Builds / RTX 5090
Ryzen 9 9950X3D + RTX 5090 + 128 GB DDR5: four big MoEs with experts offloaded to RAM (Qwen3.8 Flash-Next 78.8 tok/s)
Source: dev.to ↗📏 How measured: reported by the original source; RigBase re-checked the numbers at that source
- GPU(s)
- RTX 5090
Amazon · eBay (used) · Newegg - VRAM / unified memory (GB)
- 32
- CPU
- AMD Ryzen 9 9950X3D (16 cores)
Amazon · eBay (used) · Newegg - Motherboard
- not stated
- RAM
- 128 GB DDR5-5600 (4x32 GB)
Amazon · eBay (used) · Newegg - PSU
- not stated
- Other
- Arch Linux, NVIDIA 615.78, CUDA 13.4; GPU also drives the desktop (3 GiB kept free); runtimes keep hot experts in VRAM and the rest in system RAM
- Reported total cost
- not stated
- Power
- not stated
Measured speed (generation tokens/sec)
| Model | Quant | Engine | tok/s | Details |
|---|---|---|---|---|
| Qwen3.8 Flash-Next | Unsloth UD-Q5_K_XL, MTP drafting, experts offloaded to RAM | Strata (296b59d) | 78.8 | ⚡ spec-decode (MTP) |
| Ornith 1.5 122B | EXL3 4 bpw, 180/256 experts per layer in RAM, no drafting | TabbyAPI + ExLlamaV3 1.6.2 | 46.4 | |
| GLM-5.3 Flash | EXL3 3.05 bpw, MTP drafting, experts offloaded to RAM | Project Maya 1.1.21 | 30.1 | ⚡ spec-decode (MTP) |
| DeepSeek V4.1 Flash | Maya-S (experts 2-3 bit), partly streamed from SSD | Project Maya 1.1.21 (maya-ds4) | 24.9 |
Seppe Gadeyne (dev.to), updated measurements 22 Oct 2026; median of 8 runs (Dutch prose + coding prompt), greedy, 512 output tokens. All four are MoEs with CPU offload of experts to system RAM: Qwen3.8 Flash-Next, Ornith, GLM-5.3 Flash and DeepSeek V4.1 Flash. Prefill at ~125K prompt: 4,252 / 1,508 / 2,445 / 2,570 tok/s in the same order. Versions: Strata 296b59d, Project Maya 1.1.21, TabbyAPI d6e6227 + ExLlamaV3 1.6.2.
Current GPU street price (estimate, not the builder's cost)
New retail prices in Germany incl. VAT; used cards are often much cheaper. Prices change daily.
- 1× RTX 5090: ≈€7,140 each (≈$8,003), new, approx. price as of 2026-10-24, lowest listing on billiger.de (MSI GeForce RTX 5090 (lowest variant))
New retail prices in Germany incl. VAT; used cards are often much cheaper. Prices change daily.
Source: https://dev.to/seppegadeyne/four-big-local-llms-on-one-rtx-5090-and-128-gb-of-ram-benchmarked-side-by-side-2h8j
Blank fields = not stated in the source. We never estimate build data; GPU street prices above are separately sourced and dated.
Rent before you buy: try this model on a cloud GPU for a few dollars first: Runpod (referral link; sign up with Google to get the new-user bonus)
Quick answers: Can RTX 5090 run DeepSeek V4.1 Flash? · Can RTX 5090 run GLM-5.3 Flash? · Can RTX 5090 run Ornith 1.5 122B? · Can RTX 5090 run Qwen3.8 Flash-Next?