Feature Request: fixed VRAM allocation pool
Ninguém assumiu esta issue ainda.
Avaliação
- Dificuldade
- 5/5
- Tempo estimado
- Mais de uma semana
- Facilidade para iniciantes
- 22/100
- Tipo de issue
- Funcionalidade
- Clareza
- Razoavelmente clara
- Status de atividade
- Pouca atividade
- Stack de tecnologia
- cpp
- Domínio
- machine-learning, performance
Direção de pesquisa
The request centres on the CUDA backend's buffer allocation, so start with ggml_backend_cuda_buffer_type_alloc_buffer in ggml/src/ggml-cuda.cu and trace where cudaMalloc is called for model weights and the KV cache. The issue proposes a fixed pool sized as a fraction of VRAM, but the design (how to size it, how to handle fragmentation and freeing) is unsettled and the code in the issue is an unverified sketch. Done would need a maintainer-agreed design before any implementation.
Escrita pelo modelo de indexação a partir do texto da issue.
Descrição
Prerequisites
- I am running the latest code. Mention the version if possible as well.
- I carefully followed the README.md.
- I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
- I reviewed the Discussions, and have a new and useful enhancement to share.
Feature Description
It would be nice to be able to set the value of preallocated VRAM in percents like for vLLM (ex. 0.9 = 90%) and to use preallocated pool in this mode, so that extra software never influence beellama work.
Motivation
The problem: whatever manually crafted settings for beellama and model are made, the VRAM consumption is not fully controlled, because VRAM is allocated during context growth and beellama work and because other GUI applications (browsers, electron-based projects) also continue to consume extra VRAM. Currently Firefox can easily consume up to 1.5 GB vram and begin with 100 MB initially, so generation speed just drops to almost zero from unpredictable moment.
Possible Implementation
LLM based suggestions:
Method 2: Code Modification (Hard Physical Reservation Pool)
If you want llama.cpp to allocate a single massive contiguous block of VRAM (e.g., 20 GB out of 24 GB) via cudaMalloc at program start, and have GGML consume slices of that block instead of calling cudaMalloc dynamically:
Step 1: Create a Fixed Global Pool
In ggml/src/ggml-cuda.cu (or ggml-cuda.cpp), declare a static memory pool pointer and allocate it during CUDA backend initialization:
// Static preallocated pool
static void* g_cuda_vram_pool = nullptr;
static size_t g_vram_pool_offset = 0;
static size_t g_vram_pool_size = 0;
void init_cuda_vram_pool(size_t reserve_bytes) {
if (g_cuda_vram_pool != nullptr) return;
// Physically allocate the entire pool at start
cudaError_t err = cudaMalloc(&g_cuda_vram_pool, reserve_bytes);
if (err == cudaSuccess) {
g_vram_pool_size = reserve_bytes;
printf("CUDA: Preallocated %zx bytes of VRAM pool successfully.\n", reserve_bytes);
} else {
printf("CUDA: Failed to preallocate VRAM pool: %s\n", cudaGetErrorString(err));
}
}
Step 2: Override GGML's CUDA Buffer Allocation Function
In ggml-cuda.cu, locate the buffer allocation function used by the CUDA backend: ggml_backend_cuda_buffer_type_alloc_buffer.
Replace the underlying cudaMalloc call inside that function so it returns an offset pointer into g_cuda_vram_pool:
// Simplistic bump-allocator replacement inside CUDA backend
static void* pool_alloc(size_t size) {
// Maintain 256-byte alignment required by CUDA
size_t aligned_size = (size + 255) & ~255;
if (g_vram_pool_offset + aligned_size > g_vram_pool_size) {
printf("CUDA Pool Error: Out of preallocated VRAM!\n");
return nullptr;
}
void* ptr = (uint8_t*)g_cuda_vram_pool + g_vram_pool_offset;
g_vram_pool_offset += aligned_size;
return ptr;
}
// In your buffer allocation routine:
// Replace cudaMalloc(&ptr, size) with:
ptr = pool_alloc(size);
Step 3: Call Initialization in llama_backend_init
Call init_cuda_vram_pool(target_size_bytes) inside your initialization function (e.g., llama_backend_init) before loading model layers or allocating context buffers.
Result
When llama.cpp starts up, it immediately issues a single cudaMalloc call for your requested size. All model weights, KV cache matrices, and scratch workspaces allocated during token generation will be carved directly out of this existing buffer. No further calls to cudaMalloc or cudaFree will touch the driver or OS.
- Linguagem predominante
- C++
- Estrelas
- 1.2k
- Forks
- 80
- Merge médio
- 4d 20h
- PRs com merge (30d)
- 2
Preparar o ambiente
Este projeto não oferece contêiner de desenvolvimento, Dockerfile nem guia de contribuição, então a configuração fica por sua conta: comece pelo README e veja nosso guia da primeira contribuição para os passos gerais.
Primeiros passos
- Leia a issue inteira e depois o guia de contribuição do projeto.
- Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
- Faça um fork do repositório e trabalhe em uma branch.
- Abra um pull request que referencie o número da issue.
Mais de Anbeeld/beellama.cpp
-
[BUG] KVarN assertion failure in build_inp_mem_hybrid_iswa() due to uninitialized self_kvarn_rotAberta
Dificuldade 2/5 1-3 horas Facilidade para iniciantes 68/100
Anbeeld/beellama.cpp#205 ·
-
Dificuldade 3/5 Meio dia Facilidade para iniciantes 58/100
Anbeeld/beellama.cpp#206 ·
-
Dificuldade 3/5 1-2 dias Facilidade para iniciantes 57/100
Anbeeld/beellama.cpp#204 ·
-
Dificuldade 4/5 3-5 dias Facilidade para iniciantes 45/100
Anbeeld/beellama.cpp#203 · 2 comentários ·
-
Dificuldade 4/5 3-5 dias Facilidade para iniciantes 45/100
Anbeeld/beellama.cpp#202 · 1 comentário ·
Todas as issues de Anbeeld/beellama.cpp
Issues semelhantes
-
Dificuldade 2/5 1-3 horas Facilidade para iniciantes 72/100
CachyOS/packageinstaller#144 ·
Mantenedores costumam responder em até 1 dia
-
documentation
Dificuldade 1/5 Menos de uma hora Facilidade para iniciantes 82/100
TomerAberbach/profiler-md#951 ·
Mantenedores costumam responder em até 1 dia
-
[Bug] Default key routing (JavaStringHash) picks a different partition than the Java client for about half of keysTalvez já em andamento @0z5a assumiu hoje. Aberta
Dificuldade 2/5 1-3 horas Facilidade para iniciantes 82/100
apache/pulsar-client-go#1575 ·
Mantenedores costumam responder em até 2 dias
-
Dificuldade 2/5 1-3 horas Facilidade para iniciantes 72/100
pocoproject/poco#5764 ·
Mantenedores costumam responder em até 2 dias
-
bug cpp-core-guidelines low-priority
Dificuldade 2/5 1-3 horas Facilidade para iniciantes 62/100
NatronGitHub/openfx-supportext#164 ·
Mantenedores costumam responder em até 1 dia