Dark Mode
Hardware Specifications and FPS Prediction
Product Reviews & Feedback
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Data details the performance metrics, specifically Frame Per Second (FPS) measurements, of various video games running on different computer hardware configurations. The primary goal of the resource is to allow users to predict FPS performance based on the technical specifications of the components involved. Each observation records the FPS achieved by a game—defined by its name, displayed resolution, and quality settings—running on a system characterized by its CPU and GPU specifications, reflecting the factory state of these components.
Columns
The dataset focuses on technical specifications of the hardware, detailing 10 of the available 44 columns:
- CpuName: The identified name of the CPU. The most common name recorded is 'Intel Core i9-9900K'.
- CpuNumberOfCores: The count of physical cores present in the CPU. Values range from 2 up to 12, with a mean of 6.32.
- CpuNumberOfThreads: The total number of threads supported by the CPU. Values range from 4 to 24, averaging 10.8.
- CpuBaseClock: The base clock frequency measured in Mhz. This value is uniformly 100 for all records.
- CpuCacheL1: The total size of the level 1 cache in kilobytes (kB). The mean size is 515 kB, with a maximum recorded size of 1.15k kB.
- CpuCacheL2: The total size of the level 2 cache in kB. Values span up to 6.14k kB, with an average size of 2.51k kB.
- CpuCacheL3: The total size of the level 3 cache in kB. Values range from 3 to 64, with a mean of 18.2.
- CpuDieSize: The physical size of the die in square meters. Note that this attribute has a high percentage of missing values (53%).
- CpuFrequency: The frequency measured in Mhz. Values range from 2.8k Mhz up to 4.2k Mhz, averaging 3.59k Mhz.
- CpuMultiplier: The multiplier of the CPU. Values range from 28 to 42, with a mean of 35.9.
Distribution
The material is stored as a CSV file named
fps_benchmark.csv, which has a size of 8.45 MB. The dataset contains 24.6k valid records across most key CPU specification columns. The data structure was filtered initially to only include observations associated with the 'fps-benchmark' subset, and the corresponding column was subsequently removed. While many columns are 100% valid, the CpuDieSize attribute contains 53% missing entries.Usage
This resource is ideal for developing machine learning models aimed at predicting video game performance (FPS) based on underlying technical hardware specifications. It can be used for hardware benchmarking analysis, regression modelling, and understanding the performance bottlenecks caused by specific CPU or GPU configurations.
Coverage
The material captures FPS measurements relating the performance outcome to CPU, GPU, and Game characteristics (name, resolution, quality settings). The data covers 19 unique CPU names. The scope includes only those observations that were designated as "fps-benchmark" in the original source data.
License
Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
Who Can Use It
The dataset is intended for advanced data scientists, machine learning engineers, and analysts focused on hardware performance and video game metrics. It holds a maximum usability rating of 10.00 and is a static dataset, meaning no future updates are expected.
Dataset Name Suggestions
- Video Game FPS Performance Benchmark
- Hardware Specifications and FPS Prediction
- CPU/GPU Gaming Performance Metrics
Attributes
Original Data Source: Hardware Specifications and FPS Prediction
QUALITYLoading...