taeda 0.1.0__py3-none-any.whl

taeda/__init__.py

@@ -0,0 +1,3 @@
+from .tflops import *
+
+__all__ = ["get_tflops"]

taeda/tflops.py

@@ -0,0 +1,58 @@
+
+import torch
+import pynvml
+
+
+def get_tflops(device_num=0):
+    """
+    Estimate theoretical FP32 GPU performance in TFLOPS.
+    """
+
+    if not torch.cuda.is_available():
+        return None
+
+    if not 0 <= device_num < torch.cuda.device_count():
+        raise ValueError("Invalid GPU index")
+
+    pynvml.nvmlInit()
+
+    try:
+        gpu = torch.cuda.get_device_properties(device_num)
+        handle = pynvml.nvmlDeviceGetHandleByIndex(device_num)
+
+        # CUDA cores per SM depends on GPU architecture.
+        cores_per_sm = {
+            (5, 0): 128,
+            (5, 2): 128,
+            (5, 3): 128,
+            (6, 0): 64,
+            (6, 1): 128,
+            (6, 2): 128,
+            (7, 0): 64,
+            (7, 5): 64,
+            (8, 0): 64,
+            (8, 6): 128,
+            (8, 9): 128,
+            (9, 0): 128,
+        }
+
+        capability = (gpu.major, gpu.minor)
+        cores = cores_per_sm.get(capability)
+
+        if cores is None:
+            raise NotImplementedError(
+                f"Unknown CUDA architecture: {capability}"
+            )
+
+        cuda_cores = gpu.multi_processor_count * cores
+
+        clock_mhz = pynvml.nvmlDeviceGetClockInfo(
+            handle, pynvml.NVML_CLOCK_SM
+        )
+
+        clock_ghz = clock_mhz / 1000
+
+        return cuda_cores * clock_ghz * 2 / 1000
+
+    finally:
+        pynvml.nvmlShutdown()

taeda-0.1.0.dist-info/METADATA

@@ -0,0 +1,82 @@
+Metadata-Version: 2.4
+Name: taeda
+Version: 0.1.0
+Summary: A torch utils library
+License-Expression: MIT
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: torch
+Requires-Dist: nvidia-ml-py
+
+# Taeda
+
+**A lightweight Python library for GPU utilities and performance estimation.**
+
+Taeda is a Python package built with PyTorch and NVIDIA Management Library (NVML) to retrieve GPU information and estimate theoretical floating-point performance (TFLOPS).
+
+## Features
+
+- Detect available NVIDIA CUDA GPUs.
+- Retrieve GPU clock speeds in real time.
+- Calculate CUDA core counts.
+- Estimate theoretical FP32 TFLOPS.
+- Support multiple NVIDIA GPUs.
+- Simple Python API.
+
+## Installation
+
+```bash
+pip install taeda
+```
+
+## Requirements
+
+- Python 3.9+
+- PyTorch
+- NVIDIA GPU
+- NVIDIA drivers
+- `nvidia-ml-py`
+
+## Usage
+
+```python
+import taeda
+
+tflops = taeda.get_tflops(0)
+
+print(f"GPU Performance: {tflops:.3f} TFLOPS")
+```
+
+Example output:
+
+```text
+GPU Performance: 0.622 TFLOPS
+```
+
+## API
+
+### `get_tflops(device_num=0)`
+
+Estimates the theoretical FP32 performance of an NVIDIA GPU.
+
+| Parameter | Description |
+|---|---|
+| `device_num` | CUDA GPU index (default: `0`) |
+
+**Returns:** Estimated FP32 performance in TFLOPS (`float`), or `None` if CUDA is unavailable.
+
+**Raises:** `ValueError` if the GPU index is invalid.
+
+The clock speed is retrieved dynamically through NVML.
+
+## Limitations
+
+- TFLOPS values are theoretical estimates, not measured benchmark results.
+- Clock speeds fluctuate depending on GPU load, temperature and power limits.
+- CUDA core counts depend on GPU architecture.
+- Currently intended for supported NVIDIA CUDA GPUs.
+- Actual application performance depends on memory bandwidth, workload and other hardware limitations.
+
+## License
+
+MIT License.

taeda-0.1.0.dist-info/RECORD

@@ -0,0 +1,6 @@
+taeda/__init__.py,sha256=o9zKyMRoNJDe8U0tfTTI2EPguXZ1t0Lbc8gxNII-AMg,49
+taeda/tflops.py,sha256=3y1oV11hwLfaegnEj9w31z47SLkstV5qKF2Fy3fbIK4,1417
+taeda-0.1.0.dist-info/METADATA,sha256=JiZQ5yb8DdDIUNyobHtIKa8W0TANc7MBbIdx5YtagHU,1881
+taeda-0.1.0.dist-info/WHEEL,sha256=mOdXR9C3wMVov7cSY8PY-XCzo3SmmWCHvGQr5csJoao,91
+taeda-0.1.0.dist-info/top_level.txt,sha256=oSzjuexM7IPB2lL236IuexsuGoG1XdLvp6R1FHWjhUE,6
+taeda-0.1.0.dist-info/RECORD,,

taeda-0.1.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (84.2.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

taeda-0.1.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+taeda