NVIDIA GeForce GTX 870M

NVIDIA GeForce GTX 870M

NVIDIA GeForce GTX 870M in 2025: A Retrospective and Relevance

An overview of an outdated mobile graphics card in the era of the RTX 50 series


Introduction

In 2025, when ray tracing and neural network technologies dominate the scene, the NVIDIA GeForce GTX 870M appears as a relic from the past. Released in 2014, this mobile GPU was once a flagship for gaming laptops. But is there a place for it in the modern world? Let’s explore its features, performance, and areas of application today.


1. Architecture and Key Features

Kepler Architecture: The Legacy of the 2010s

The GTX 870M is built on the Kepler architecture (28 nm), which looks archaic in 2025. It comprises 1344 CUDA cores and boasts a clock speed of up to 967 MHz. Modern features like RTX (ray tracing), DLSS, or FidelityFX are absent — these technologies emerged later, starting with the RTX 20 series.

Manufacturing Process and Limitations

The 28-nm manufacturing process significantly lags behind today’s 4-nm NVIDIA chips. This leads to high power consumption and heat generation with modest performance. For comparison, the RTX 4050 Mobile (2023) utilizes a 5-nm process and consumes less power while delivering 4 to 5 times more FPS.


2. Memory: What Does the GTX 870M Offer?

GDDR5 and 192-bit Bus

The card is equipped with either 3 GB or 6 GB of GDDR5 memory with a bandwidth of 120 GB/s (192-bit bus). For games from 2014 to 2016, this was sufficient, but by 2025 even indie projects like Hades II require at least 4 GB of VRAM. Modern AAA titles (GTA VI, Starfield) with high-quality textures can consume 6 GB in seconds.

Buffer Issues

The limited memory size and low bandwidth lead to FPS drops in open-world games and complex scene rendering. For example, in Cyberpunk 2077 (on low settings), the GTX 870M delivers 15-20 FPS at 1080p due to insufficient VRAM.


3. Gaming Performance: Nostalgia or Torment?

1080p: Minimum Survival

In older titles (The Witcher 3, CS:GO), the card shows 40-60 FPS on medium settings. However, in Elden Ring (minimum settings), FPS rarely exceeds 25-30. Higher resolutions like 1440p and 4K are unattainable due to lack of power.

Ray Tracing: No Support

RTX features that have become standard by 2025 are absent here. Even with mods like Reshade, mimicking ray tracing effects is impossible — the Kepler architecture isn’t designed for such computations.


4. Professional Use Cases: Is There Potential?

CUDA: A Drop in the Ocean

With 1344 CUDA cores, it's meager for modern tasks. Rendering a medium scene in Blender takes 2-3 hours versus 10-15 minutes on the RTX 4060. Basic functions are sufficient for video editing in DaVinci Resolve, but rendering a 4K video with effects becomes a challenge.

OpenCL and Scientific Computing

The card supports OpenCL, but its performance in MATLAB or TensorFlow (via CUDA) is too low for serious projects. For training neural networks or simulations, cloud solutions are preferable.


5. Power Consumption and Heat Generation

TDP 100W: Inefficient by 2025 Standards

With a TDP of 100W, the GTX 870M runs hotter than modern counterparts. In laptops, this often leads to throttling — a reduction in speeds due to overheating.

Cooling Recommendations

If using the GTX 870M in a PC (via an external enclosure), a case with 3-4 fans and an airflow of ≥50 CFM is necessary. The ideal operating temperature is below 80°C.


6. Comparison with Competitors

AMD Radeon R9 M290X: Similar Issues

The main competitor from 2014, the R9 M290X, has also become outdated. It slightly handles Vulkan projects (Doom Eternal) better but falls short in DirectX 12.

Modern Alternatives: RTX 3050 Mobile

Even the budget RTX 3050 Mobile (2021) outperforms the GTX 870M by 3-4 times. It supports DLSS 3, consumes less power, and easily runs games at 1080p on high settings.


7. Practical Tips

Power Supply: At Least 450W

For a desktop PC with GTX 870M, a 450W PSU (80+ Bronze) is needed. Ensure the motherboard supports PCIe 3.0 x16.

Drivers: Seek Legacy Versions

Official driver support ended in 2021. Look for updated versions on forums like TechPowerUp or use modified drivers (e.g., from enthusiasts).

OS Compatibility

Windows 10 is the optimal choice. Windows 11 may cause conflicts due to the lack of a TPM module in older laptops.


8. Pros and Cons

Pros:

- Low price: New units are not sold, but used ones can be found for $30-50.

