AINAS
UGREEN DXP8800 Ultra 8-bay NAS

UGREEN DXP8800 Ultra

AI ReadyUGREEN

Why AI Ready: Intel Core 7 150U (10C/12T, no NPU) with Intel Iris Xe Graphics (96EU), internal PCIe Gen4 x4 slot and 2x Thunderbolt 4 for a GPU, and up to 96GB DDR5.

Updated

Can the UGREEN DXP8800 Ultra run local AI?

The UGREEN DXP8800 Ultra can run small local models, with limits. Its Intel Core 7 150U and 96GB memory ceiling suit 70B class models at 4-bit with room left for longer context (Estimated), so treat it as a storage-first NAS that can host light AI tools.

It also has a GPU path (PCIe x4 slot), so a graphics card can be added later for much faster inference.

Local LLM support is documented by the manufacturer.

The short answer

The DXP8800 Ultra suits buyers who want an eight-bay NAS that also hosts local AI in Docker. It takes up to 96GB of DDR5 SO-DIMM 5200 memory and ships with 8GB, so the memory is expandable. With that ceiling, Mistral Large 2 123B fits at Q4 (Estimated). The Intel Core 7 150U has 10 cores and Intel Iris Xe Graphics (96EU). For a dedicated GPU there is a PCIe Gen4 x4 slot and 2 Thunderbolt 4 ports, and UGREEN documents GPU use in Docker. Two 10GbE ports and 2 M.2 slots suit fast storage and model files. UGOS Pro supports Docker and virtual machines, and Ollama and Open WebUI have vendor guides. The AINAS Score is 61.

The main catch is that inference on the DXP8800 Ultra is CPU-only until a graphics card is added. There is no NPU, and UGREEN rates local LLM support as limited, so the Intel Core 7 150U carries every model by itself. It ships with 8GB, which means the larger models need a memory upgrade toward the 96GB ceiling first (Estimated). Buyers who want fast chat with big models should plan for a dedicated GPU in the PCIe Gen4 x4 slot or over one of the 2 Thunderbolt 4 ports, which UGREEN documents for Docker. The power supply wattage is not published, so card size and power draw need checking first. The warranty length is also unconfirmed, which matters to anyone who wants written cover terms before committing.

Current offers

  • Amazon US

    8GB

    Ships from the US; delivery and import charges shown at checkout

    $1,779.99

    Verified

    In stock

    Buy on Amazon

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.co.uk or Amazon.com, as applicable, at the time of purchase will apply to the purchase of this product.

Price history

Current
$1,779.99
30-day average
$1,779.99
90-day low
$1,779.99
All-time tracked low
$1,779.99
$1,769.00$1,780.00$1,791.0020 AugToday
Last in-stock Amazon US price we verified each day. Dashed line: 30-day average. Days out of stock are not drawn as prices.

Key specifications

CPU
Intel Core 7 150U
Source: Manufacturer
GPU
Intel Iris Xe Graphics (96EU)
Source: Manufacturer
Included RAM
8GB DDR5 SO-DIMM 5200
Source: Manufacturer
Maximum RAM
96GB
Source: Manufacturer
Networking
2 x 10GbE
Source: Manufacturer
PCIe
1x PCIe Gen4 x4
Source: Manufacturer
GPU expansion
PCIe x4 slot
Source: Official docs
Power supply
Not stated
Source: Manufacturer
Docker
Yes
VM support
Yes
Released
2026
Segment
Thunderbolt

Can the DXP8800 Ultra run local LLMs?

Each capability shows how it was confirmed: Verified by AINAS, Vendor confirmed, Community confirmed or Unverified.

CapabilitySupportedEvidenceNote
Local LLMLimitedVendor confirmedCPU-only inference via Ollama in Docker per UGREEN KB; no NPU.
Ollama YesVendor confirmedUGREEN KB Docker Compose guide for Ollama/Open WebUI on UGOS Pro.
Open WebUI YesVendor confirmedOfficial UGOS Pro Open WebUI guide.
RAGLimitedVendor confirmedRAG via Open WebUI document chat only.
AI searchLimitedVendor confirmedPhotos description and OCR search via Model Manager; document semantic search applicability per model not stated.
GPU expansion YesVendor confirmedInternal PCIe Gen4 x4 slot or Thunderbolt 4 eGPU; UGREEN KB documents dedicated GPU use in Docker.
Hardware accelerationLimitedVendor confirmedQuick Sync on Iris Xe; CN page cites 8K transcoding; no NPU.
AI photos YesVendor confirmedAI photo album: faces, places, objects, duplicates.
AI videoLimitedVendor confirmedSurveillance Center local AI detection (people, vehicle, package, pet); features vary by model.
Surveillance AI YesVendor confirmedSurveillance Center supports DXP series.

What AI models can the DXP8800 Ultra run?

Calculated from the DXP8800 Ultra's memory and each model's published size at each quantization. Rows marked Estimated are calculations, not benchmarks. Check any model.

