```python
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.utils.cpp_extension import load_inline

_cpp_source = r"""
#include <torch/extension.h>
torch::Tensor bias_hardswish_relu_cuda(torch::Tensor x, torch::Tensor bias);
"""

_cuda_source = r"""
#include <torch/extension.h>
#include <ATen/cuda/CUDAContext.h>
#include <c10/cuda/CUDAGuard.h>
#include <c10/cuda/CUDAException.h>
#include <cuda_runtime.h>
#include <algorithm>

__device__ __forceinline__ float activate(float x, float bias) {
    x = x + bias;
    float y = (x * fminf(fmaxf(x + 3.0f, 0.0f), 6.0f)) / 6.0f;
    return y < 0.0f ? 0.0f : y;
}

template<bool AlignedChannels>
__global__ void bias_hardswish_relu_kernel(
    const float* __restrict__ input,
    const float* __restrict__ bias,
    float* __restrict__ output,
    int64_t count,
    int64_t spatial,
    int channels
) {
    const int64_t index =
        static_cast<int64_t>(blockIdx.x) * blockDim.x + threadIdx.x;
    const int64_t stride =
        static_cast<int64_t>(blockDim.x) * gridDim.x;
    const int64_t vector_count = count / 4;

    const float4* input4 = reinterpret_cast<const float4*>(input);
    float4* output4 = reinterpret_cast<float4*>(output);

    for (int64_t i = index; i < vector_count; i += stride) {
        float4 v = input4[i];
        if (AlignedChannels) {
            float b = bias[(i / (spatial / 4)) % channels];
            v.x = activate(v.x, b);
            v.y = activate(v.y, b);
            v.z = activate(v.z, b);
            v.w = activate(v.w, b);
        } else {
            const int64_t j = i * 4;
            v.x = activate(v.x, bias[(j / spatial) % channels]);
            v.y = activate(v.y, bias[((j + 1) / spatial) % channels]);
            v.z = activate(v.z, bias[((j + 2) / spatial) % channels]);
            v.w = activate(v.w, bias[((j + 3) / spatial) % channels]);
        }
        output4[i] = v;
    }

    for (int64_t i = vector_count * 4 + index; i < count; i += stride) {
        output[i] = activate(input[i], bias[(i / spatial) % channels]);
    }
}

torch::Tensor bias_hardswish_relu_cuda(
    torch::Tensor x, torch::Tensor bias
) {
    TORCH_CHECK(x.is_cuda() && bias.is_cuda(), "Expected CUDA tensors");
    TORCH_CHECK(x.device() == bias.device(), "Device mismatch");
    TORCH_CHECK(
        x.scalar_type() == at::kFloat && bias.scalar_type() == at::kFloat,
        "Expected float32 tensors"
    );
    TORCH_CHECK(x.dim() == 4 && x.is_contiguous(), "Expected contiguous NCHW");
    TORCH_CHECK(
        bias.dim() == 1 && bias.is_contiguous() && bias.numel() == x.size(1),
        "Invalid bias"
    );

    const c10::cuda::CUDAGuard guard(x.device());
    auto output = torch::empty_like(x);
    const int64_t count = x.numel();

    if (count > 0) {
        constexpr int threads = 256;
        const int blocks = static_cast<int>(
            std::min<int64_t>((count + 4 * threads - 1) / (4 * threads), 4096)
        );
        const int64_t spatial = x.size(2) * x.size(3);
        const int channels = static_cast<int>(x.size(1));
        auto stream = at::cuda::getCurrentCUDAStream();

        if (spatial % 4 == 0) {
            bias_hardswish_relu_kernel<true><<<blocks, threads, 0, stream>>>(
                x.data_ptr<float>(), bias.data_ptr<float>(),
                output.data_ptr<float>(), count, spatial, channels
            );
        } else {
            bias_hardswish_relu_kernel<false><<<blocks, threads, 0, stream>>>(
                x.data_ptr<float>(), bias.data_ptr<float>(),
                output.data_ptr<float>(), count, spatial, channels
            );
        }
        C10_CUDA_KERNEL_LAUNCH_CHECK();
    }

    return output;
}
"""

_fused_ops = load_inline(
    name="conv_bias_hardswish_relu_fused_v2",
    cpp_sources=_cpp_source,
    cuda_sources=_cuda_source,
    functions=["bias_hardswish_relu_cuda"],
    extra_cflags=["-O3"],
    extra_cuda_cflags=["-O3"],
    verbose=False,
)


class _BiasHardSwishReLU(torch.autograd.Function):
    @staticmethod
    def forward(ctx, x, bias):
        if ctx.needs_input_grad[0] or ctx.needs_input_grad[1]:
            ctx.save_for_backward(x, bias)
        return _fused_ops.bias_hardswish_relu_cuda(x, bias)

    @staticmethod
    def backward(ctx, grad_output):
        x, bias = ctx.saved_tensors
        z = x + bias.view(1, -1, 1, 1)
        slope = torch.where(
            z <= 0,
            torch.zeros_like(z),
            torch.where(z <= 3, z / 3 + 0.5, torch.ones_like(z)),
        )
        grad = grad_output * slope
        grad_x = grad if ctx.needs_input_grad[0] else None
        grad_bias = grad.sum((0, 2, 3)) if ctx.needs_input_grad[1] else None
        return grad_x, grad_bias


class ModelNew(nn.Module):
    def __init__(self, in_channels, out_channels, kernel_size):
        super().__init__()
        self.conv = nn.Conv2d(in_channels, out_channels, kernel_size)

    def forward(self, x):
        use_custom = (
            x.is_cuda
            and x.dtype == torch.float32
            and self.conv.weight.dtype == torch.float32
            and self.conv.bias.dtype == torch.float32
            and not torch.is_autocast_enabled()
        )
        if not use_custom:
            return F.relu(F.hardswish(self.conv(x)))

        y = F.conv2d(
            x,
            self.conv.weight,
            None,
            self.conv.stride,
            self.conv.padding,
            self.conv.dilation,
            self.conv.groups,
        )
        if y.is_contiguous():
            return _BiasHardSwishReLU.apply(y, self.conv.bias)

        y = y + self.conv.bias.view(1, -1, 1, 1)
        return F.relu(F.hardswish(y))
```