Author: Harold Fritts Published: 2026-09-01T18:42:59Z Content type: news Language: en Sources: [StorageReview.com]() Topics: [Amazon Web Services](), [Amazon EC2](), [Arm](), [cpu](), [Hardware](), [Nitro System](), [virtualization](), [Nitro Isolation Engine](), [Database](), [Caching](), [Kubernetes](), [Formal verification]() Tags: [amazon-ec2](), [amazon-eks](), [arm](), [aws](), [c-plus-plus](), [caching](), [cloud](), [database](), [docker](), [enterprise](), [formal-verification](), [go](), [hardware](), [java](), [microservices](), [nitro-isolation-engine](), [nitro-system](), [node-js](), [nvme](), [processors](), [python](), [time]() ### AI overview AWS has generally released the Amazon EC2 R9g and R9gd instances, powered by AWS Graviton5 processors. These Arm-based, memory-optimized instances offer higher per-vCPU performance, faster memory, larger cache capacity, and increased networking and Amazon EBS bandwidth for database, caching, analytics, container, and microservices workloads. R9gd adds local NVMe SSD storage for low-latency local capacity. Both families use the AWS Nitro System and Nitro Isolation Engine, with AWS citing formal verification for the isolation properties. ### Source excerpt AWS has made its Amazon EC2 R9g and R9gd instances generally available, introducing memory-optimized systems powered by AWS Graviton5 processors. The new Arm-based instances target database, caching, analytics, container, and microservices workloads that require high memory bandwidth and improved per-vCPU performance. Compared with Graviton4-based R8g instances, AWS states that R9g can deliver up to 25% The post AWS Launches Graviton5-Based EC2 R9g and R9gd Memory-Optimized Instances appeared first on StorageReview.com. ## How Object Storage Performance Affects GPU Training Workloads DevFeed: [How Object Storage Performance Affects GPU Training Workloads]() Original publisher: [Read original article]() Author: Sridhar Rajarao Published: 2026-08-30T00:00:00Z Content type: article Language: en Sources: [Sridhar Rajarao]() Topics: [GPU](), [cloud-infrastructure](), [Caching](), [Disk image]() Tags: [ai](), [aws](), [cache](), [compute](), [gpu](), [nvme](), [object-storage](), [performance](), [storage](), [training-data]() ### AI overview This article explains how object-storage latency, network transfers, many small files, and concurrent requests can leave GPUs idle during AI training. It recommends sharding small samples into larger files, reusing connections, and caching data near the GPU; it also notes AWS's suggested shard range of roughly 100 MB to 1 GB for many training workloads. ### Source excerpt The GPU does not care that object storage is durable and scalable. It cares whether the next batch of data arrives before it goes idle. ## KV vs Prefix vs Prompt vs Semantic Caching DevFeed: [KV vs Prefix vs Prompt vs Semantic Caching]() Original publisher: [Read original article]() Author: Avi Chawla Published: 2026-08-27T20:01:47Z Content type: article Language: en Sources: [Daily Dose of Data Science]() Topics: [Caching](), [Large Language Model](), [Neo4j]() Tags: [caching](), [embedding](), [knowledge-graph](), [latency](), [llm](), [memory](), [nvme](), [object-storage](), [retrieval](), [server](), [token](), [vector]() ### AI overview The article distinguishes KV caching, prefix caching, prompt caching, and semantic caching in LLM systems. It explains what each cache stores, how exact-match and fuzzy-match behavior differ, and how cache misses affect cost, latency, and correctness. It also describes Konig's tiered storage approach for many small knowledge graphs, though the supplied text is truncated. ### Source excerpt ...explained with best practices in production. ## Debugging my new network, when 10 Gigabit Ethernet Runs at 300 Megabits DevFeed: [Debugging my new network, when 10 Gigabit Ethernet Runs at 300 Megabits]() Original publisher: [Read original article]() Published: 2026-08-27T15:25:48Z Content type: article Language: en Sources: [Scott Hanselman]() Topics: [Homelab](), [debugging](), [UniFi UNAS Pro](), [intel](), [Windows](), [NVMe]() Tags: [debugging](), [homelab](), [intel](), [musings](), [nvme](), [smb](), [ssd](), [storage](), [unifi-unas-pro](), [windows]() ### AI overview A Windows desktop connected to a 10GbE homelab network was limited to