NVIDIA GeForce GTX 1080 Ti vs NVIDIA GeForce RTX 3070 Comparison

The Verdict

The recorded data shows a clear overall winner: the NVIDIA GeForce RTX 3070 wins 9 of 10 head-to-head benchmarks against the NVIDIA GeForce GTX 1080 Ti. The average benchmark score favors the RTX 3070 at 17,208 versus 15,548 for the GTX 1080 Ti, a meaningful gap of roughly 10.7%. In terms of percentile ranking among all GPUs, the RTX 3070 sits at the 61st percentile, while the GTX 1080 Ti sits at the 58th percentile. The RTX 3070 is the choice for most modern workloads, especially DirectX 12 and compute-heavy tasks, where its architectural advantages show up clearly. However, the GTX 1080 Ti retains one significant area of supremacy: Vulkan performance, where it leads by 48.1%. Users targeting Vulkan-based titles or older API scenarios should consider the GTX 1080 Ti. For everything else, the RTX 3070 is the stronger pick based on the benchmark results. The GTX 1080 Ti also offers more memory (11 GB versus 8 GB) and higher memory bandwidth (484.4 GB/s versus 448.0 GB/s), which may matter in specific capacity-limited scenarios, but the RTX 3070 compensates with a newer architecture and higher raw compute throughput.

Architecture Differences

The two cards come from different NVIDIA generations and manufacturing processes. The RTX 3070 uses the GA104 chip on the Ampere architecture, built on an 8 nm process at Samsung. It packs 17,400 million transistors into a 392 mm² die, yielding a transistor density of 44.4M per mm². The GTX 1080 Ti uses the GP102 chip on the Pascal architecture, built on a 16 nm process at TSMC. It contains 11,800 million transistors on a larger 471 mm² die, resulting in a lower density of 25.1M per mm². The newer process allows the RTX 3070 to fit more transistors in a smaller area, which directly contributes to its higher compute capabilities.

The RTX 3070 features 5,888 shading units, 184 texture mapping units, and 96 raster output units. It also includes dedicated ray tracing cores (46) and tensor cores (184), which are absent from the GTX 1080 Ti. The GTX 1080 Ti offers 3,584 shading units, 224 TMUs, and 88 ROPs. While the GTX 1080 Ti has more TMUs, the RTX 3070 has more shaders and ROPs. Clock speeds are similar: the RTX 3070 runs at a base of 1500 MHz and boosts to 1725 MHz, while the GTX 1080 Ti runs at 1481 MHz base and 1582 MHz boost. The RTX 3070's higher boost clock, combined with its larger shader count, explains its FP32 throughput of 20.31 TFLOPS versus 11.34 TFLOPS for the GTX 1080 Ti. The FP16 ratio also differs dramatically: the RTX 3070 achieves 20.31 TFLOPS FP16 (1:1 ratio with FP32), while the GTX 1080 Ti manages only 177.2 GFLOPS FP16 (1:64 ratio), making the RTX 3070 far more capable in half-precision workloads.

Memory configurations diverge as well. The RTX 3070 uses 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s bandwidth. The GTX 1080 Ti uses 11 GB of GDDR5X on a 352-bit bus, delivering 484.4 GB/s. Despite the GTX 1080 Ti's higher capacity and bandwidth, the RTX 3070's faster memory clock (14 Gbps effective versus 11 Gbps effective) compensates in many tests. The bus interface also differs: PCIe 4.0 x16 on the RTX 3070 versus PCIe 3.0 x16 on the GTX 1080 Ti. Display outputs are similar, but the RTX 3070 supports HDMI 2.1, while the GTX 1080 Ti uses HDMI 2.0. The RTX 3070 also supports DirectX 12 Ultimate (12_2), while the GTX 1080 Ti is limited to DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4. Power requirements differ: the RTX 3070 has a 220 W TDP with a single 12-pin connector and a suggested 550 W PSU, whereas the GTX 1080 Ti has a 250 W TDP with a 6-pin plus 8-pin connector and a suggested 600 W PSU. Physically, the RTX 3070 is shorter at 242 mm versus 267 mm for the GTX 1080 Ti, though both are dual-slot cards.

Where Each One Wins

The RTX 3070 dominates in nearly every measured category. Its largest victory comes in Geekbench OpenCL, where it scores 112,821 versus 67,929, a 66.1% lead. This indicates a substantial advantage in general-purpose compute workloads, likely driven by the higher shader count and newer architecture. In 3DMark Steel Nomad (DirectX 12), the RTX 3070 scores 3,162 versus 2,231, a 41.7% lead, showing strong performance in modern game engines that leverage DirectX 12 features. Passmark DirectX 12 also favors the RTX 3070, with a score of 85 versus 66, a 28.8% advantage. The RTX 3070 also wins in DirectX 10 (150 versus 118, 27.1% lead), DirectX 11 (182 versus 151, 20.5% lead), and Passmark G3D (22,214 versus 18,600, 19.4% lead). Compute performance in Passmark GPU Compute shows a 16.2% lead (11,195 versus 9,632). Even older APIs favor the RTX 3070: DirectX 9 scores 247 versus 231, a 6.9% lead, and 2D performance is slightly better (1,001 versus 939, 6.6% lead).

