EVGA GeForce RTX 3080 FTW3 Ultra Gaming, 10G-P5-3897-KL

RTX 3080 Review & Specifications

The RTX 3080 is based on the Ampere architecture and manufactured using the 8 nm nm process node. It was launched approximately 5 years, 9 months ago, originally priced at $699. It features 8704 cores, including 38 compute units (SMs), 272 texture mapping units (TMUs), and 96 render output units (ROPs). For ray tracing and AI-accelerated tasks, it also includes 68 dedicated RT cores and 272 Tensor cores, enabling advanced rendering and upscaling features where supported. It has a base clock of 1440 MHz and a boost clock reaching up to 1710 MHz. The RTX 3080 comes equipped with 10 GB of GDDR6X memory across a 320-bit bus. Memory runs at 1188 MHz and delivers a total memory bandwidth of 760 GB/s.

This GPU has a very high thermal design power (TDP) of 320 W. A PC case with excellent airflow and a high-quality, high-wattage power supply are essential for stable operation.

RTX 3080 Performance

RTX 3080 is a mid-range graphics card, scoring 50.92% in relative gaming performance. However, the limited 10 GB of VRAM may lead to severe memory bottlenecks at 4K, even with DLSS or FSR. It’s best suited for high FPS in 1080p or smooth 1440p performance, but not recommended for demanding 4K gaming.

The RTX 3080 pairs a relatively powerful GPU core with a moderate VRAM size of 10 GB — a configuration that's becoming increasingly common. However, this pairing has sparked significant discussion in the gaming and tech review communities. Many modern GPUs with strong processing capabilities are equipped with insufficient VRAM (often less than 16 GB), which prevents them from realizing their full potential. As a result, GPUs like this one often deliver excellent performance in 1080p gaming at ultra settings, and perform well at 1440p in many titles. However, as modern games demand more VRAM — particularly due to high-resolution textures, ray tracing, and expansive open worlds — this limited memory becomes a bottleneck. In 4K gaming, performance can drop significantly due to VRAM exhaustion, leading to underutilization of the GPU core itself. Technologies like DLSS can mitigate some of the performance loss at higher resolutions, although this may come at the cost of visual fidelity. Content creators will find it suitable for HD and QHD editing, small-scale 3D work, and digital painting. Importantly, NVIDIA’s support for CUDA, cuDNN, and TensorRT gives this GPU a distinct edge in AI and ML workloads. It can handle inference, fine-tuning, and training of small to intermediate models with good efficiency, provided batch sizes and precision are tuned for memory constraints. An ideal starting point for AI learners and developers.

In terms of gaming value, EVGA GeForce RTX 3080 FTW3 Ultra Gaming, 10G-P5-3897-KL provides excellent performance per dollar, reaching 87.2% efficiency relative to the best in this category. It performs well in 1080p and 1440p, but due to the limited VRAM, it may struggle with future games or long-term use.