roughly 100-200 megabits per second despite 10-gigabit link indications. Direct testing traced the problem to the Windows/NIC path: increasing the Intel E610 driver's receive buffers eliminated discarded packets and improved receive throughput, while disabling IPv4 Large Send Offload V2 raised transmit throughput to 7.03 Gbit/sec. ### Source excerpt I've been moving my home storage over to a UniFi UNAS Pro 8 as part of a larger homelab moderization. My main IRONHEART (the ultimate PC from a few years back) desktop now has an Intel E610-XT2 10GbE card, the NAS is on 10GbE, and there's a Minisforum MS-01 miniPC on the same network with a 10GbE SFP+ connection running Immich and Portainer and a few other things. Everything says 10 gigabit. Windows says 10 gigabit. UniFi says 10 gigabit. SMB copies are using the correct NIC but my file copies are running at around 100 to 200 megabits per second which is sad making. Naturally, I blamed the NAS, and the spinning rust within. The UNAS has six 16 TB spinning disks in RAID 6 and a pair of NVMe SSDs being used as cache. I'm also running Immich on the MS-01, with its photo library living on the UNAS, so there are lots of thumbnails, metadata reads, and little background writes happening. All seem like reasonable suspects. I switched the UNAS SSD cache from read-write to read-only. No meaningful difference. I stopped Immich completely. No difference. I looked at iostat; the disks weren't saturated. We looked at SMB signing and Windows Defender network scanning. Still slow. Then I stopped testing the NAS and ran iperf3 directly between the Windows desktop and the MS-01: iperf3 -c 192.168.1.222 -P 4 133 Mbit/sec Oops. The reverse test was better, but still wrong: iperf3 -c 192.168.1.222 -P 4 -R 1.33 Gbit/sec That's weird. Now the disks, SMB, Immich, RAID, and the NAS itself were completely out of the equation. This was a Windows/NIC problem and it's weirdly asymmetrical. Looking at the Intel adapter statistics shows me... Get-NetAdapterStatistics -Name "Ethernet - 10 Gig Intel" There were nearly a million ReceivedDiscardedPackets. During one ten-second iperf3 test, the counter increased by another 268. Why? The E610 driver had its receive buffers at the default 512, although it supported up to 4096. I increased them. I love an increased buffer. Set-NetAdapterAdvancedProperty ## Sandisk 600 and 800 NAS SSDs Revealed DevFeed: [Sandisk 600 and 800 NAS SSDs Revealed]() Original publisher: [Read original article]() Author: Rob Andrews Published: 2026-08-20T12:00:27Z Content type: news Language: en Sources: [NAS Compares]() Topics: [Hardware](), [NVMe](), [M.2 NVMe](), [Specifications]() Tags: [asustor-ssd](), [m-2-nvme](), [nas-600](), [nas-800](), [nas-drive-ssd](), [nas-hardware](), [nas-ssd](), [network-attached-storage-nas](), [nvme](), [qnap-ssd](), [sandisk](), [sandisk-1000](), [sandisk-200](), [sandisk-2026](), [sandisk-2027](), [sandisk-400](), [sandisk-600](), [sandisk-800](), [sandisk-nas-600](), [sandisk-nas-800](), [sandisk-nas-ssd](), [sandisk-ssd](), [specifications](), [ssd](), [ssd-for-nas](), [ssd-nas-drive](), [synology-ssd](), [ugreen-ssd](), [uncategorised](), [wd-2026](), [wd-2027](), [wd-nas-ssd](), [wd-red-nas-ssd](), [wd-red-ssd]() ### AI overview Sandisk has introduced the NAS 600 and NAS 800 SSD families for NAS systems. The NAS 600 is a 2.5-inch SATA SSD for storage and all-flash arrays, while the NAS 800 is an M.2 2280 PCIe 5.0 NVMe SSD for caching, tiered storage and primary storage in compatible systems. ### Source excerpt New SSDs for NAS Released by Sandisk - Too late? Sandisk has introduced 2 SSD families designed specifically for NAS use, covering both conventional 2.5-inch SATA storage and higher-performance M.2 NVMe applications. The SANDISK NAS 600 and NAS 800 series are intended for prosumers, creative professionals and small businesses using always-on systems for shared storage, [...] ## QNAP TS-262A and TS-462A NAS Released - New x62A Series Replaces the Older TS-262/TS-462 DevFeed: [QNAP TS-262A and TS-462A NAS Released - New x62A Series Replaces the Older TS-262/TS-462]() Original publisher: [Read original article]() Author: Rob Andrews Published: 2026-08-12T16:00:09Z Content type: news Language: en Sources: [NAS Compares]() Topics: [QNAP](), [TS-462A](), [Intel Celeron N5095](), [TS-462](), [intel](), [systems]() Tags: [2-5gbe](), [2-bay-nas](), [4-bay-nas](), [home-nas](), [intel](), [intel-celeron-n5095](), [nas](), [network-attached-storage-nas](), [news](), [news-2026](), [nvme](), [plex-nas](), [qnap](), [qnap-nas](), [qnap-ts-262a](), [qnap-ts-462a](), [qts](), [systems](), [ts-262](), [ts-262a](), [ts-462](), [ts-462a](), [uncategorised](), [value]() ### AI overview QNAP has listed the TS-262A and TS-462A, new desktop NAS systems that appear to replace the older TS-262 and TS-462. The main reported change is a move from the dual-core Intel Celeron N4505 to the quad-core Intel Celeron N5095. ### Source excerpt QNAP has now listed the TS-262A and TS-462A, two new Intel-powered desktop NAS systems that appear to replace the older TS-262 and TS-462 in the value-focused x62 family. The headline change is the move from the older dual-core Intel Celeron N4505 platform to a quad-core Intel Celeron N5095, while keeping the same core idea: affordable [...] ## Monitor Your Drive Health with Performance Co-Pilot on Fedora DevFeed: [Monitor Your Drive Health with Performance Co-Pilot on Fedora]() Original publisher: [Read original article]() Author: Paul Evans Published: 2026-08-10T08:00:00Z Content type: article Language: en Sources: [Fedora Magazine]() Topics: [Fedora](), [Monitoring](), [Grafana](), [telemetry]() Tags: [data](), [diagnostics](), [drive](), [farm](), [fedora-project-community](), [grafana](), [hdd](), [health](), [home-server](), [metrics](), [monitor](), [monitoring](), [nvme](), [open-source](), [performance](), [performance-metrics](), [self-hosted](), [server](), [smart](), [ssd](), [telemetry](), [wwid]() ### AI overview This article explains how to monitor drive health on Fedora using Performance Co-Pilot (PCP). It covers SMART metric collection for HDDs, SSDs, and NVMe drives, NVMe error-log decoding, WWID-based tracking, Seagate FARM telemetry, continuous metric collection, and Grafana dashboards for viewing drive-health trends. ### Source excerpt You wake up one morning to find your home server unresponsive, after some investigation you discover a failed NVMe drive taking your self-hosted services and data with it. Perhaps you're a system administrator and a workstation's SSD has been silently accumulating errors for months, and now a user is reporting corrupted files. Drive failures are [...] ## Building a Modern Analytics Stack Around ClickHouse DevFeed: [Building a Modern Analytics Stack Around ClickHouse]() Original publisher: [Read original article]() Author: Sebastian Insausti Published: 2026-07-01T10:35:21Z Content type: article Language: en Sources: [SeveralNines]() Topics: [clickhouse](), [Databases](), [Open Source](), [MySQL](), [PostgreSQL](), [real-time](), [NVMe](), [Replication](), [IO]() Tags: [analytics](), [analytics-stack](), [clickhouse](), [cta](), [database-general](), [databases](), [distributed](), [mysql](), [nvme](), [olap](), [open-source](), [postgresql](), [real-time]() ### AI overview This article explains how ClickHouse fits into a modern analytics stack alongside relational databases, Redis, and Prometheus. It describes ClickHouse as an open-source column-oriented DBMS for analytical workloads and discusses columnar storage, distributed processing, replication, and near-real-time ingestion. ### Source excerpt Historically, relational databases did double duty. The same PostgreSQL / MySQL instance that handled your application's writes also answered your business questions. A well-indexed schema, a few GROUP BY reports, done. And that works, right up until it doesn't: the reports get slower, the dashboards start eating the same I/O budget as the application; or, [...] The post Building a Modern Analytics Stack Around ClickHouse appeared first on Severalnines. ## For whom the door-bell tolls DevFeed: [For whom the door-bell tolls]() Original publisher: [Read original article]() Author: Kyle Bader Published: 2026-06-11T00:00:00Z Content type: article Language: en Sources: [Ceph Blog]() Topics: [Caching](), [Large Language Model](), [Inference Performance](), [Low-Latency Inference](), [GPU](), [cpu]() Tags: [2025](), [ai](), [architecture](), [blog-post](), [cache](), [caching](), [ceph](), [cpu](), [en-article](), [en-blog-post](), [gpu](), [gpu-initated](), [inference](), [llm](), [nvme](), [rados](), [rados-nkv](), [storage]() ### AI