96GB max RAM · 5 bays · 10GbE
AINAS Score 65
$919.00
Amazon US · Verified
Ships from the US; delivery and import charges shown at checkout

AI ReadyASUSTOR
Why AI Ready: AMD Ryzen 7 Pro 7745 (8 cores, 2-core Radeon iGPU, no NPU) with a 192GB memory ceiling meets the 64GB memory line; GPU path is not documented (PCIe 5.0 x8 and x4 slots listed for NIC and SAS cards, and ASUSTOR says its NAS has no discrete GPU), so not native.
Updated
Can the ASUSTOR Lockerstor 12R Pro Gen2 (AS7212RDX) run local AI?
Only lightly. Expect 70B class models at higher-quality quantization, or large mixture-of-experts models at 4-bit (Estimated) on the ASUSTOR Lockerstor 12R Pro Gen2 (AS7212RDX): AMD Ryzen 7 Pro 7745 does the work with up to 192GB of memory, which is fine for experiments rather than a daily assistant.
Without any GPU option, the processor carries the whole inference load.
Local LLM support is documented by the manufacturer.
The Lockerstor 12R Pro Gen2 suits small businesses and home lab users who want a 12-bay rackmount NAS that can also host private AI services on the CPU. It has an AMD Ryzen 7 Pro 7745 with 8 cores, 16GB of DDR5-4800 ECC memory installed and a ceiling of 192GB across four slots. It has 2 x 10GbE and 2 x 1GbE, one M.2 slot, Docker and VirtualBox support, and a 5 year warranty. Ollama and Open WebUI are vendor confirmed on ADM 5. Estimated: Qwen3 235B A22B fits at Q4 once memory is expanded. The AINAS Score is 64.
The one thing that rules it out for some buyers is the lack of a GPU path. ASUSTOR states its NAS models have no discrete GPU, and the two PCIe 5.0 slots are documented for network and SAS cards only. Local inference is CPU-only, and ASUSTOR recommends an external GPU host for heavier work. Anyone who wants GPU-accelerated LLMs, image generation or video work on the NAS itself should look for a system with a documented PCIe GPU slot, OCuLink or Thunderbolt connection, or a higher-TOPS NPU. The Lockerstor 12R Pro Gen2 has none of these. It has no NPU and no Thunderbolt, USB4 or OCuLink ports.
No verified price
We have not verified a listing for the ASUSTOR Lockerstor 12R Pro Gen2 (AS7212RDX) at a retailer we track. The specifications below stand on their own; a price appears here as soon as a listing is confirmed.
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.
Each capability shows how it was confirmed: Verified by AINAS, Vendor confirmed, Community confirmed or Unverified.
| Capability | Supported | Evidence | Note |
|---|---|---|---|
| Local LLM | Limited | Vendor confirmed | ASUSTOR App Central lists Ollama + Open WebUI (v0.11.4.r01, ADM 5.0+) for the AS72 series. CPU-only inference on AMD Ryzen 7 Pro 7745. ASUSTOR recommends an external GPU host for inference. |
| Ollama | Yes | Vendor confirmed | Ollama + Open WebUI app lists AS72 as a supported series. |
| Open WebUI | Yes | Vendor confirmed | Bundled with the Ollama + Open WebUI App Central package. |
| RAG | Limited | Vendor confirmed | ASUSTOR markets NAS as RAG knowledge base storage with vector databases (Chroma, pgvector, Weaviate) via Docker. It recommends an external compute host for inference. |
| AI search | No | Unverified | No AI or semantic file search feature found in ADM 5.1 material. |
| GPU expansion | No | Vendor confirmed | ASUSTOR states its NAS models do not feature a discrete GPU. PCIe slots are documented for network and SAS cards only on the spec page. |
| Hardware acceleration | Limited | Unverified | CPU has a small Radeon iGPU, but ASUSTOR documents no GPU compute or transcoding use on this model. No NPU. HDMI is service only. |
| AI photos | Yes | Vendor confirmed | Photo Gallery app lists AS72; ADM 5.1 adds facial recognition grouping powered by locally run AI. |
| AI video | No | Unverified | No AI video feature found in ADM documentation. |
| Surveillance AI | Yes | Vendor confirmed | Surveillance Center apps (1663 and 1661) list AS72; 4 free camera licences, more purchasable, per spec page. |
Calculated from the Lockerstor 12R Pro Gen2 (AS7212RDX)'s memory and each model's published size at each quantization. Rows marked Estimated are calculations, not benchmarks. Check any model.
