Ollama runs out of memory

DevOps and infra2 posts from 2 people2d active+11 post in the last 7 days, 1 the 7 days before (steady)Unverified

Posts per day2 posts, Oct 8 to Oct 21

The posts behind it

2, newest first

GitHub issuesOct 15 ROCm: gemma4:26b-a4b with experts offloaded to the CPU aborts in hipGraphInstantiate (out of memory), then segfaults on every load; GGML_CUDA_DISABLE_GRAPHS=1 avoids it...processing and `created context checkpoint 2 of 32`, including with `num_batch 128`. **Workaround:** `Environment="GGML_CUDA_DISABLE_GRAPHS=1"` in the service. No crashes since, in long-prompt tests (up to about 7,000 tokens) and a run of real prompts, and no slowdown we could...ollama/ollama · dspl1236Open ↗GitHub issuesOct 8 Clef / Clef Flash: n_ubatch is forced to n_ctx, so default num_ctx 16384 fails to load (OOM / GGML_ASSERT)...Flash: n_ubatch is forced to n_ctx, compute buffer scales with num_ctx and the model fails to load with the default num_ctx 16384 Windows 11, AMD Ryzen AI 9 HX 370, Radeon 890M (iGPU, gfx1150, ~27 GB UMA), 7.6 GB system RAM visible to Windows...ollama/ollama · Altandreas1 repliesOpen ↗

Companies and products named

Named2 posts

Ollama

About this problem

Evidence

2 posts from 2 people in 1 place, about 1 a week over 8 days. Mostly on ollama/ollama.

Frustration Frustration 0 of 3· Seen on GitHub issues

How it was grouped

Posts that state a pain and match the "oom" rule about Ollama. First post Oct 8, 2026, latest Oct 15, 2026. Unverified: fewer than 3 posts, or one person or place only. Method

The same problem elsewhere

Elsewhere9 postsGemini runs out of memory Elsewhere9 postsJobs and services run out of memory Elsewhere5 postsGitHub runs out of memory Elsewhere4 postsNVIDIA runs out of memory Elsewhere4 postsDuckDB runs out of memory Elsewhere4 postsApache Iceberg runs out of memory Elsewhere3 postsPolars runs out of memory Elsewhere2 postsvLLM runs out of memory Elsewhere2 postsllama.cpp runs out of memory Elsewhere2 postsApache Airflow runs out of memory Elsewhere2 postsMistral runs out of memory

Other Ollama problems

Same product15 postsOllama upgrades break working setups Same product6 postsTrouble with Ollama pull Same product27 postsRunning Ollama locally hits memory and speed limits Same product9 postsOllama error messages are cryptic Same product13 postsTrouble with Ollama chat Same product2 postsTrouble with Ollama silently Same product34 postsOllama users keep filing bugs Same product5 postsTrouble with Ollama empty Same product4 postsServing models on Ollama is harder than it should be Same product4 postsOllama runs out of context on real work Same product3 postsTrouble with Ollama text Same product3 postsOllama fails to return valid structured output

History

AddedOct 20, 2026Ollama runs out of memory (2 posts)