overview The article explores using Ceph RADOS through NVMe key-value commands to support GPU-initiated storage access for LLM KV caching. It connects content-addressable caching with the efficiency of GPU-initiated NVMe access and discusses a possible vendor-specific extension for executing classes against objects. ### Source excerpt For whom the door-bell tolls In a previous post we extolled the benefits of KV caching, a technique to save the KV states from the prefill step of LLM-based inference to reduce time to first token (TTFT) and skip redundant computation. I co-presented this with Tushar Gohad at Cephalocon. Since then I've been thinking a lot about how to improve the state of the art. Really move the needle. We've made strides in a lot of areas in Ceph, especially the work going into Fast EC -- if you don't know what I'm talking about you should check it out, it promises huge benefits to a large category of workloads. That's not what we're here for today, though. Last year there was one paper that stuck in my mind, which is fairly remarkable because I read on the order of 130. That paper was GPU-Initiated On-Demand High-Throughput Storage Access in the BaM System Architecture. I struggled with this. It describes a system where a CPU loads a kernel into the GPU that allows the GPU to serve as an NVMe initiator. The struggle was rooted in the fact that block just felt like the wrong interface for KV caching. If you use block, then you need a lookup table that maps the hash of the sequences representing a cache block to a particular (device, offset, length) tuple. It begs for a content-addressable approach with no centralized lookup or coordination. On the other hand, what was described in the paper was just flat-out more electrically efficient. I couldn't have my cake and eat it too. The idea: RADOS, spoken as NVMe key-value ¶ In 2025 we also saw the first ratified version of the NVMe key-value command set specification. We already have a Ceph implementation of NVMe/TCP that leverages SPDK, and it seemed like we could add support fairly easily for the key-value command set. For those who don't know the internal mechanics of Ceph, the native API operates against RADOS objects. RADOS is richer than most object stores: it supports reads and writes to arbitrary offsets, deletes, key-value sto ## v20.2.1 Tentacle released DevFeed: [v20.2.1 Tentacle released]() Original publisher: [Read original article]() Author: Yuri Weinstein Published: 2026-04-06T00:00:00Z Content type: release Language: en Sources: [Ceph Blog]() Topics: [bug](), [configuration](), [monitor](), [Command-line interface](), [API](), [dashboards](), [JSON]() Tags: [2026](), [api](), [blog-post](), [bug](), [cli](), [en-article](), [en-blog-post](), [json](), [monitor](), [nvme](), [osd](), [peering](), [policy](), [recovery](), [reef](), [release](), [rgw](), [tentacle](), [ui](), [workflows]() ### AI overview Ceph v20.2.1, the first minor release in the Tentacle series, is recommended for all users. The release fixes storage, recovery, peering, validation, and dashboard issues; restricts unsafe fast EC configurations; improves NVMeoF Gateway failover; adds listener and subsystem CLI commands; introduces a transient exclusive RBD lock policy; and updates dashboard navigation and forms. ### Source excerpt This is the first minor release in the Tentacle series. We recommend that all users update to this release. Release Date ¶ April 06, 2026 Notable Changes ¶ OSD / BlueStore ¶ EC Recovery: Fixed a length calculation bug in erase_after_ro_offset() that caused empty shards to retain data, leading to shard_size >= tobj_size assertion failures when recovering small objects in EC pools. BlueFS Volume Selector: Updated the BlueFS volume selector to properly account for file size changes when recovering the WAL in envelope mode. BlueFS: Fixed a bug where stat() missed the actual file size update after indexing WAL envelope files. Monitor (mon) ¶ Fast EC Restrictions: Denied the ability to enable EC optimizations ("fast EC") for non-4K-aligned chunk sizes. Unaligned chunk sizes handled by fast EC perform poorly and suffer from bugs, so attempts to force this configuration are now rejected. Peering: Ensured ceph pg repeer proposes a correctly sized pg temp, as optimized EC cannot cope with mismatched sizes. NVMeoF