Serving Qwen 3.8 27B with Unsloth
Qwen 3.8 27B has 27 billion parameters. Unsloth targets fine-tuning and quantisation. This page gives the memory required and the platforms that qualify.
What is the minimum memory?
Qwen 3.8 27B totals 27 billion parameters. In NVFP4 its weights take about 15.5 GB including the runtime margin: 27 billion parameters × 0.5 byte (NVFP4) = 13.5 GB of weights; × 1.15 runtime margin = 15.5 GB. In FP8 the footprint doubles, to about 31.0 GB (27 billion parameters × 1 byte (FP8) = 27.0 GB of weights; × 1.15 runtime margin = 31.0 GB.) The attention cache is not calculated on this page. These models’ architectures (compressed latent, linear or sparse attention, Mamba layers) share no per-token formula; it is measured on the machine, with the target context and concurrency [TO BE MEASURED].
What is Unsloth for?
Unsloth targets fine-tuning and quantisation. See the Unsloth and Qwen 3.8 27B fact sheets.
On which platforms?
- Qwen 3.8 27B on DGX Spark
- Qwen 3.8 27B on Mac Studio Ultra
- Qwen 3.8 27B on DGX Station
- Qwen 3.8 27B on RTX PRO 6000 server
- Qwen 3.8 27B on H200 SXM server
- Qwen 3.8 27B on B200 SXM
How much memory for Qwen 3.8 27B with Unsloth?
About 15.5 GB in NVFP4 for the weights, excluding the attention cache.
Does Unsloth suit Qwen 3.8 27B?
Unsloth serves Qwen 3.8 27B as soon as the machine offers at least 15.5 GB of memory, the weight footprint in NVFP4. 8 of the 8 platforms in the catalogue meet that bar.
Method and sources
Memory footprints are calculated, not measured: a reading on real hardware may differ depending on the engine and the exact weight format. Every figure below carries its source and the date of its last check.
- Qwen 3.8 27B: 27 billion parameters, 262,144-token context, Apache 2.0 licence. Hugging Face repository (config.json, API, model card), checked on 2 September 2026. Dense model: every parameter computes on each token. The BF16 repository holds 27.8 billion elements (56 GB).
- Bytes per parameter: NVFP4 0.5 (4 bits per parameter, hence 0.5 byte; 4-bit published weights weigh 0.54 to 0.56 byte per parameter with their scales (DeepSeek V4 Pro 865 GB for 1,599 billion, Kimi K3 1,561 GB for 2,780 billion)); FP8 1 (8 bits per parameter (E4M3), hence 1 byte, block scales not counted). Source: connaissance/faits.yaml, quantifications family, checked on 2026-08-31.
- Runtime margin × 1.15: QDNA operating assumption, not measured: 15% above the weights for activations, buffers and fragmentation.
- “Fits” threshold at 75% of memory: QDNA assumption, not measured, which keeps the remaining quarter for the attention cache and concurrency.