ModelBest quantizationRuns?Memory neededExpected performanceVerification
Llama 3.2 3BLlama, 3.21BQ8Yes13GBNot measuredEstimated
Llama 3.1 8BLlama, 8.03BQ8Yes19GBNot measuredEstimated
Qwen3 8BQwen, 8.19BQ8Yes19GBNot measuredEstimated
Gemma 3 12BGemma, 11.8BQ8Yes24GBNot measuredEstimated
Qwen3 14BQwen, 14.8BQ8Yes28GBNot measuredEstimated
Phi-4 14BPhi, 14.7BQ8Yes28GBNot measuredEstimated
Mistral Small 3 24BMistral, 23.6BQ8Yes38GBNot measuredEstimated
Gemma 3 27BGemma, 27BQ8Yes44GBNot measuredEstimated
Qwen3 30B A3BQwen, 30.5BQ8Yes47GBNot measuredEstimated
Qwen3 32BQwen, 32.8BQ8Yes50GBNot measuredEstimated
DeepSeek-R1 32B (Qwen distill)DeepSeek, 32.8BQ8Yes50GBNot measuredEstimated
Llama 3.3 70BLlama, 70.6BQ5Yes68GBNot measuredEstimated
Llama 4 Scout 17Bx16ELlama, 109BQ4Yes89GBNot measuredEstimated
Qwen2.5 72BQwen, 72.7BQ5Yes73GBNot measuredEstimated
Mistral Large 2 123BMistral, 123BQ4Yes96GBNot measuredEstimated

Workload fit

One verdict per workload, from the DXP8800 Ultra's use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.

Local LLMGood fit

Good for local LLM work, with one caveat on the software route. Ollama runs in Docker per UGREEN's guide, but local LLM support is limited to CPU-only inference with no NPU, and 96GB of memory holds large models (Estimated).

Private RAGGood fit

Good for private RAG, within a narrow scope. The vendor confirms RAG only as limited, through Open WebUI document chat, and Open WebUI and Ollama are both vendor confirmed with official UGOS Pro guides.

AI searchOK

AI search is rated OK, though coverage is partial. Support is limited: the vendor confirms photo description and OCR search through Model Manager, but does not state whether document semantic search applies for each model, so check before relying on it.

AI photosGood fit

AI photos are a good fit. The vendor confirms an AI photo album covering faces, places, objects and duplicates, on a 10-core Intel Core 7 150U with Iris Xe graphics and 2 M.2 slots for fast storage.

SurveillanceGood fit

Good for surveillance. Surveillance Center supports the DXP series and is vendor confirmed, with local AI detection of people, vehicles, packages and pets, though the related video AI row is limited and features vary by model.

Home labGood fit

Good for a home lab. UGOS Pro supports Docker and virtual machines on a 10-core Intel Core 7 150U, and the system adds two 10GbE ports, 2 Thunderbolt 4 ports and a PCIe Gen4 x4 slot for expansion.

BusinessGood fit

Good for business, mainly as a file and app server. It has 8 drive bays, two 10GbE ports, up to 96GB of memory and UGOS Pro with Docker and VM support, but ECC memory is not offered.

Creative AIGood fit

Good for creative AI, provided you add a GPU. A PCIe Gen4 x4 slot or Thunderbolt 4 eGPU is vendor confirmed, with dedicated GPU use documented in Docker, while built-in Quick Sync acceleration is limited and there is no NPU.

How much RAM does the DXP8800 Ultra support?

The DXP8800 Ultra ships with 8GB of DDR5 SO-DIMM 5200 and supports up to 96GB. Upgrading to the maximum lifts the model ceiling from small models up to around 7B parameters to 70B class models at 4-bit with room left for longer context (both Estimated).

As a rule of thumb, a 4-bit model needs about 0.6GB per billion parameters plus room for context. After leaving around 8GB for the NAS operating system, 96GB leaves space for a model of roughly 146B parameters at most (Estimated).

Can you add a GPU to the DXP8800 Ultra?

Partly. The DXP8800 Ultra has a 1x PCIe Gen4 x4, which suits low-profile cards and network or storage upgrades. Only small single-slot GPUs fit, so expect modest gains.

Is the DXP8800 Ultra suitable for RAG?

With limits. The DXP8800 Ultra can host a small RAG stack such as AnythingLLM or Open WebUI with a compact embedding model, but memory and compute restrict how large the answering model can be. It suits a personal document library more than a team knowledge base.

Networking is 2 x 10GbE. That is ample for indexing large libraries and for several people querying at once.

Limitations

  • There is no OCuLink port, so a dedicated GPU must go in the single PCIe Gen4 x4 slot, with only four lanes, or connect over one of the Thunderbolt 4 ports.
  • Inference on the system itself is CPU-only with no NPU, and local LLM support is rated limited. Faster work needs a dedicated GPU in the PCIe Gen4 x4 slot or over Thunderbolt 4.
  • ECC memory is not supported. The system ships with 8GB of DDR5 SO-DIMM and is expandable to 96GB, so larger models need a memory upgrade.
  • Warranty length and power supply wattage are not published or confirmed, so check both before you buy.
  • RAG is limited to Open WebUI document chat, and AI search is limited, with document semantic search not stated for each model.