Gateway: Added a new nvme-gw listeners command to display all existing listeners (including auto-listeners) inside a pool/group. NVMeoF Failover: Overhauled the NVMeoF Gateway fast-failover logic. Beacon timeouts are now evaluated within prepare_beacon to support shorter intervals, and the mechanism for detecting monitor slowness was improved. librbd & rbd-mirror ¶ RBD: Introduced a new RBD_LOCK_MODE_EXCLUSIVE_TRANSIENT policy for rbd_lock_acquire(). This is a low-level interface intended to allow a peer to grab exclusive lock manually for short periods of time with other peers pausing their activity and waiting for the lock to be released rather than instantly aborting I/O and returning an error. It's possible to switch from RBD_LOCK_MODE_EXCLUSIVE to RBD_LOCK_MODE_EXCLUSIVE_TRANSIENT policy and vice versa even if the lock is already held. Ceph Object Gateway (RGW) ¶ Multi-Part Operations: Fixed conditional validation handling in MultiWrite, Delete, and MultiDelete ## RocksDB Compression in Ceph: Space Savings with No Performance Cost DevFeed: [RocksDB Compression in Ceph: Space Savings with No Performance Cost]() Original publisher: [Read original article]() Author: Daniel Alexander Parkes, Anthony D'Atri Published: 2025-12-17T00:00:00Z Content type: article Language: en Sources: [Ceph Blog]() Topics: [Compression](), [Database](), [Hardware](), [ibm]() Tags: [blog-post](), [ceph](), [compression](), [configuration](), [cost](), [devices](), [en-article](), [en-blog-post](), [ibm](), [mon](), [nvme](), [osd](), [performance](), [rados](), [reef](), [rocksdb](), [space](), [storage]() ### AI overview This article reports Ceph performance tests showing that enabling RocksDB compression can substantially reduce metadata database space, especially for smaller objects, without harming throughput or resource consumption. The tests used IBM Storage Ceph 7.1 with BlueStore OSDs, HDD object storage, and NVMe devices for the RocksDB WAL and database. ### Source excerpt Introduction ¶ In the world of data storage, engineers and architects constantly face a fundamental dilemma: the trade-off between performance and efficiency. It's a balancing act. When you want to save space, you typically enable features like compression, but the common assumption is that this will cost you performance, a CPU cycle tax that slows throughput. But what if you could significantly reduce your metadata storage footprint without slowing things down? This search for an answer to this question started with research work from Mark Nelson, who published a blog post on ceph.io that covers RocksDB tuning in depth, exploring RocksDB compression with positive results. These promising results sparked a conversation on the upstream GitHub about enabling compression by default; a link to the PR is available here. To build on the previous investigation, the Ceph performance team ran tests on a robust hardware configuration running IBM Storage Ceph 7.1 (Reef). The cluster used the BlueStore OSDs for an erasure-coded (EC 4+2) pool, with a hybrid OSD storage setup: HDDs for object data and fast NVMe drives for the BlueStore WAL+DB. To understand the test, it's helpful to know what the WAL+DB is. In modern Ceph, the BlueStore storage engine manages all data on the OSDs (physical devices). To do this, it must maintain a vast catalog of internal metadata: think of it as a high-speed index that quickly locates every piece of data. RocksDB, a high-performance key-value database, manages this critical index. In our hybrid cluster, the RocksDB database runs on the fast NVMe deviceses, while the actual object data resides on the slower HDDs. Because this metadata can grow very large, RocksDB's efficiency, how much space it consumes on those expensive NVMe drives, is a critical factor in the cluster's overall cost and performance. Our test, therefore, focuses on a simple, high-stakes question: Can we compress this metadata to save space without paying a performance penalty? Ex ## How io\_uring improves database performance DevFeed: [How io\_uring improves database performance]() Original publisher: [Read original article]() Published: 2025-12-11T00:00:00Z Content type: tutorial Language: en Sources: [Gaurav Sarma's Blog]() Topics: [io\_uring](), [Databases](), [Linux Kernel](), [NVMe]() Tags: [asynchronous](), [database-performance](), [io](), [io-uring](), [nvme](), [performance]() ### AI overview This article explains how io_uring improves database performance through shared submission and completion queues, batching, fewer system calls, zero-copy operation, and features such as SQPoll and registered buffers. It also discusses DMA and the relevance of fast NVMe storage. ### Source excerpt . [How io_uring Improves Database Performance](how-iouring-improves-database-performance-cover... ## Why I Ditched the Cloud for On-Prem Infrastructure DevFeed: [Why I Ditched the Cloud for On-Prem Infrastructure]() Original publisher: [Read original article]() Author: Shivam Anand Published: 2025-10-25T16:30:21Z Content type: article Language: en Sources: [Shivam Anand - DevOps & Cloud Engineering Blog]() Topics: [on-prem](), [cloud-infrastructure](), [migration](), [networking](), [BGP](), [Firewall](), [Hardware](), [Artificial Intelligence]() Tags: [ai](), [ai-workloads-on-prem](), [bgp](), [bgp-and-asn-networking](), [cloud-connectivity](), [cloud-cost-optimisation-2025](), [cloud-infrastructure](), [cloud-infrastructure-savings](), [cloud-repatriation-case-study](), [cloud-to-on-prem-migration](), [enterprise-data-centre-setup](), [firewall](), [hardware](), [it-infrastructure-2025](), [managed-cloud-vs-on-prem](), [migration](), [multi-site-high-availability](), [network-cost-reduction](), [networking](), [nixi-ip-block-india](), [nvme](), [on-prem](), [on-prem-infrastructure-benefits](), [open-source-monitoring-stack](), [proxmox-vs-vmware-migration](), [subhash-choudhary-dukaan-cto]() ### AI overview The article recounts a migration from cloud infrastructure to on-premise data centres after cloud cost-optimization efforts produced minimal results. It describes the networking, hardware, and software setup, and reports that the monthly cloud bill fell from INR 40 lakhs to INR 3 lakhs after deployment. ### Source excerpt Explore why migrating from costly cloud infrastructure to efficient on-premise data centres can drastically reduce monthly expenses and complexity, especially for data-heavy AI workloads. Learn about real-world experiences, technical setups, and best practices for on-prem cloud repatriation in 2025. ## Migrating from VMware to Proxmox VE at 3,000-Core Scale DevFeed: [Migrating from VMware to Proxmox VE at 3,000-Core Scale]() Original publisher: [Read original article]() Author: Shivam Anand Published: 2025-08-23T08:25:07Z Content type: article Language: en Sources: [Shivam Anand - DevOps & Cloud Engineering Blog]() Topics: [Proxmox](), [virtualization](), [migration](), [Automation](), [Command-line interface](), [ceph](), [NVMe]() Tags: [automation](), [ceph](), [cli](), [kvm](), [migration](), [nvme](), [proxmox](), [virtualization-cloud-infrastructure](), [vmware]() ### AI overview The article describes a three-month migration of 3,000 CPU cores, 8 TB of RAM, and production and archival storage from VMware to Proxmox VE. It explains the use of NFS, Ceph, and a custom CLI orchestrator to manage the resulting infrastructure. ### Source excerpt Discover how I migrated 3,000 cores, 8TB RAM from VMware to Proxmox VE, built a custom CLI orchestrator, and leveraged Ceph storage for seamless management. ## Beelink ME Mini Tested as a Low-Power Proxmox LXC Server DevFeed: [Beelink ME Mini Tested as a Low-Power Proxmox LXC Server]() Original publisher: [Read original article]() Author: Techno Tim Published: 2025-08-09T13:00:00Z Content type: opinion Language: en Sources: [Techno Tim]() Topics: [Proxmox](), [virtualization](), [Server]() Tags: [intel](), [low-power](), [mini-pc](), [monitoring](), [nvme](), [proxmox](), [server](), [storage](), [virtualization]() ### AI overview The article evaluates the Beelink ME Mini as a low-power Proxmox LXC server rather than only a NAS. It covers the hardware, Proxmox installation, cooling, NVMe bandwidth, power use, and a ZFS performance test. ### Source excerpt Everyone thinks this is a NAS. I don't. It's a low power Proxmox LXC Server and it's awesome! 📺 Watch Video Intro The Beelink ME Mini has been generating attention as a small form factor NAS, and with its compact footprint and six NVMe slots, it's easy to see why. But after spending time with it, I think it has another role worth exploring: virtualization. While it's marketed for storage... [Next page]()