ollama - 💡(How to fix) Fix CUDA error on RTX 3060 Laptop 6GB with qwen3.5:4b — fixed by updating from 0.17.1 to 0.17.7 [4 comments, 2 participants]

Error: listen tcp 0.0.0.0:11435: bind: Only one usage of each socket address (protocol/network address/port) is normally permitted. time=2026-03-11T17:41:44.284-04:00 level=INFO source=routes.go:1663 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GGML_VK_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:0 OLLAMA_DEBUG:INFO OLLAMA_EDITOR: OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11435 OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:C:\Users\USER\.ollama\models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NO_CLOUD:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_REMOTES:[ollama.com] OLLAMA_SCHED_SPREAD:false OLLAMA_VULKAN:false ROCR_VISIBLE_DEVICES:]" time=2026-03-11T17:41:44.285-04:00 level=INFO source=routes.go:1665 msg="Ollama cloud disabled: false" time=2026-03-11T17:41:44.292-04:00 level=INFO source=images.go:473 msg="total blobs: 51" time=2026-03-11T17:41:44.293-04:00 level=INFO source=images.go:480 msg="total unused blobs removed: 0" time=2026-03-11T17:41:44.294-04:00 level=INFO source=routes.go:1718 msg="Listening on [::]:11435 (version 0.17.1)" time=2026-03-11T17:41:44.295-04:00 level=INFO source=runner.go:67 msg="discovering available GPUs..." time=2026-03-11T17:41:44.309-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 42901" time=2026-03-11T17:41:44.517-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 42910" time=2026-03-11T17:41:44.711-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 42916" time=2026-03-11T17:41:44.800-04:00 level=INFO source=runner.go:106 msg="experimental Vulkan support disabled. To enable, set OLLAMA_VULKAN=1" time=2026-03-11T17:41:44.802-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 42922" time=2026-03-11T17:41:44.802-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 42921" time=2026-03-11T17:41:44.988-04:00 level=INFO source=types.go:42 msg="inference compute" id=GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 filter_id="" library=CUDA compute=8.6 name=CUDA0 description="NVIDIA GeForce RTX 3060 Laptop GPU" libdirs=ollama,cuda_v13 driver=13.1 pci_id=0000:01:00.0 type=discrete total="6.0 GiB" available="4.4 GiB" time=2026-03-11T17:41:44.988-04:00 level=INFO source=routes.go:1768 msg="vram-based default context" total_vram="6.0 GiB" default_num_ctx=4096 [GIN] 2026/03/11 - 17:42:04 | 200 | 4.8694ms | 127.0.0.1 | GET "/api/tags" [GIN] 2026/03/11 - 17:42:05 | 200 | 146.8504ms | 127.0.0.1 | POST "/api/show" [GIN] 2026/03/11 - 17:42:05 | 200 | 144.9813ms | 127.0.0.1 | POST "/api/show" time=2026-03-11T17:42:05.719-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --port 16818" time=2026-03-11T17:42:05.894-04:00 level=INFO source=cpu_windows.go:148 msg=packages count=1 time=2026-03-11T17:42:05.894-04:00 level=INFO source=cpu_windows.go:164 msg="efficiency cores detected" maxEfficiencyClass=1 time=2026-03-11T17:42:05.894-04:00 level=INFO source=cpu_windows.go:195 msg="" package=0 cores=14 efficiency=8 threads=20 time=2026-03-11T17:42:05.986-04:00 level=INFO source=server.go:247 msg="enabling flash attention" time=2026-03-11T17:42:05.986-04:00 level=INFO source=server.go:431 msg="starting runner" cmd="C:\Users\USER\AppData\Local\Programs\Ollama\ollama.exe runner --ollama-engine --model C:\Users\USER\.ollama\models\blobs\sha256-dec52a44569a2a25341c4e4d3fee25846eed4f6f0b936278e3a3c900bb99d37c --port 16823" time=2026-03-11T17:42:05.989-04:00 level=INFO source=sched.go:491 msg="system memory" total="39.7 GiB" free="25.6 GiB" free_swap="22.4 GiB" time=2026-03-11T17:42:05.989-04:00 level=INFO source=sched.go:498 msg="gpu memory" id=GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 library=CUDA available="3.9 GiB" free="4.4 GiB" minimum="457.0 MiB" overhead="0 B" time=2026-03-11T17:42:05.989-04:00 level=INFO source=server.go:757 msg="loading model" "model layers"=33 requested=-1 time=2026-03-11T17:42:06.024-04:00 level=INFO source=runner.go:1411 msg="starting ollama engine" time=2026-03-11T17:42:06.025-04:00 level=INFO source=runner.go:1446 msg="Server listening on 127.0.0.1:16823" time=2026-03-11T17:42:06.035-04:00 level=INFO source=runner.go:1284 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Enabled KvSize:4096 KvCacheType: NumThreads:6 GPULayers:33[ID:GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 Layers:33(0..32)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2026-03-11T17:42:06.069-04:00 level=INFO source=ggml.go:136 msg="" architecture=qwen35 file_type=Q4_K_M name="" description="" num_tensors=883 num_key_values=52 load_backend: loaded CPU backend from C:\Users\USER\AppData\Local\Programs\Ollama\lib\ollama\ggml-cpu-alderlake.dll ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA GeForce RTX 3060 Laptop GPU, compute capability 