| Model | Best quantization | Runs? | Memory needed | Expected performance | Verification |
|---|---|---|---|---|---|
| Llama 3.2 3BLlama, 3.21B | Q8 | Yes | 13GB | Not measured | Estimated |
| Llama 3.1 8BLlama, 8.03B | Q8 | Yes | 19GB | Not measured | Estimated |
| Qwen3 8BQwen, 8.19B | Q8 | Yes | 19GB | Not measured | Estimated |
| Gemma 3 12BGemma, 11.8B | Q8 | Yes | 24GB | Not measured | Estimated |
| Qwen3 14BQwen, 14.8B | Q8 | Yes | 28GB | Not measured | Estimated |
| Phi-4 14BPhi, 14.7B | Q8 | Yes | 28GB | Not measured | Estimated |
| Mistral Small 3 24BMistral, 23.6B | Q8 | Yes | 38GB | Not measured | Estimated |
| Gemma 3 27BGemma, 27B | Q8 | Yes | 44GB | Not measured | Estimated |
| Qwen3 30B A3BQwen, 30.5B | Q8 | Yes | 47GB | Not measured | Estimated |
| Qwen3 32BQwen, 32.8B | Q8 | Yes | 50GB | Not measured | Estimated |
| DeepSeek-R1 32B (Qwen distill)DeepSeek, 32.8B | Q8 | Yes | 50GB | Not measured | Estimated |
| Llama 3.3 70BLlama, 70.6B | Q8 | Yes | 98GB | Not measured | Estimated |
| Llama 4 Scout 17Bx16ELlama, 109B | Q8 | Yes | 149GB | Not measured | Estimated |
| Qwen2.5 72BQwen, 72.7B | Q8 | Yes | 101GB | Not measured | Estimated |
| Mistral Large 2 123BMistral, 123B | Q8 | Yes | 164GB | Not measured | Estimated |
One verdict per workload, from the Lockerstor 12R Pro Gen2 (AS7212RDX)'s use-case scores and capability rows. Unknown means we cannot confirm it, not that it probably works.
The Lockerstor 12R Pro Gen2 is a good fit for local LLM work, with limits. ASUSTOR lists Ollama and Open WebUI for ADM 5, but inference is CPU-only, and large models need the memory upgraded beyond 16GB.
Private RAG is a good fit here. ASUSTOR describes vector databases such as Chroma, pgvector and Weaviate running through Docker, which the system supports, though RAG support is marked limited and inference is best handled by an external host.
AI search is a poor fit. No AI or semantic file search feature was found in ADM 5.1 material, and that capability row is Unverified, so it is unconfirmed rather than supported.
AI photos is a good fit. The Photo Gallery app lists the AS72 series, and ADM 5.1 adds facial recognition grouping powered by locally run AI, which ASUSTOR confirms.
Surveillance is a good fit. The Surveillance Center apps list the AS72 series, and ASUSTOR confirms 4 free camera licences, with extra licences sold separately.
Home lab use is a good fit. Docker and VirtualBox are both listed for the system, and it offers up to 192GB of ECC memory, 12 bays and two 10GbE ports.
Business use is a good fit. ECC memory, redundant 550W supplies, 12 bays, two 10GbE ports and a 5 year warranty suit shared storage and light services in a rack.
Creative AI is rated good, but the GPU side needs care. GPU expansion is marked no, and hardware acceleration is limited and Unverified, so work depends on CPU inference or an external GPU host.
Plan on upgrading. 16GB is fitted as standard, while the Lockerstor 12R Pro Gen2 (AS7212RDX) accepts 192GB of DDR5-4800 ECC UDIMM, ECC included, which is the difference between 7B to 8B class models and 70B class models at higher-quality quantization, or large mixture-of-experts models at 4-bit (Estimated).
Our estimate for the biggest model that fits in 192GB is about 306B parameters at 4-bit, after setting aside 8GB for the system. Long context windows reduce that figure (Estimated).
