ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-21 23:19:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

RPC: support multiple devices including cpu (#1024)

Browse Source 
* RPC support multiple devices

* rpc : update documentation (#16441)

Update the README file to match the newly added functionality of
exposing multiple devices from a single server.

Co-authored-by: Diego Devesa <slarengh@gmail.com>

# Conflicts:
#	examples/rpc/README.md

* Remove memory settings

* rpc : cache and reuse compute graphs (#15405)

Store the last computed graph and reuse it when possible.
Also do not return response from GRAPH_COMPUTE and assume it always
completes successfully. If this this is not the case, the server closes
the connection. This saves us a network round trip to the server.

* Add -cpu to include cpu backend

---------

Co-authored-by: firecoperana <firecoperana>
Co-authored-by: Radoslav Gerganov <rgerganov@gmail.com>
This commit is contained in:
firecoperana
2025-11-30 11:48:02 -06:00
committed by GitHub
parent 52adcf1e90
commit e89064e657
8 changed files with 734 additions and 381 deletions
Download Patch File Download Diff File Expand all files Collapse all files

68
examples/rpc/README.md
View File

@@ -4,7 +4,7 @@
> This example and the RPC backend are currently in a proof-of-concept development stage. As such, the functionality is fragile and
> insecure. **Never run the RPC server on an open network or in a sensitive environment!**
The `rpc-server` allows running `ggml` backend on a remote host.
The `rpc-server` allows exposing `ggml` devices on a remote host.
The RPC backend communicates with one or several instances of `rpc-server` and offloads computations to them.
This can be used for distributed LLM inference with `llama.cpp` in the following way:
@@ -14,27 +14,34 @@ flowchart TD
rpcb---|TCP|srvb
rpcb-.-|TCP|srvn
subgraph hostn[Host N]
srvn[rpc-server]-.-backend3["Backend (CUDA,Metal,etc.)"]
srvn[rpc-server]<-.->dev4["CUDA0"]
srvn[rpc-server]<-.->dev5["CPU"]
end
subgraph hostb[Host B]
srvb[rpc-server]---backend2["Backend (CUDA,Metal,etc.)"]
srvb[rpc-server]<-->dev3["Metal"]
end
subgraph hosta[Host A]
srva[rpc-server]---backend["Backend (CUDA,Metal,etc.)"]
srva[rpc-server]<-->dev["CUDA0"]
srva[rpc-server]<-->dev2["CUDA1"]
end
subgraph host[Main Host]
ggml[llama.cpp]---rpcb[RPC backend]
local["Local devices"]<-->ggml[llama-cli]
ggml[llama-cli]<-->rpcb[RPC backend]
end
style hostn stroke:#66,stroke-width:2px,stroke-dasharray: 5 5
classDef devcls fill:#5B9BD5
class local,dev,dev2,dev3,dev4,dev5 devcls
```
Each host can run a different backend, e.g. one with CUDA and another with Metal.
You can also run multiple `rpc-server` instances on the same host, each with a different backend.
By default, `rpc-server` exposes all available accelerator devices on the host.
If there are no accelerators, it exposes a single `CPU` device.
## Usage
On each host, build the corresponding backend with `cmake` and add `-DGGML_RPC=ON` to the build options.
For example, to build the CUDA backend with RPC support:
### Remote hosts
On each remote host, build the backends for each accelerator by adding `-DGGML_RPC=ON` to the build options.
For example, to build the `rpc-server` with support for CUDA accelerators:
```bash
mkdir build-rpc-cuda
@@ -43,36 +50,49 @@ cmake .. -DGGML_CUDA=ON -DGGML_RPC=ON
cmake --build . --config Release
```
Then, start the `rpc-server` with the backend:
When started, the `rpc-server` will detect and expose all available `CUDA` devices:
```bash
$ bin/rpc-server -p 50052
create_backend: using CUDA backend
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: CUDA_USE_TENSOR_CORES: yes
$ bin/rpc-server
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 T1200 Laptop GPU, compute capability 7.5, VMM: yes
Starting RPC server on 0.0.0.0:50052
Device 0: NVIDIA GeForce RTX 5090, compute capability 12.0, VMM: yes