8.6, VMM: yes, ID: GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 load_backend: loaded CUDA backend from C:\Users\USER\AppData\Local\Programs\Ollama\lib\ollama\cuda_v13\ggml-cuda.dll time=2026-03-11T17:42:06.155-04:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX_VNNI=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=750,800,860,870,890,900,1000,1030,1100,1200,1210 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(clang) time=2026-03-11T17:42:06.782-04:00 level=INFO source=runner.go:1284 msg=load request="{Operation:fit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Enabled KvSize:4096 KvCacheType: NumThreads:6 GPULayers:19[ID:GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 Layers:19(13..31)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2026-03-11T17:42:07.109-04:00 level=INFO source=runner.go:1284 msg=load request="{Operation:alloc LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Enabled KvSize:4096 KvCacheType: NumThreads:6 GPULayers:19[ID:GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 Layers:19(13..31)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2026-03-11T17:42:07.732-04:00 level=INFO source=runner.go:1284 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Enabled KvSize:4096 KvCacheType: NumThreads:6 GPULayers:19[ID:GPU-21f58bd8-9811-7bcf-7a46-f0f7345d0d29 Layers:19(13..31)] MultiUserCache:false ProjectorPath: MainGPU:0 UseMmap:false}" time=2026-03-11T17:42:07.732-04:00 level=INFO source=ggml.go:482 msg="offloading 19 repeating layers to GPU" time=2026-03-11T17:42:07.732-04:00 level=INFO source=ggml.go:486 msg="offloading output layer to CPU" time=2026-03-11T17:42:07.732-04:00 level=INFO source=ggml.go:494 msg="offloaded 19/33 layers to GPU" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="2.3 GiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:245 msg="model weights" device=CPU size="3.9 GiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="812.8 MiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:256 msg="kv cache" device=CPU size="571.4 MiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="756.6 MiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:267 msg="compute graph" device=CPU size="630.8 MiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=device.go:272 msg="total memory" size="8.8 GiB" time=2026-03-11T17:42:07.732-04:00 level=INFO source=sched.go:566 msg="loaded runners" count=1 time=2026-03-11T17:42:07.732-04:00 level=INFO source=server.go:1350 msg="waiting for llama runner to start responding" time=2026-03-11T17:42:07.733-04:00 level=INFO source=server.go:1384 msg="waiting for server to become available" status="llm server loading model" time=2026-03-11T17:42:08.985-04:00 level=INFO source=server.go:1388 msg="llama runner started in 3.00 seconds" time=2026-03-11T17:42:09.068-04:00 level=WARN source=runner.go:187 msg="truncating input prompt" limit=4096 prompt=6979 keep=4 new=4096 CUDA error: invalid argument current device: 0, in function ggml_cuda_cpy at C:\a\ollama\ollama\ml\backend\ggml\ggml\src\ggml-cuda\cpy.cu:438 cudaMemcpyAsyncReserve(src1_ddc, src0_ddc, ggml_nbytes(src0), cudaMemcpyDeviceToDevice, main_stream) C:\a\ollama\ollama\ml\backend\ggml\ggml\src\ggml-cuda\ggml-cuda.cu:94: CUDA error time=2026-03-11T17:42:09.291-04:00 level=ERROR source=server.go:1610 msg="post predict" error="Post "http://127.0.0.1:16823/completion\": read tcp 127.0.0.1:16828->127.0.0.1:16823: wsarecv: An existing connection was forcibly closed by the remote host." [GIN] 2026/03/11 - 17:42:09 | 500 | 3.6940788s | 127.0.0.1 | POST "/api/chat" time=2026-03-11T17:42:09.504-04:00 level=ERROR source=server.go:304 msg="llama runner terminated" error="exit status 1"

Fix Plan

To resolve the issue, update Ollama to version 0.17.7. This version fix has been confirmed to resolve the CUDA error and 500 response when calling Ollama remotely.

Steps to Update

1. Stop the Ollama service: Ensure that any running Ollama instances are stopped before proceeding.
2. Download the latest Ollama version: Navigate to the official Ollama download page and download version 0.17.7.
3. Install the new version: Follow the installation instructions for the new version, making sure to select the correct options for your setup, including the port number (in this case, 11435).
4. Restart the Ollama service: Once the update is complete, restart the Ollama service.

Verification

After updating to version 0.17.7, verify that the issue is resolved by:

• Calling Ollama remotely from another machine on the local network.
• Checking for any CUDA errors or 500 responses.

Extra Tips

• Ensure that your NVIDIA drivers are up to date, as outdated drivers can cause compatibility issues with CUDA applications.
• If you continue to experience issues after updating Ollama, consider checking the system logs for any errors related to CUDA or the Ollama service.