ECC support protects against silent memory errors on a box that runs around the clock.
A graphics card is not an option on the Lockerstor 12R Pro Gen2 (AS7212RDX), which has neither a PCIe slot nor OCuLink. Inference runs on the CPU, with AMD Radeon Graphics (2 cores) handling media tasks, so keep to small models.
As the storage half of a RAG setup it works well; as the model host it is limited. Small embedding models run fine on the Lockerstor 12R Pro Gen2 (AS7212RDX), while the answering model needs more compute than it has.
For network access it has 2 x 10GbE + 2 x 1GbE. Fast enough that the network will not be the bottleneck.
| Configuration | RAM | SSD | Notes |
|---|---|---|---|
| 16GB | 16GB | None | Diskless; 1x 16GB ECC DDR5-4800 pre-installed, 4 slots. |
The closest ASUSTOR model by bay count and AINAS Score. Only the specifications that differ are listed.
Best fit: home lab owners who want one box for storage, containers, media and a few AI services. That is where the Lockerstor 12R Pro Gen2 (AS7212RDX) scores highest (Home Lab 75/100), followed by Business (72/100).
Less suited to buyers who expect to add a graphics card later.
Other AI Ready systems closest in overall score.
96GB max RAM · 5 bays · 10GbE
AINAS Score 65
$919.00
Amazon US · Verified
Ships from the US; delivery and import charges shown at checkout
128GB max RAM · 18 bays · 2 x 10GbE
AINAS Score 63
$5,099.00
Amazon US · Verified
Ships from the US; delivery and import charges shown at checkout
128GB max RAM · 22 bays · 2 x 10GbE
AINAS Score 63
$6,209.00
Amazon US · Verified
Ships from the US; delivery and import charges shown at checkout
Workloads it suits
Yes, Estimated, but only after a memory upgrade. Qwen3 235B A22B needs 142GB at Q4, and the system takes up to 192GB. At Q8 it needs 250GB, so it does not fit (Estimated). This is a calculation, not a benchmark. Inference is CPU-only, and the system ships with 16GB.
192GB is the maximum. The system ships with one 16GB DDR5-4800 ECC UDIMM and has 4 slots, so it can be expanded. Larger models need that extra memory, since the 16GB installed only holds smaller ones. Check the memory type before buying modules: it takes ECC UDIMMs.
Yes. It uses DDR5-4800 ECC UDIMM memory, and the 16GB module installed is ECC. ECC helps catch and correct memory errors, which matters for a NAS that stores data and runs services around the clock. Any memory you add should also be ECC UDIMM.
No. ASUSTOR states its NAS models do not feature a discrete GPU. The PCIe 5.0 x8 and x4 slots are documented for optional 10/25/40/50GbE and SAS cards. It also has no OCuLink, Thunderbolt or USB4 port for an external GPU.
12 drive bays and 1 M.2 slot. The bays sit in a rackmount chassis, and the single M.2 slot takes one NVMe drive. The only configuration is diskless with 16GB of ECC memory, so you supply every drive. For more drive connectivity, the PCIe 5.0 x8 and x4 slots are documented for SAS cards as well as 10/25/40/50GbE network cards. Power comes from a 550W supply, and the warranty is 5 years.
Yes to both. The Docker Engine app lists the AS72 series, and the VirtualBox app lists it too. Docker is what ASUSTOR points to for running vector databases for RAG. The system runs ADM 5, and the 8-core Ryzen 7 Pro 7745 handles the workloads.
Yes. Ollama support is documented by the manufacturer. See the capability table above for the source.
Yes. Open WebUI support is documented by the manufacturer. See the capability table above for the source.
Yes. AI photos support is documented by the manufacturer. See the capability table above for the source.
Yes. Surveillance AI support is documented by the manufacturer. See the capability table above for the source.
Partly. Hardware acceleration support is not yet verified. See the capability table above for the source.
It has 2 x 10GbE + 2 x 1GbE built in, and the 1x PCIe 5.0 x8 (x8 link), 1x PCIe 5.0 x4 (x4 link); for optional 10/25/40/50GbE and SAS cards can take a faster network card. The fastest port is 2 x 10GbE.
Every field records its source. Specifications were last checked on . Spot an error? Send a correction with a manufacturer link.