Starting RPC server v3.0.0
endpoint : 127.0.0.1:50052
local cache : n/a
Devices:
CUDA0: NVIDIA GeForce RTX 5090 (32109 MiB, 31588 MiB free)
```
When using the CUDA backend, you can specify the device with the `CUDA_VISIBLE_DEVICES` environment variable, e.g.:
You can control the set of exposed CUDA devices with the `CUDA_VISIBLE_DEVICES` environment variable or the `--device` command line option. The following two commands have the same effect:
```bash
$ CUDA_VISIBLE_DEVICES=0 bin/rpc-server -p 50052
$ bin/rpc-server --device CUDA0 -p 50052
```
This way you can run multiple `rpc-server` instances on the same host, each with a different CUDA device.
### Main host
On the main host build `llama.cpp` only with `-DGGML_RPC=ON`:
On the main host build `llama.cpp` with the backends for the local devices and add `-DGGML_RPC=ON` to the build options.
Finally, when running `llama-cli` or `llama-server`, use the `--rpc` option to specify the host and port of each `rpc-server`:
```bash
mkdir build-rpc
cd build-rpc
cmake .. -DGGML_RPC=ON
cmake --build . --config Release
$ llama-cli -hf ggml-org/gemma-3-1b-it-GGUF -ngl 99 --rpc 192.168.88.10:50052,192.168.88.11:50052
```
Finally, use the `--rpc` option to specify the host and port of each `rpc-server`:
By default, llama.cpp distributes model weights and the KV cache across all available devices -- both local and remote -- in proportion to each device's available memory.
You can override this behavior with the `--tensor-split` option and set custom proportions when splitting tensor data across devices.
```bash
$ bin/llama-cli -m ../models/tinyllama-1b/ggml-model-f16.gguf -p "Hello, my name is" --repeat-penalty 1.0 -n 64 --rpc 192.168.88.10:50052,192.168.88.11:50052 -ngl 99
```
By default, the cache is stored in the `$HOME/.cache/llama.cpp/rpc` directory and can be controlled via the `LLAMA_CACHE` environment variable.
### Troubleshooting
Use the `GGML_RPC_DEBUG` environment variable to enable debug messages from `rpc-server`:
```bash
$ GGML_RPC_DEBUG=1 bin/rpc-server
```

208
examples/rpc/rpc-server.cpp
View File

@@ -32,6 +32,7 @@
#include <fstream>
#include <filesystem>
#include <codecvt>
#include <regex>
namespace fs = std::filesystem;
@@ -145,22 +146,24 @@ static std::string fs_get_cache_directory() {
}
struct rpc_server_params {
std::string host = "127.0.0.1";
int port = 50052;
size_t backend_mem = 0;
bool use_cache = false;
int n_threads = std::max(1U, std::thread::hardware_concurrency() / 2);
std::string host = "127.0.0.1";
int port = 50052;
bool use_cache = false;
bool use_cpu = false;
int n_threads = std::max(1U, std::thread::hardware_concurrency() / 2);
std::vector<std::string> devices;
};
static void print_usage(int /*argc*/, char** argv, rpc_server_params params) {
fprintf(stderr, "Usage: %s [options]\n\n", argv[0]);
fprintf(stderr, "options:\n");
fprintf(stderr, " -h, --help show this help message and exit\n");
fprintf(stderr, " -t, --threads number of threads for the CPU backend (default: %d)\n", params.n_threads);
fprintf(stderr, " -H HOST, --host HOST host to bind to (default: %s)\n", params.host.c_str());
fprintf(stderr, " -p PORT, --port PORT port to bind to (default: %d)\n", params.port);
fprintf(stderr, " -m MEM, --mem MEM backend memory size (in MB)\n");
fprintf(stderr, " -c, --cache enable local file cache\n");
fprintf(stderr, " -h, --help show this help message and exit\n");
fprintf(stderr, " -t, --threads N number of threads for the CPU device (default: %d)\n", params.n_threads);
fprintf(stderr, " -d, -dev, --device <dev1,dev2,...> comma-separated list of devices\n");
fprintf(stderr, " -cpu enable cpu backend\n");
fprintf(stderr, " -h, -H, --host, --Host HOST host to bind to (default: %s)\n", params.host.c_str());
fprintf(stderr, " -p, -P, --port, --Port PORT port to bind to (default: %d)\n", params.port);
fprintf(stderr, " -c, --cache enable local file cache\n");
