llm-kb
← Back to social
Community

User benchmarks AMD EPYC 9374F for LLM inference, finds 48-thread sweet spot

A user replaced their EPYC 9135 with a cheap 9374F (8 CCDs) for LLM inference. Initial benchmarks showed no decoding advantage until they used 48 threads; 64 or 32 threads performed worse than the 9135 in some scenarios. The 9374F is worse for gaming.

6 engagement·1 source·Sat, Jul 11, 2026, 01:03 PM
The user found a deal on a 9374F to replace their bottle-necked 9135. Benchmarks with Unsloth GLM-5.2-UD-IQ4_XS on llama.cpp with 4800 DDR5 showed that using 48 threads gave the best performance, while 64 or 32 threads sometimes underperformed the 9135. The 9374F is noted to be much worse for gaming.

Entities

llama.cpp(tool)Reddit(company)AMD EPYC 9374F(model)AMD EPYC 9135(model)Unsloth GLM-5.2-UD-IQ4_XS(model)

Related

CommunitySun, Jul 12, 2026, 04:38 AM

User tests LLMs on Intel 285HX CPU-only mini-PC, finds Llama-Swap incompatible with SYCL

A user set up a homelab server on an MS-02 mini-PC with an Intel Core Ultra 285HX and 64GB RAM, testing Qwen3, Qwen3.6, and Gemma4 via Llama.cpp. They found Llama-Swap's quick swapping helpful but incompatible with SYCL, complicating testing. The post shares early impressions of running LLMs on a CPU-only system.

4 engagement·1 source·reddit
CommunitySun, Jul 12, 2026, 04:30 AM

User seeks advice on adding second cheap GPU to run larger local models

A Reddit user running Gemma 4 26B-A4B at 12-15 t/s on an RTX 3060 (12 GB) finds the model insufficiently intelligent and wants to upgrade to a 31B model, which runs at only 1.5 t/s. They ask the community about the benefits of adding a second cheap GPU to improve performance for local LLM inference.

7 engagement·1 source·reddit
ProductfeaturedSun, Jul 5, 2026, 08:45 PM

Rust+CUDA LLM inference engine for RTX 5090, optimized with NVFP4, MoE, and speculative decoding

A from-scratch LLM inference engine written in Rust and CUDA, specifically tuned for the RTX 5090 Laptop GPU (Blackwell sm_120a). It supports NVFP4 quantization, mixture-of-experts (MoE), and multi-token prediction (MTP) speculative decoding, achieving up to 1.6x speedup over llama.cpp on target models. Designed for developers who need maximum inference performance on a single high-end laptop GPU.

260 engagement·1 source·github
CommunitySun, Jul 12, 2026, 05:43 AM

User runs 100B+ MoE LLMs on low-end laptop using NVMe swap and Q3 quantization

A Reddit user with a low-spec laptop (i7-8750H, 20GB RAM, GTX 1050 4GB) reports successfully running 100B+ parameter MoE models by offloading parameters to a Samsung NVMe SSD via mmap, using Q3 quantization and quantized KV cache (Q4_0). They note that dense models are unusable on their hardware, but MoE models work with experts offloaded to CPU.

22 engagement·1 source·reddit
BenchmarkSun, Jul 12, 2026, 10:47 AM

User benchmarks 4x RTX 5060 Ti with SGLang for Qwen3.6 27B, finds better concurrency than vLLM

A user shared benchmark results showing that SGLang handles higher concurrency better than vLLM when running Qwen3.6 27B (INT8 with bf16 KV cache) on a 4x RTX 5060 Ti (64GB VRAM) setup. The test achieved 200 successful requests at 8 concurrency over 348.87 seconds, processing 61,870 input tokens and generating 44,525 tokens. This provides a practical reference for others considering multi-GPU configurations with these consumer cards.

3 engagement·1 source·reddit