- Support for DirectX 12 (Feature Level 11_0) for basic tasks.

Cons:

- No support for modern APIs (DirectX 12 Ultimate, Vulkan 1.3).

- Limited performance in gaming and professional applications.


9. Final Conclusion: Who Is the GTX 870M For?

This graphics card is suitable for:

1. Retro gaming enthusiasts looking to run 2010s projects on original hardware.

2. Owners of older laptops needing a GPU upgrade (if the design allows).

3. Office tasks: handling documents, browsing, simple video editing.

In 2025, the GTX 870M holds more historical value than practical gaming capability. For comfortable gaming or graphics work, it's better to choose a budget RTX 3050 or AMD Radeon RX 6600. However, if you feel nostalgic for Skyrim or Dota 2, this card can still provide you with hours of modest enjoyment.

Basic

Label Name
NVIDIA
Platform
Mobile
Launch Date
March 2014
Model Name
GeForce GTX 870M
Generation
GeForce 800M
Base Clock
941MHz
Boost Clock
967MHz
Bus Interface
MXM-B (3.0)
Transistors
3,540 million
TMUs
?
Texture Mapping Units (TMUs) serve as components of the GPU, which are capable of rotating, scaling, and distorting binary images, and then placing them as textures onto any plane of a given 3D model. This process is called texture mapping.
112
Foundry
TSMC
Process Size
28 nm
Architecture
Kepler

Memory Specifications

Memory Size
3GB
Memory Type
GDDR5
Memory Bus
?
The memory bus width refers to the number of bits of data that the video memory can transfer within a single clock cycle. The larger the bus width, the greater the amount of data that can be transmitted instantaneously, making it one of the crucial parameters of video memory. The memory bandwidth is calculated as: Memory Bandwidth = Memory Frequency x Memory Bus Width / 8. Therefore, when the memory frequencies are similar, the memory bus width will determine the size of the memory bandwidth.
192bit
Memory Clock
1250MHz
Bandwidth
?
Memory bandwidth refers to the data transfer rate between the graphics chip and the video memory. It is measured in bytes per second, and the formula to calculate it is: memory bandwidth = working frequency × memory bus width / 8 bits.
120.0 GB/s

Theoretical Performance

Pixel Rate
?
Pixel fill rate refers to the number of pixels a graphics processing unit (GPU) can render per second, measured in MPixels/s (million pixels per second) or GPixels/s (billion pixels per second). It is the most commonly used metric to evaluate the pixel processing performance of a graphics card.
27.08 GPixel/s
Texture Rate
?
Texture fill rate refers to the number of texture map elements (texels) that a GPU can map to pixels in a single second.
108.3 GTexel/s
FP64 (double)
?
An important metric for measuring GPU performance is floating-point computing capability. Double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy, while single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
108.3 GFLOPS
FP32 (float)
?
An important metric for measuring GPU performance is floating-point computing capability. Single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks, while double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
2.547 TFLOPS

Miscellaneous

Shading Units
?
The most fundamental processing unit is the Streaming Processor (SP), where specific instructions and tasks are executed. GPUs perform parallel computing, which means multiple SPs work simultaneously to process tasks.
1344
L1 Cache
16 KB (per SMX)
L2 Cache
384KB
TDP
100W
Vulkan Version
?
Vulkan is a cross-platform graphics and compute API by Khronos Group, offering high performance and low CPU overhead. It lets developers control the GPU directly, reduces rendering overhead, and supports multi-threading and multi-core processors.
1.1
OpenCL Version
3.0
OpenGL
4.6
DirectX
12 (11_0)
CUDA
3.0
Power Connectors
None
Shader Model
5.1
ROPs
?
The Raster Operations Pipeline (ROPs) is primarily responsible for handling lighting and reflection calculations in games, as well as managing effects like anti-aliasing (AA), high resolution, smoke, and fire. The more demanding the anti-aliasing and lighting effects in a game, the higher the performance requirements for the ROPs; otherwise, it may result in a sharp drop in frame rate.
24

Benchmarks

FP32 (float)
Score
2.547 TFLOPS
Blender
Score
121.28
Hashcat
Score
45978 H/s

Compared to Other GPU

FP32 (float) / TFLOPS
2.693 +5.7%
2.585 +1.5%
2.509 -1.5%
2.429 -4.6%
Blender
1506.77 +1142.4%
848 +599.2%
194 +60%
Hashcat / H/s
52572 +14.3%
49571 +7.8%
45589 -0.8%
44442 -3.3%