Search results

Feature Request: Add VK_EXT_memory_priority support for model allocations (Vulkan backend)

enhancementNew feature or requestNew feature or request

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

Feature

#17605 In ggml-org/llama.cpp;

· hikki-gd opened on Nov 29, 2025

Eval bug: Qwen Models segfault on Vulkan

bugSomething isn't workingSomething isn't working

high severityUsed to report high severity bugs in llama.cpp (Malfunctioning hinder important workflow)Used to report high severity bugs in llama.cpp (Malfunctioning hinder important workflow)

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

Bug

#17561 In ggml-org/llama.cpp;

· leuchtetgruen opened on Nov 27, 2025

Eval bug: Vulkan issues starting with 439342ea0be347ff279ec204719794df3b3108f6

bugSomething isn't workingSomething isn't working

testingEverything test relatedEverything test related

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#17438 In ggml-org/llama.cpp;

· ddpasa opened on Nov 22, 2025

Eval bug: AMD driver 25.11.1 crashes with llamacpp built with Vulkan SDK 1.4.328.1

bugSomething isn't workingSomething isn't working

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#17432 In ggml-org/llama.cpp;

· Tetramatrix opened on Nov 21, 2025

Eval bug: Vulkan - Gemma3n-E2B-Q4_K_M model crashes llama-cli during evaluation [Intel IGPU]

bugSomething isn't workingSomething isn't working

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#17389 In ggml-org/llama.cpp;

· virajwad opened on Nov 19, 2025

Eval bug: Vulkan regression with -fa 1

bug-unconfirmed

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#17366 In ggml-org/llama.cpp;

· eiffel31 opened on Nov 18, 2025

Eval bug: Performance in current vulkan binaries is BAD

bug-unconfirmed

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#16317 In ggml-org/llama.cpp;

· openconstruct opened on Sep 28, 2025

Misc. bug: Critical Crash and Performance Degradation in Vulkan Build since Release B6524

bugSomething isn't workingSomething isn't working

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#16301 In ggml-org/llama.cpp;

· Basten7 opened on Sep 28, 2025

Misc. bug: Vulkan backend shows negative scaling at low batch sizes with MOE models

research 🔬

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#16134 In ggml-org/llama.cpp;

· Mushoz opened on Sep 20, 2025

Eval bug: Vulkan backend hangs forever with NVIDIA GPU on FreeBSD

bugSomething isn't workingSomething isn't working

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#15996 In ggml-org/llama.cpp;

· var1ati0n opened on Sep 14, 2025

Research: Performance differences between Metal (macOS) and Vulkan (Linux)

Apple Metalhttps://en.wikipedia.org/wiki/Metal_(API)https://en.wikipedia.org/wiki/Metal_(API)

research 🔬

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#10982 In ggml-org/llama.cpp;

· asahilina opened on Dec 26, 2024