fprintf(stderr, "\n");
}
@@ -168,7 +171,7 @@ static bool rpc_server_params_parse(int argc, char** argv, rpc_server_params& pa
std::string arg;
for (int i = 1; i < argc; i++) {
arg = argv[i];
if (arg == "-H" || arg == "--host") {
if (arg == "-H" || arg == "-h" || arg == "--host" || arg == "--Host") {
if (++i >= argc) {
return false;
}
@@ -184,7 +187,25 @@ static bool rpc_server_params_parse(int argc, char** argv, rpc_server_params& pa
return false;
}
}
else if (arg == "-p" || arg == "--port") {
else if (arg == "-d" || arg == "-dev" || arg == "--device") {
if (++i >= argc) {
return false;
}
const std::regex regex{ R"([,/]+)" };
std::string dev_str = argv[i];
std::sregex_token_iterator iter(dev_str.begin(), dev_str.end(), regex, -1);
std::sregex_token_iterator end;
for (; iter != end; ++iter) {
try {
params.devices.push_back(*iter);
}
catch (const std::exception&) {
fprintf(stderr, "error: invalid device: %s\n", iter->str().c_str());
return false;
}
}
}
else if (arg == "-p" || arg == "-P" || arg == "--port" || arg == "--Port") {
if (++i >= argc) {
return false;
}
@@ -196,11 +217,8 @@ static bool rpc_server_params_parse(int argc, char** argv, rpc_server_params& pa
else if (arg == "-c" || arg == "--cache") {
params.use_cache = true;
}
else if (arg == "-m" || arg == "--mem") {
if (++i >= argc) {
return false;
}
params.backend_mem = std::stoul(argv[i]) * 1024 * 1024;
else if (arg == "-cpu") {
params.use_cpu = true;
}
else if (arg == "-h" || arg == "--help") {
print_usage(argc, argv, params);
@@ -215,11 +233,18 @@ static bool rpc_server_params_parse(int argc, char** argv, rpc_server_params& pa
return true;
}
static ggml_backend_t create_backend(const rpc_server_params& params) {
static ggml_backend_t create_cpu_backend(const rpc_server_params& params) {
fprintf(stderr, "%s: using CPU backend\n", __func__);
ggml_backend_t backend = ggml_backend_cpu_init();
ggml_backend_cpu_set_n_threads(backend, params.n_threads);
return backend;
}
static ggml_backend_t create_gpu_backend(const rpc_server_params& params, uint32_t device) {
ggml_backend_t backend = NULL;
#ifdef GGML_USE_CUDA
fprintf(stderr, "%s: using CUDA backend\n", __func__);
backend = ggml_backend_cuda_init(0, nullptr); // init device 0
fprintf(stderr, "%s: using CUDA backend: CUDA%d\n", __func__, device);
backend = ggml_backend_cuda_init(device, nullptr); // init device
if (!backend) {
fprintf(stderr, "%s: ggml_backend_cuda_init() failed\n", __func__);
}
@@ -231,34 +256,113 @@ static ggml_backend_t create_backend(const rpc_server_params& params) {
}
#elif GGML_USE_VULKAN
fprintf(stderr, "%s: using Vulkan backend\n", __func__);
backend = ggml_backend_vk_init(0); // init device 0
backend = ggml_backend_vk_init(device); // init device 0
if (!backend) {
fprintf(stderr, "%s: ggml_backend_vulkan_init() failed\n", __func__);
}
#elif GGML_USE_SYCL
fprintf(stderr, "%s: using SYCL backend\n", __func__);
backend = ggml_backend_sycl_init(0); // init device 0
backend = ggml_backend_sycl_init(device); // init device 0
if (!backend) {
fprintf(stderr, "%s: ggml_backend_sycl_init() failed\n", __func__);
}
#endif
// if there aren't GPU Backends fallback to CPU backend
if (!backend) {
fprintf(stderr, "%s: using CPU backend\n", __func__);
backend = ggml_backend_cpu_init();
ggml_backend_cpu_set_n_threads(backend, params.n_threads);
}
//if (!backend) {
// fprintf(stderr, "%s: using CPU backend\n", __func__);
// backend = ggml_backend_cpu_init();
// ggml_backend_cpu_set_n_threads(backend, params.n_threads);
//}
return backend;
}
static void get_backend_memory(size_t * free_mem, size_t * total_mem) {
static int32_t find_device_idx(const std::string& str) {
std::regex pattern(R"((\d+)$)"); // Match digits at the end
std::smatch matches;
int number = -1;
if (std::regex_search(str, matches, pattern)) {
number = std::stoi(matches[1]);
}
return number;
}
static size_t get_gpu_backend_count(const rpc_server_params& params) {
