Quick Answer
The GeForce RTX 3070 and RTX 2080 are both high-performance graphics cards from different generations. The RTX 3070, based on the newer Ampere architecture, generally offers performance that is comparable to or slightly exceeds the previous-generation RTX 2080. This makes the RTX 3070 a notable example of a newer mid-range card matching the performance of a previous-generation flagship.
GeForce RTX 3070 vs GeForce RTX 2080: Full Comparison
Introduction
When evaluating graphics cards, comparing models across different generations can provide insight into the pace of technological progress. The GeForce RTX 2080 was a high-end offering from the Turing generation, while the RTX 3070 arrived as a more accessible option in the subsequent Ampere lineup. This comparison examines their architectural differences, performance metrics, and feature sets to clarify their positions relative to each other, helping users understand the trade-offs and advancements between the two.
Architecture and Core Specifications
The fundamental difference between these cards lies in their underlying architecture. The RTX 2080 utilizes the Turing architecture, while the RTX 3070 is built on the newer Ampere architecture. This generational shift brings changes in core design, efficiency, and feature implementation.
- RTX 2080 (Turing): Features TU104 GPU with 2944 CUDA Cores, 46 RT Cores for ray tracing, and 368 Tensor Cores for AI-driven tasks like DLSS. It typically came with 8GB of GDDR6 memory on a 256-bit bus.
- RTX 3070 (Ampere): Features GA104 GPU with 5888 CUDA Cores, a significant increase on paper. It includes 46 RT Cores (2nd Gen) and 184 Tensor Cores (3rd Gen). It also typically features 8GB of GDDR6 memory on a 256-bit bus.
The Ampere architecture’s 2nd Gen RT Cores and 3rd Gen Tensor Cores are designed to be more efficient at ray tracing and DLSS computations compared to their Turing counterparts.
Performance in Gaming and Applications
In synthetic benchmarks and real-world gaming tests, the performance relationship between these two cards is well-defined.
- Traditional Rasterization (Non-RT): The RTX 3070 typically matches or slightly outperforms the RTX 2080 in games at resolutions like 1440p and 1080p. The performance uplift is generally attributed to the higher CUDA core count and architectural improvements of Ampere.
- Ray Tracing Performance: With dedicated hardware for ray tracing (RT Cores), both cards are capable. However, the RTX 3070’s 2nd Gen RT Cores often provide a performance advantage in ray-traced titles, sometimes by a more noticeable margin than in traditional gaming.
- DLSS Performance: The RTX 3070’s 3rd Gen Tensor Cores can provide more efficient upscaling when DLSS (Deep Learning Super Sampling) is enabled, which can lead to higher frame rates in supported games compared to the RTX 2080 using the same quality setting.
Power, Thermals, and Features
Efficiency and supported features are other key areas of comparison.
- Power Consumption: The RTX 3070 has a typical board power (TBP) rating of 220W, while the RTX 2080 is rated at 215W. In practice, their power draw under load is often similar, though actual consumption can vary based on the specific card model and cooling solution.
- Thermal Design: Both cards require robust cooling. Thermal performance is highly dependent on the custom design from the board partner (e.g., dual-fan or triple-fan coolers) rather than the GPU model itself.
- Feature Support: Both cards support real-time ray tracing, DLSS, and NVIDIA’s suite of technologies like NVIDIA Reflex. The RTX 3070, being newer, may have longer driver support and optimizations for future game titles.
Comparison Table
| Feature | GeForce RTX 3070 | GeForce RTX 2080 |
|---|---|---|
| GPU Architecture | Ampere (GA104) | Turing (TU104) |
| CUDA Cores | 5888 | 2944 |
| RT Cores | 46 (2nd Generation) | 46 (1st Generation) |
| Tensor Cores | 184 (3rd Generation) | 368 (2nd Generation) |
| Base / Boost Clock | ~1500 MHz / ~1725 MHz* | ~1515 MHz / ~1710 MHz* |
| VRAM | 8 GB GDDR6 | 8 GB GDDR6 |
| Memory Bus | 256-bit | 256-bit |
| Memory Bandwidth | 448 GB/s | 448 GB/s |
| Typical Board Power (TBP) | 220 W | 215 W |
| Recommended PSU | 650 W | 650 W |
| PCIe Interface | PCIe 4.0 | PCIe 3.0 |
| Key Technologies | Ray Tracing, DLSS (3rd Gen Tensor), NVIDIA Reflex | Ray Tracing, DLSS (2nd Gen Tensor), NVIDIA Reflex |
*Clock speeds can vary significantly between different custom models from board partners.
Frequently Asked Questions (FAQ)
What is the main performance difference between the RTX 3070 and RTX 2080?
The RTX 3070 generally offers performance that is similar to or slightly better than the RTX 2080 in most games, particularly at 1440p resolution. Its advantage can be more pronounced in titles that utilize ray tracing and DLSS due to its newer generation cores.
Does the RTX 3070 have more VRAM than the RTX 2080?
No, both cards typically feature 8GB of GDDR6 memory. The memory configuration (bus width and bandwidth) is also identical between the two models.
Which card is more power-efficient, the RTX 3070 or RTX 2080?
Their rated power consumption is very close (220W vs 215W). The RTX 3070’s Ampere architecture is often considered more efficient on a performance-per-watt basis, meaning it can deliver similar or higher performance for a comparable power draw.
Is PCIe 4.0 support on the RTX 3070 a major advantage?
For most current games and applications, the difference between PCIe 3.0 (RTX 2080) and PCIe 4.0 (RTX 3070) is minimal. The advantage of PCIe 4.0 may become more relevant for specific professional workloads or with future technologies that can saturate the higher bandwidth.
Final Thoughts
The comparison between the GeForce RTX 3070 and RTX 2080 illustrates a common pattern in GPU generations: a newer mid-range model often reaches the performance level of a previous-generation flagship. The RTX 3070 achieves this through its advanced Ampere architecture, offering comparable traditional performance while typically providing enhanced efficiency in ray tracing and DLSS tasks. The RTX 2080 remains a capable card, but the architectural benefits of the RTX 3070 generally position it as the more modern option. The choice between them often depends on specific market conditions and the value placed on newer feature sets versus established performance.