Home›Hardware›Xiaomi AI Cube, 80 GB — engineering prototype
Xiaomi AI Cube, 80 GB — engineering prototype
mini-PC with NPU80 GiB341 GB/s40.0 TFLOPS
You cannot buy this one. Everything else in this hub is on sale; this is an engineering prototype with no price and no release date. We measure it anyway, because we wanted to know what it would run if it ever reaches a shop, and the way to find out is the same arithmetic we use for every other machine here. Spoiler: quite a lot.
Once the system takes its share, this machine leaves 77.00 GiB for the model: room for 22 of the 26 models we measure at the reference compression, and the heaviest one that fits is nemotron-3.5-super-120b. And how fast does it answer? That’s not down to the chip, but to its 341 GB/s of memory bandwidth (our estimate: the manufacturer does not publish it).
Left for the model
Estimated
Memory bandwidth
Estimated
Models that fit
Measured by Local AI Scope
Videos: Xiaomi AI Cube, 80 GB — engineering prototype 5
The best walk-throughs we found. Press play and it opens right here.
The videos live on YouTube, and nothing loads from there until you press play (no-cookie mode). The thumbnail is served by us.
The memory you can actually use
Unified memory is shared with the operating system and everything you have open. The reserve here is deliberately conservative; you can change it in the tool.
| Memory | GiB |
|---|---|
| Memory on the box | 80 |
| Reserved for the system | −3.00 |
| Left for the model | 77.00 |
160 GB is what the D100 chip supports, not what this machine carries. The unit Xiaomi demonstrated has 80 GB, and 80 is what every figure on this page is computed from. The difference is not cosmetic: with 160 GB this machine would take 24 of the 26 models instead of 22, and the largest one it holds would be deepseek-v4-flash at 127.28 GiB instead of nemotron-3.5-super-120b at 65.29 GiB.
Memory bandwidth
Two different limits, and they are easy to mix up. While the model is writing its answer, bandwidth rules: every token means reading the weights it uses again — all of them in a dense model; in a mixture of experts, the shared layers plus only the active experts — so tokens per second is roughly bandwidth divided by that amount. While the model is reading your document, compute rules: the whole input is processed at once. That is why a laptop with no GPU can take minutes before the first word appears and then type at a tolerable pace.
341 GB/s — Estimated · the manufacturer does not publish this figure at all; ours is a conservative estimate, not a specification.
The 1.22 TB/s is a real Xiaomi figure, just not for this memory. It is the near-memory bandwidth Xiaomi quotes for the O100 accelerator, over DRAM stacked right on the compute die — a small, very fast pool, not the 80 GB where a 120B model actually sits. The bandwidth of those 80 GB is a number Xiaomi has not published, so the figure on this page is our own assumption and is labelled as one. Same story with the 200 TOPS: they belong to the O3’s NPU and count 8-bit integer operations — another unit, on another chip — so we keep them out of the compute line above. Tempting, we know.
Compute: 40.0 TFLOPS (estimate for this class of machine) — Estimated · the manufacturer does not publish this figure at all; ours is a conservative estimate, not a specification. The figures in this line do not all come from the same measure across machines, because each manufacturer publishes a different one. Compare them between machines only with that in mind.
Which models fit here, and which do not
Fitting means three things at once fit in the memory left over: the model file, its context memory, and the runtime’s working headroom. Context memory is computed from each model’s declared attention pattern layer by layer, not from the generic formula — which is why several models fit here that the generic formula would rule out.
