rtx4090

Qwen3.8-Flash-Next on 2× RTX 3090: with 128 GB RAM up to 4,191 tok/s prefill · 111.5 tok/s decode (131K prompt), full 256K window at 2,865 tok/s; with 64 GB RAM 3,410 tok/s prefill · 84 tok/s decode. New: opt-in uncensored mode (runtime abliteration, no new weights).

Working Guide for Passthrough tested on intel i7 13700k and RTX 4090

A hands-on guide for AI builders: make your own RTX 4090D/5090 GPU server that’s fast and efficient.

Device-native Qwen3-TTS inference for AMD RX 7900 XTX and NVIDIA RTX 4090 with explicit HIP/CUDA kernels, streaming and Resident execution.

RTX 4090 performance optimization nodes for ComfyUI

Open-source compute driver for RTX 40 series GPUs on macOS - Pure AI/ML power

Experimental 256 MiB BAR1 P2P on 48 GB RTX 4090: alignment and budget fixes, correctness probes, and vLLM measurements.

Rent cloud GPUs from your terminal. Deploy in 30 seconds. npm i -g computegpu

🌠 End-to-end deep learning systems study: custom PyTorch ResNet-18 built from scratch, CUDA kernel profiling with torch.profiler, and multi-GPU PyTorch DistributedDataParallel (DDP) benchmarking on 2x NVIDIA RTX 4090 GPUs (achieving 1.72× speedup & 86% scaling efficiency).

Validated Qwen3.8-27B agent-serving profile for one 48 GiB RTX 4090: DFlash2, KVarN, prefix caching, long-context fairness, and reproducible benchmarks.

Experimental Strata fork for Qwen3.8-Flash-Next with paired RTX 4090 benchmarks and opt-in CPU decode optimizations

Matrix-free 3D SIMP topology optimization with fused gather-GEMM-scatter CUDA kernels on NVIDIA RTX 4090. Companion code for arXiv:2604.18020.

GRPO training that runs until you stop it on a single RTX4090 with vllm 0.26.0 (Linux Only).

Extreme Rank 2048 Van Gogh style LoRA for FLUX2-dev Trained on a single 24GB RTX 4090 via Viking Engine async memory manager

Dynamic open-source Google Sheets tables inspired by the famous Hive Systems ones. Just input the H/s/GPU data.

Reproducible Qwen3.8-27B GGUF benchmark study on an NVIDIA RTX 4090 running Windows

Practical RTX 4090 local LLM workflow benchmark for coding, repair loops, browser runtime validation, vision extraction, and context/VRAM limits.

Viking Engine: High-performance LoRA training engine for consumer hardware. Server-class speed, universal model support, optimized for RTX 4090.

Run Qwen3-TTS 12Hz models natively on 24GB AMD RX 7900 XTX or NVIDIA RTX 4090 GPUs with minimal latency.