BEST SELLERS

In Server & Components

Keep in mind this card has its limitations, I am looking at it from the stand point of this verse a RTX3090 used (about the same price). No way I am spending $1000 on a used card that can die tmw even with Cuda support. So now I get 32GB vs 24GB ram. I am primary running windows and LMstudio and Ollma worked fine after installing all the intel software (like 4 large installs) Vulkan support gave me around 100T/KS for qwen 3.5-35b. Out of the box with standard setup. This running on an Intel i7-11th gen with 64GB DDR4 and PCI 4 x8 interface.. The biggest hit is on how fast I can load the model in memory. The card is very quite with the fan. I suppose other people got loud fans.

16 GB of ECC VRAM allows for larger LLMs and context windows Needs only 70w of total board power (no external power connector required) Low cost compared to what Nvidia is offering

This is an excellent product and CyberPower is an excellent company.

Works seamlessly to plug in old drives and keep moving forward.

Works as expected

32GB VRAM ITs not a cutdown 9070 XT, its 98% the same speed in gaming and obviously much greater for large memory applications like pytorch/generative ai. Still boosts right up to 3ghz, can be undervolted, no stupid zero rpm fan idle Fan defaults to a silent 1000rpm idle and unlike most blower gpus it really doesnt need to ramp up path 50% pwm (2500rpm) to cool a sustained 300w. Games exactly like a 9070 xt.

Easy to set-up Easy to be used by any devices Easily support 6 users ( 4 iphones + 2 android + 5 window laptop + 2 macosx + 2 fire tv) backup and streaming)

This thing is truly amazing. Can run models up to 200 billion parameters. (Most of mine are between 70-120b, but my 24b "devstral" model is very useful for simple coding.) For a "personal AI" it's perfect. I use it to assist with sensitive client matters that I do not want to "share" with public AI services. Yes, it is brittle, but it is not hard to perform factory-restore using Nvidia's image - they have clear instructions posted and AI can help you if you need it. Be sure to use Nvidia's tool to create the boot thumb-drive. The unique hardware and driver situation makes this computer incompatible with other distros or OSes.

- Easy to setup - Plenty of Support - Works on Ubuntu/Win11 easily Ollama/LM Studio - Gemma 4:26b - Flux 2

Works as advertised at a reasonable price.

Best NAS OS DSM 7.3, user friendly, robust, slick. Can use any HDD brand with DSM 7.3. Support RAID and SHR.

Excellent solid hardware (2.5GB LAN, M.2 Storage for cashing, I/O Ports) at a great price. Installation of hardware (drives, etc.) done well. UGREEN NAS app make it simple and quick to get the unit up and running.

much quicker nas than my ds223 have the option to have a warm spare quiet

Product came with everything needed in the box no missing anything. Basic manual but to the point. I've worked with Supermicro in the past and when starting those motherboards I would constantly have fan issues and have go in with ipmitool to fix it. On this board that issue didn't happen, came up right away didn't have issues with ports just needed to make sure to setup the jumpers right to use m.2, mini sas to Sata and OCuLink. Its my first time building with EPYC I got a good deal on a 7313p no issues so far. running Unraid first time changing the main hardware with Unraid was fantastic.

Decent price for the hardware, sleek design, upgradable RAM, software updates are frequent, drive bays are solid, can control fan speed for the CPU

I was impressed with how the unit was packaged, the included accessories and the unit's design. UGREEN'S attention to details is top notch. I easily installed both of my drives on the unit's HDD rails, slid them in and booted up. I opted for a local configuration only (no UGREEN ID or web access) over LAN. UGOS is minimal but effective: configuration, diagnostics, basic apps, user management, etc. I also installed the UGREEN NAS app on my iPhone, which also works good.

Room for everything, usb ports are great