size_t count = 0;
#if defined(GGML_USE_CUDA)
count = ggml_backend_cuda_get_device_count();
#elif defined(GGML_USE_SYCL)
count = ggml_backend_sycl_get_device_count();
#elif defined(GGML_USE_VULKAN)
count = ggml_backend_vk_get_device_count();
#elif defined(GGML_USE_CANN)
return ggml_backend_cann_get_device_count();
#endif
return count;
}
static std::vector<ggml_backend_t> get_devices(const rpc_server_params& params) {
std::vector<ggml_backend_t> devices;
if (!params.devices.empty()) {
for (auto device : params.devices) {
int32_t device_id;
ggml_backend_t dev;
if (params.use_cpu && device == "CPU" ) {
dev = create_cpu_backend(params);
} else {
device_id = find_device_idx(device);
if (device_id < 0) {
fprintf(stderr, "error: unknown device: %s\n", device.c_str());
continue;
}
dev = create_gpu_backend(params, device_id);
}
if (dev) {
devices.push_back(dev);
} else {
fprintf(stderr, "error: unknown device: %s\n", device.c_str());
}
}
}
else {
for (size_t i = 0; i < get_gpu_backend_count(params); i++) {
ggml_backend_t dev = create_gpu_backend(params, i);
if (dev) {
devices.push_back(dev);
}
}
// cpu backend at last
if (params.use_cpu || devices.empty()) {
ggml_backend_t dev = create_cpu_backend(params);
if (dev) {
devices.push_back(dev);
}
}
}
return devices;
}
static void get_cpu_backend_memory(size_t * free_mem, size_t * total_mem) {
#ifdef _WIN32
MEMORYSTATUSEX status;
status.dwLength = sizeof(status);
GlobalMemoryStatusEx(&status);
*total_mem = status.ullTotalPhys;
*free_mem = status.ullAvailPhys;
#else
long pages = sysconf(_SC_PHYS_PAGES);
long page_size = sysconf(_SC_PAGE_SIZE);
*total_mem = pages * page_size;
*free_mem = *total_mem;
#endif
}
static void get_backend_memory(uint32_t device, size_t * free_mem, size_t * total_mem) {
#ifdef GGML_USE_CUDA
ggml_backend_cuda_get_device_memory(0, free_mem, total_mem);
ggml_backend_cuda_get_device_memory(device, free_mem, total_mem);
#elif GGML_USE_VULKAN
ggml_backend_vk_get_device_memory(0, free_mem, total_mem);
ggml_backend_vk_get_device_memory(device, free_mem, total_mem);
#elif GGML_USE_SYCL
ggml_backend_sycl_get_device_memory(0, free_mem, total_mem);
ggml_backend_sycl_get_device_memory(device, free_mem, total_mem);
#else
#ifdef _WIN32
MEMORYSTATUSEX status;
@@ -292,20 +396,27 @@ int main(int argc, char * argv[]) {
fprintf(stderr, "\n");
}
ggml_backend_t backend = create_backend(params);
if (!backend) {
fprintf(stderr, "Failed to create backend\n");
auto devices = get_devices(params);
if (devices.empty()) {
fprintf(stderr, "No backend found\n");
return 1;
}
std::string endpoint = params.host + ":" + std::to_string(params.port);
size_t free_mem, total_mem;
if (params.backend_mem > 0) {
free_mem = params.backend_mem;
total_mem = params.backend_mem;
}
else {
get_backend_memory(&free_mem, &total_mem);
std::vector<size_t> free_mem, total_mem;
for (size_t i = 0; i < devices.size(); i++) {
size_t free, total;
const char* name = ggml_backend_name(devices[i]);
if (std::string(name) == "CPU") {
get_cpu_backend_memory(&free, &total);
} else {
int32_t idx = find_device_idx(name);
get_backend_memory((uint32_t) idx, &free, &total);
}
free_mem.push_back(free);
total_mem.push_back(total);
}
const char * cache_dir = nullptr;
std::string cache_dir_str;
if (params.use_cache) {
@@ -316,14 +427,7 @@ int main(int argc, char * argv[]) {
}
cache_dir = cache_dir_str.c_str();
}
printf("Starting RPC server v%d.%d.%d\n",
RPC_PROTO_MAJOR_VERSION,
RPC_PROTO_MINOR_VERSION,
RPC_PROTO_PATCH_VERSION);
printf(" endpoint : %s\n", endpoint.c_str());
printf(" local cache : %s\n", cache_dir ? cache_dir : "n/a");
printf(" backend memory : %zu MB\n", free_mem / (1024 * 1024));
ggml_backend_rpc_start_server(backend, endpoint.c_str(), cache_dir, free_mem, total_mem);
ggml_backend_free(backend);
ggml_backend_rpc_start_server(endpoint.c_str(), cache_dir, devices.size(), devices.data(),
free_mem.data(), total_mem.data());
return 0;
}