The NVIDIA GeForce GTX 470 is a highly efficient graphics card designed for competitive gaming and professional use. Its powerful Graphics Processor, the GF100-275-A3 variant, operates on the cutting-edge Fermi architecture and is manufactured by TSMC using a 40 nm process. With 3.1 billion transistors, a density of 5.9M/mm², and a die size of 529mm², this card is designed to provide top-notch performance for demanding applications.
The GeForce GTX 470 was released on March 26th, 2010 and quickly became one of the most popular graphics cards on the market. It is the predecessor of the GTX 580, which succeeded it in NVIDIA's product line. After a successful run, production of this card has now ended, making it a sought-after collector's item for enthusiasts. Its launch price of 349 USD made it a competitive option for gamers looking for a high-performance card without breaking the bank.
Equipped with a PCI Express 2.0 x16 bus interface, dual-slot design, and a length of 241mm, the GTX 470 offers seamless compatibility with most modern motherboards. It boasts impressive clock speeds, with the GPU clock operating at 608 MHz, the shader clock at 1215 MHz, and the memory clock at 837 MHz. This combination allows for smooth and responsive performance, even when running demanding tasks.
The GTX 470 comes with 1280 MB of high-speed GDDR5 memory and has a 320-bit memory bus, delivering a bandwidth of 133.9 GB/s. This ensures quick loading times, smooth gameplay, and seamless multitasking. Its render configuration includes 448 shading units, 56 TMUs, 40 ROPs, and 14 SMs, as well as 64 KB of L1 cache per SM and 640 KB of L2 cache. These specs contribute to the card's impressive theoretical performance, with a pixel rate of 17.02 GPixel/s, a texture rate of 34.05 GTexel/s, and 1,089 GFLOPS of FP32 performance.
The NVIDIA GeForce GTX 470 also excels in its board design, featuring a dual-slot width, a TDP of 215 W, and a recommended PSU of 550 W. It is equipped with two DVI outputs and requires two 6-pin power connectors. Its board number, P1025 SKU 0006, ensures easy identification and compatibility with other NVIDIA products.
This graphics card supports the latest graphics features, including DirectX 12, OpenGL 4.6, OpenCL 1.1, and Shader Model 5.1. It also offers support for CUDA 2.0, which allows for parallel processing and significantly improves certain tasks such as video rendering and scientific computations. While it does not support Vulkan, it still offers impressive performance for its time.
Designed for high-performance graphics, the GF100 GPU used in the GTX 470 comes with various improvements such as updated CUDA cores and a unified shader architecture. Its powerful design was adopted by various manufacturers in their retail boards, and it can be found in nine different models, each with its unique features.
In conclusion, the NVIDIA GeForce GTX 470 is a highly efficient graphics card designed for top-tier gaming and professional use. It comes equipped with a powerful GF100 GPU, high-speed GDDR5 memory, and impressive clock speeds, making it a reliable and powerful option for demanding tasks. Its board design, support for the latest graphics features, and compatibility with other NVIDIA products make it a versatile choice for enthusiasts and professionals alike.
NVIDIA GeForce GTX 470 - Full Technical Specifications
Graphics Processor
| GPU Variant | GF100-275-A3 |
|---|---|
| Architecture | Fermi |
| Foundry | TSMC |
| Process Size | 40 nm |
| Transistors | 3,100 million |
| Density | 5.9M / mm² |
| Die Size | 529 mm² |
| Chip Package | BGA-1980 |
Graphics Card
| Release Date | Mar 26th, 2010 |
|---|---|
| Production | End-of-life |
| Launch Price | 349 USD |
| Bus Interface | PCIe 2.0 x16 |
| Reviews | 48 in our database |
Clock Speeds
| GPU Clock | 608 MHz |
|---|---|
| Shader Clock | 1215 MHz |
| Memory Clock | 837 MHz |
Memory
| Memory Size | 1280 MB |
|---|---|
| Memory Type | GDDR5 |
| Memory Bus | 320 bit |
| Bandwidth | 133.9 GB/s |
Render Config
| Shading Units | 448 |
|---|---|
| TMUs | 56 |
| ROPs | 40 |
| SM Count | 14 |
| L1 Cache | 64 KB (per SM) |
| L2 Cache | 640 KB |
Theoretical Performance
| Pixel Rate | 17.02 GPixel/s |
|---|---|
| Texture Rate | 34.05 GTexel/s |
| FP32 (float) | 1,089 GFLOPS |
| FP64 (double) | 136.1 GFLOPS |
Board Design
| Slot Width | Dual-slot |
|---|---|
| Length | 241 mm |
| TDP | 215 W |
| Suggested PSU | 550 W |
| Outputs | 2x DVI |
| Power Connectors | 2x 6-pin |
| Board Number | P1025 SKU 0006 |
Graphics Features
| DirectX | 12 |
|---|---|
| OpenGL | 4.6 |
| OpenCL | 1.1 |
| Vulkan | N/A |
| CUDA | 2.0 |
| Shader Model | 5.1 |