Which DXP8800 Ultra configurations are sold?

ConfigurationRAMSSDNotes
8GB8GBNoneDiskless, 8GB DDR5, 128GB system SSD.

How is the DXP8800 Ultra different from the DXP8800 Plus?

The closest UGREEN model by bay count and AINAS Score. Only the specifications that differ are listed.

  • Processor: Intel Core 7 150U on the DXP8800 Ultra, Intel Core i5-1235U on the DXP8800 Plus
  • Integrated graphics: Intel Iris Xe Graphics (96EU) on the DXP8800 Ultra, Intel Iris Xe Graphics on the DXP8800 Plus
  • Memory ceiling: 96GB on the DXP8800 Ultra, 64GB on the DXP8800 Plus
  • AINAS Score: 61 on the DXP8800 Ultra, 60 on the DXP8800 Plus

See the full UGREEN DXP8800 Plus page

Who should buy the DXP8800 Ultra?

The DXP8800 Ultra scores best for Home Lab (68/100) and Business (66/100). It suits home lab owners who want one box for storage, containers, media and a few AI services.

What are the alternatives to the DXP8800 Ultra?

Other AI Ready systems closest in price.

Synology FS6400

Synology FS6400

Synology

AI Ready

512GB max RAM · 24 bays · 2 x 10GbE

AINAS Score 61

No verified price

View
Synology HD6500

Synology HD6500

Synology

AI Ready

512GB max RAM · 60 bays · 2 x 10GbE

AINAS Score 61

No verified price

View
Synology SA6400

Synology SA6400

Synology

AI ReadyDiscontinued

1024GB max RAM · 12 bays · 2 x 10GbE

AINAS Score 61

No verified price

View

Where does the DXP8800 Ultra rank?

Common questions

Can the DXP8800 Ultra run Mistral Large 2 123B?

Yes, at Q4 (Estimated). Mistral Large 2 123B needs 73GB at Q4, and the DXP8800 Ultra takes up to 96GB of memory, so the 8GB it ships with must be upgraded first. At Q8 it needs 130GB, which does not fit (Estimated). Qwen3 32B needs 20GB at Q4 and fits with more headroom (Estimated). These are calculations, not benchmarks, and inference without an added GPU is CPU-only.

Can you upgrade the memory in the DXP8800 Ultra?

Yes. The memory is expandable. The DXP8800 Ultra ships with 8GB of DDR5 SO-DIMM 5200 and takes up to 96GB. UGREEN lists 2 slots for the modules, and fitting both is how you reach the 96GB ceiling. Larger models need more memory, so the upgrade matters most for local AI. ECC is not supported. Check the module type before buying.

Can the DXP8800 Ultra take a graphics card?

Yes. The DXP8800 Ultra has one PCIe Gen4 x4 slot and 2 Thunderbolt 4 ports, so a dedicated GPU can go inside or in an external enclosure. UGREEN's guide documents dedicated GPU use in Docker, vendor confirmed. There is no OCuLink port. Without a GPU, the built-in Intel Iris Xe Graphics (96EU) does the work, hardware acceleration there is rated limited, and there is no NPU.

Does the DXP8800 Ultra support ECC memory?

No. The DXP8800 Ultra uses DDR5 SO-DIMM 5200 without ECC support. It ships with 8GB and takes up to 96GB, and at that ceiling Mistral Large 2 123B fits at Q4 (Estimated). Buyers storing business records across its 8 bays who need error-correcting memory should look for a system that supports ECC. The diskless 8GB configuration includes a 128GB system SSD, so the 2 M.2 slots stay free for data.

Does the DXP8800 Ultra run Docker?

Yes. UGOS Pro supports Docker and virtual machines. UGREEN publishes a Docker Compose guide for Ollama and Open WebUI on UGOS Pro, vendor confirmed. Docker is also the route UGREEN documents for using a dedicated GPU. Local inference on the system itself is CPU-only with no NPU, so local LLM support is rated limited rather than full.

Can the DXP8800 Ultra run Ollama?

Yes. Ollama support is documented by the manufacturer. See the capability table above for the source.

Does Open WebUI work on the DXP8800 Ultra?

Yes. Open WebUI support is documented by the manufacturer. See the capability table above for the source.

Does the DXP8800 Ultra do AI photo recognition?

Yes. AI photos support is documented by the manufacturer. See the capability table above for the source.

Can the DXP8800 Ultra run AI surveillance?

Yes. Surveillance AI support is documented by the manufacturer. See the capability table above for the source.

Does the DXP8800 Ultra have hardware AI acceleration?

Partly. Hardware acceleration support is documented by the manufacturer. See the capability table above for the source.

How fast is the DXP8800 Ultra's network?

It has 2 x 10GbE built in, and the 1x PCIe Gen4 x4 can take a faster network card. The fastest port is 2 x 10GbE.

Where do these specifications come from?

Every field records its source. Specifications were last checked on . Spot an error? Send a correction with a manufacturer link.