The GTX 1080 Ti wins only in Geekbench Vulkan, scoring 40,511 versus 21,022, a 48.1% advantage. This is a substantial reversal and suggests that the Pascal architecture has a stronger Vulkan driver path or that the benchmark favors its specific memory subsystem. Users running Vulkan-based applications, such as certain Linux gaming configurations or specific emulators, may see better results with the GTX 1080 Ti. The GTX 1080 Ti also has more VRAM (11 GB versus 8 GB) and higher bandwidth (484.4 GB/s versus 448.0 GB/s), which could help in texture-heavy scenes or workloads that approach the 8 GB limit of the RTX 3070, though the head-to-head benchmarks do not show a scenario where this is decisive.

FAQ

Q: Which card is faster in DirectX 12 workloads?

A: The RTX 3070 is faster. In 3DMark Steel Nomad (DirectX 12), it scores 3,162 against 2,231, a 41.7% lead. In Passmark DirectX 12, it scores 85 against 66, a 28.8% advantage. The RTX 3070 also supports DirectX 12 Ultimate (12_2), while the GTX 1080 Ti is limited to DirectX 12 (12_1).

Q: Does the GTX 1080 Ti win in any benchmark?

A: Yes, the GTX 1080 Ti wins in Geekbench Vulkan, scoring 40,511 versus 21,022, a 48.1% margin. This is its only win out of 10 head-to-head benchmarks.

Q: How do the compute capabilities compare?

A: The RTX 3070 is substantially stronger in compute. It scores 112,821 in Geekbench OpenCL versus 67,929 for the GTX 1080 Ti, a 66.1% lead. In Passmark GPU Compute, the RTX 3070 scores 11,195 versus 9,632, a 16.2% lead. The RTX 3070 also has 20.31 TFLOPS FP32 and FP16 (1:1), while the GTX 1080 Ti has 11.34 TFLOPS FP32 and only 177.2 GFLOPS FP16 (1:64).

Q: Which card has more memory and bandwidth?

A: The GTX 1080 Ti has more memory: 11 GB of GDDR5X versus 8 GB of GDDR6 on the RTX 3070. The GTX 1080 Ti also has higher bandwidth at 484.4 GB/s versus 448.0 GB/s. However, the RTX 3070 has a faster effective memory clock (14 Gbps versus 11 Gbps).

Q: What are the power requirements?

A: The RTX 3070 has a 220 W TDP with a single 12-pin connector and a suggested 550 W PSU. The GTX 1080 Ti has a 250 W TDP with a 6-pin plus 8-pin connector and a suggested 600 W PSU. Both are dual-slot cards.

Q: How do the cards compare in overall average score?

A: The RTX 3070 has an average benchmark score of 17,208, while the GTX 1080 Ti has 15,548. The RTX 3070 sits at the 61st percentile of all GPUs, and the GTX 1080 Ti sits at the 58th percentile.

Head-to-Head Benchmarks

The largest win for the RTX 3070 is in Geekbench OpenCL, with a score of 112,821 versus 67,929, a 66.1% delta. This is a massive gap and indicates that the Ampere architecture's compute throughput is far ahead of Pascal's in this workload. The second-largest win is in 3DMark Steel Nomad (DirectX 12), where the RTX 3070 scores 3,162 versus 2,231, a 41.7% lead. This benchmark is representative of modern game engines, and the RTX 3070's ability to leverage DirectX 12 Ultimate features (including ray tracing cores and tensor cores) likely explains the result. Passmark DirectX 12 shows a 28.8% lead (85 versus 66), reinforcing the DirectX 12 advantage. Passmark DirectX 10 shows a 27.1% lead (150 versus 118), and Passmark DirectX 11 shows a 20.5% lead (182 versus 151). In Passmark G3D, the RTX 3070 scores 22,214 versus 18,600, a 19.4% lead, which is a broad 3D performance indicator. Passmark GPU Compute shows a 16.2% lead (11,195 versus 9,632). The smallest wins for the RTX 3070 are in DirectX 9 (247 versus 231, 6.9%) and Passmark G2D (1,001 versus 939, 6.6%), showing that even in legacy 2D and older API scenarios, the newer card holds an edge.

The GTX 1080 Ti's sole victory is in Geekbench Vulkan, scoring 40,511 versus 21,022, a 48.1% margin. This is a large delta and suggests that the Pascal architecture has a particular strength in Vulkan execution, possibly due to driver maturity or memory subsystem design. It is notably the GTX 1080 Ti's memory bandwidth (484.4 GB/s) is higher than the RTX 3070's (448.0 GB/s), which may help in Vulkan's low-level memory access patterns. However, this single win does not offset the RTX 3070's dominance in the other nine benchmarks. The overall win count is 9 for the RTX 3070 and 1 for the GTX 1080 Ti, and the average scores (17,208 versus 15,548) reflect the RTX 3070's superiority in the aggregate. Users should weigh the Vulkan result against the broader DirectX and compute wins when making a choice.