Getting Started on Intel GPU

Hardware Prerequisite

Supported OS	Validated Hardware
Linux	Intel® Client GPUs / Intel® Data Center GPU Max Series
Windows	Intel® Client GPUs
WSL2 (experimental feature)	Intel® Client GPUs

Intel GPUs support (Prototype) is ready in PyTorch* 2.6 for Intel® Client GPUs and Intel® Data Center GPU Max Series on both Linux and Windows, which brings Intel GPUs and the SYCL* software stack into the official PyTorch stack with consistent user experience to embrace more AI application scenarios.

Software Prerequisite

To use PyTorch on Intel GPUs, you need to install the Intel GPUs driver first. For installation guide, visit Intel GPUs Driver Installation.

Please skip the Intel® Deep Learning Essentials installation section if you install from binaries. For building from source, please refer to PyTorch Installation Prerequisites for Intel GPUs for both Intel GPU Driver and Intel® Deep Learning Essentials Installation.

Installation

Binaries

Now that we have Intel GPU Driver installed, use the following commands to install pytorch, torchvision, torchaudio on Linux.

For release wheels

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/xpu

For nightly wheels

pip3 install --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/xpu

From Source

Now that we have Intel GPU Driver and Intel® Deep Learning Essentials installed. Follow guides to build pytorch, torchvision, torchaudio from source.

Build from source for torch refer to PyTorch Installation Build from source.

Build from source for torchvision refer to Torchvision Installation Build from source.

Build from source for torchaudio refert to Torchaudio Installation Build from source.

Check availability for Intel GPU

To check if your Intel GPU is available, you would typically use the following code:

import torch
torch.xpu.is_available()  # torch.xpu is the API for Intel GPU support

If the output is False, double check driver installation for Intel GPUs.

Minimum Code Change

If you are migrating code from cuda, you would change references from cuda to xpu. For example:

# CUDA CODE
tensor = torch.tensor([1.0, 2.0]).to("cuda")

# CODE for Intel GPU
tensor = torch.tensor([1.0, 2.0]).to("xpu")

The following points outline the support and limitations for PyTorch with Intel GPU:

1. Both training and inference workflows are supported.

2. Both eager mode and torch.compile is supported.

3. Data types such as FP32, BF16, FP16, and Automatic Mixed Precision (AMP) are all supported.

Examples

This section contains usage examples for both inference and training workflows.

Inference Examples

Here is a few inference workflow examples.

Inference with FP32

import torch
import torchvision.models as models

model = models.resnet50(weights="ResNet50_Weights.DEFAULT")
model.eval()
data = torch.rand(1, 3, 224, 224)

model = model.to("xpu")
data = data.to("xpu")

with torch.no_grad():
    model(data)

print("Execution finished")

Inference with AMP

import torch
import torchvision.models as models

model = models.resnet50(weights="ResNet50_Weights.DEFAULT")
model.eval()
data = torch.rand(1, 3, 224, 224)

model = model.to("xpu")
data = data.to("xpu")

with torch.no_grad():
    d = torch.rand(1, 3, 224, 224)
    d = d.to("xpu")
    # set dtype=torch.bfloat16 for BF16
    with torch.autocast(device_type="xpu", dtype=torch.float16, enabled=True):
        model(data)

print("Execution finished")

Inference with torch.compile

import torch
import torchvision.models as models
import time

model = models.resnet50(weights="ResNet50_Weights.DEFAULT")
model.eval()
data = torch.rand(1, 3, 224, 224)
ITERS = 10

model = model.to("xpu")
data = data.to("xpu")

for i in range(ITERS):
    start = time.time()
    with torch.no_grad():
        model(data)
        torch.xpu.synchronize()
    end = time.time()
    print(f"Inference time before torch.compile for iteration {i}: {(end-start)*1000} ms")

model = torch.compile(model)
for i in range(ITERS):
    start = time.time()
    with torch.no_grad():
        model(data)
        torch.xpu.synchronize()
    end = time.time()
    print(f"Inference time after torch.compile for iteration {i}: {(end-start)*1000} ms")

print("Execution finished")

Training Examples

Here is a few training workflow examples.