| Model | Compression shown | Model size | 8k | 32k | 128k | Largest context that fits |
|---|---|---|---|---|---|---|
| qwen3-1.7b | Q4_K_M |
1.03 GiB | yes | yes | — | 32k |
| qwen3-4b-2507 | Q4_K_M |
2.33 GiB | yes | yes | yes | 256k |
| gemma4-e2b | Q4_K_M |
2.89 GiB | yes | yes | yes | 128k |
| gemma4-e4b | Q4_K_M |
4.63 GiB | yes | yes | yes | 128k |
| qwen3-8b | Q4_K_M |
4.68 GiB | yes | yes | — | 32k |
| granite-4.2-8b | Q4_K_M |
4.98 GiB | yes | yes | yes | 128k |
| qwen3.5-9b | Q4_K_M |
5.29 GiB | yes | yes | yes | 256k |
| gemma4-12b | Q4_K_M |
6.63 GiB | yes | yes | yes | 256k |
| gpt-oss-20b | Q4_K_M |
10.83 GiB | yes | yes | yes | 128k |
| mistral-small-24b | Q4_K_M |
13.35 GiB | yes | yes | yes | 128k |
| qwen3.8-27b | Q4_K_M |
15.33 GiB | yes | yes | yes | 256k |
| muse-glimmer-30b | Q4_K_M |
15.61 GiB | yes | yes | yes | 128k |
| qwen3.6-27b | Q4_K_M |
15.66 GiB | yes | yes | yes | 256k |
| gemma4-26b-a4b | Q4_K_M |
15.78 GiB | yes | yes | yes | 256k |
| gemma4-31b | Q4_K_M |
17.07 GiB | yes | yes | yes | 256k |
| qwen3-coder-30b | Q4_K_M |
17.28 GiB | yes | yes | yes | 256k |
| qwen3.6-35b-a3b | Q4_K_M |
20.61 GiB | yes | yes | yes | 256k |
| flint-40b | Q4_K_M |
23.61 GiB | yes | yes | yes | 256k |
| nemotron-3.5-lightning-30b | Q4_K_M |
23.73 GiB | yes | yes | yes | 256k |
| gpt-oss-120b | Q4_K_M |
58.46 GiB | yes | yes | yes | 128k |
| llama4-scout-17b | Q4_K_M |
60.87 GiB | yes | yes | yes | 128k |
| nemotron-3.5-super-120b | Q4_K_M |
65.29 GiB | yes | yes | yes | 256k |
| qwen3.8-flash-next | IQ4_XS |
87.25 GiB | no | no | no | — |
| deepseek-v4-flash | IQ4_XS |
127.28 GiB | no | no | no | — |
| glm-5.3-flash | Q4_K_M |
187.04 GiB | no | no | no | — |
| glm-5.2 | Q4_K_M |
433.83 GiB | no | no | no | — |
Sizes are the real published files at the reference compression, one step per row: Q4_K_M where the author publishes it, the nearest neighbour where they do not. Every row states which one it is showing. Q4_K_M is our quality floor: below it the loss is audible in the answers. A model that would only fit here at a harsher compression is listed as not fitting, on purpose. A dash in a context column means the model itself does not offer that context, so there is nothing to fit.
Models that fit — 22 of the 26 models measured
- qwen3-1.7b
Q4_K_M— context 32k tokens - qwen3-4b-2507
Q4_K_M— context 256k tokens - gemma4-e2b
Q4_K_M— context 128k tokens - gemma4-e4b
Q4_K_M— context 128k tokens - qwen3-8b
Q4_K_M— context 32k tokens - granite-4.2-8b
Q4_K_M— context 128k tokens - qwen3.5-9b
Q4_K_M— context 256k tokens - gemma4-12b
Q4_K_M— context 256k tokens - gpt-oss-20b
Q4_K_M— context 128k tokens - mistral-small-24b
Q4_K_M— context 128k tokens - qwen3.8-27b
Q4_K_M— context 256k tokens - muse-glimmer-30b
Q4_K_M— context 128k tokens - qwen3.6-27b
Q4_K_M— context 256k tokens - gemma4-26b-a4b
Q4_K_M— context 256k tokens - gemma4-31b
Q4_K_M— context 256k tokens - qwen3-coder-30b
Q4_K_M— context 256k tokens - qwen3.6-35b-a3b
Q4_K_M— context 256k tokens - flint-40b
Q4_K_M— context 256k tokens - nemotron-3.5-lightning-30b
Q4_K_M— context 256k tokens - gpt-oss-120b
Q4_K_M— context 128k tokens - llama4-scout-17b
Q4_K_M— context 128k tokens - nemotron-3.5-super-120b
Q4_K_M— context 256k tokens
Models that do not fit — 4 of the 26 models measured
- qwen3.8-flash-next
IQ4_XS— short by 11.02 GiB · withllama.cppv0.5.0 and later leaving its n-gram table on disk, it does fit, up to 256k - deepseek-v4-flash
IQ4_XS— short by 51.29 GiB - glm-5.3-flash
Q4_K_M— short by 110.80 GiB - glm-5.2
Q4_K_M— short by 358.12 GiB
The machines on either side
- Mac with M4 Max and 64 GB — 61.00 GiB left for the model, 20 / 26 models
- Mac mini with M5 Pro and 64 GB — 61.00 GiB left for the model, 20 / 26 models
- PC with Ryzen AI Max+ 395 and 128 GB unified — 96.00 GiB left for the model, 23 / 26 models
- Mac Studio with M5 Max and 128 GB — 125.00 GiB left for the model, 23 / 26 models
Price and availability
What a machine costs decides as much as what fits inside it, so the price belongs on this page. What follows is only what the manufacturer publishes, with the market it applies to and the day it was read.
There is no price: this is an engineering prototype, not a product on sale.
If you came for its price: what you can buy today with 64 to 192 GB instead, and which models each of those machines runs, is in Xiaomi AI Cube alternatives.
What we earn on this page
Nothing, today. There is no affiliate link on this page and no commercial agreement behind any figure on it. If a buying link ever appears here, it will be marked as such and the commission declared in this same block. Two rules do not change on that day: the ranking on this site is computed before commercial availability is applied, and «best» never means «pays most».