Train with FP32

import torch
import torchvision

LR = 0.001
DOWNLOAD = True
DATA = "datasets/cifar10/"

transform = torchvision.transforms.Compose(
    [
        torchvision.transforms.Resize((224, 224)),
        torchvision.transforms.ToTensor(),
        torchvision.transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
    ]
)
train_dataset = torchvision.datasets.CIFAR10(
    root=DATA,
    train=True,
    transform=transform,
    download=DOWNLOAD,
)
train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=128)
train_len = len(train_loader)

model = torchvision.models.resnet50()
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=LR, momentum=0.9)
model.train()
model = model.to("xpu")
criterion = criterion.to("xpu")

print(f"Initiating training")
for batch_idx, (data, target) in enumerate(train_loader):
    data = data.to("xpu")
    target = target.to("xpu")
    optimizer.zero_grad()
    output = model(data)
    loss = criterion(output, target)
    loss.backward()
    optimizer.step()
    if (batch_idx + 1) % 10 == 0:
         iteration_loss = loss.item()
         print(f"Iteration [{batch_idx+1}/{train_len}], Loss: {iteration_loss:.4f}")
torch.save(
    {
        "model_state_dict": model.state_dict(),
        "optimizer_state_dict": optimizer.state_dict(),
    },
    "checkpoint.pth",
)

print("Execution finished")

Train with AMP

Note: Training with GradScaler requires hardware support for FP64. FP64 is not natively supported by the Intel® Arc™ A-Series Graphics. If you run your workloads on Intel® Arc™ A-Series Graphics, please disable GradScaler.

import torch
import torchvision

LR = 0.001
DOWNLOAD = True
DATA = "datasets/cifar10/"

use_amp=True

transform = torchvision.transforms.Compose(
    [
        torchvision.transforms.Resize((224, 224)),
        torchvision.transforms.ToTensor(),
        torchvision.transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
    ]
)
train_dataset = torchvision.datasets.CIFAR10(
    root=DATA,
    train=True,
    transform=transform,
    download=DOWNLOAD,
)
train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=128)
train_len = len(train_loader)

model = torchvision.models.resnet50()
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=LR, momentum=0.9)
scaler = torch.amp.GradScaler(device="xpu", enabled=use_amp)

model.train()
model = model.to("xpu")
criterion = criterion.to("xpu")

print(f"Initiating training")
for batch_idx, (data, target) in enumerate(train_loader):
    data = data.to("xpu")
    target = target.to("xpu")
    # set dtype=torch.bfloat16 for BF16
    with torch.autocast(device_type="xpu", dtype=torch.float16, enabled=use_amp):
        output = model(data)
        loss = criterion(output, target)
    scaler.scale(loss).backward()
    scaler.step(optimizer)
    scaler.update()
    optimizer.zero_grad()
    if (batch_idx + 1) % 10 == 0:
         iteration_loss = loss.item()
         print(f"Iteration [{batch_idx+1}/{train_len}], Loss: {iteration_loss:.4f}")

torch.save(
    {
        "model_state_dict": model.state_dict(),
        "optimizer_state_dict": optimizer.state_dict(),
    },
    "checkpoint.pth",
)

print("Execution finished")

Train with torch.compile

import torch
import torchvision

LR = 0.001
DOWNLOAD = True
DATA = "datasets/cifar10/"

transform = torchvision.transforms.Compose(
    [
        torchvision.transforms.Resize((224, 224)),
        torchvision.transforms.ToTensor(),
        torchvision.transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
    ]
)
train_dataset = torchvision.datasets.CIFAR10(
    root=DATA,
    train=True,
    transform=transform,
    download=DOWNLOAD,
)
train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=128)
train_len = len(train_loader)

model = torchvision.models.resnet50()
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=LR, momentum=0.9)
model.train()
model = model.to("xpu")
criterion = criterion.to("xpu")
model = torch.compile(model)

print(f"Initiating training with torch compile")
for batch_idx, (data, target) in enumerate(train_loader):
    data = data.to("xpu")
    target = target.to("xpu")
    optimizer.zero_grad()
    output = model(data)
    loss = criterion(output, target)
    loss.backward()
    optimizer.step()
    if (batch_idx + 1) % 10 == 0:
         iteration_loss = loss.item()
         print(f"Iteration [{batch_idx+1}/{train_len}], Loss: {iteration_loss:.4f}")
torch.save(
    {
        "model_state_dict": model.state_dict(),
        "optimizer_state_dict": optimizer.state_dict(),
    },
    "checkpoint.pth",
)

print("Execution finished")