NVIDIA GeForce GTX 780 Ti vs NVIDIA GeForce RTX 2080 Comparison

The NVIDIA GeForce GTX 780 Ti and NVIDIA GeForce RTX 2080 represent two distinct eras of NVIDIA GPU design, separated by five years of architectural evolution. The GTX 780 Ti, built on the Kepler architecture, was a flagship of the GeForce 700 generation, while the RTX 2080, based on Turing, introduced dedicated ray tracing and tensor cores. Benchmark data shows the RTX 2080 delivers overwhelming performance advantages in modern compute workloads, yet the GTX 780 Ti still holds a place in the database's percentile rankings, sitting at the 70th percentile versus the RTX 2080's 68th percentile. The average benchmark score for the GTX 780 Ti is 24,236, while the RTX 2080 averages 22,895, a counterintuitive result driven by the specific test suites available for each card.

Where Each One Wins

The RTX 2080 wins every head-to-head benchmark recorded in the database, and its dominance is not marginal. In Geekbench OpenCL, the RTX 2080 scores 91,313 against the GTX 780 Ti's 27,326, a delta of -70.1% from the perspective of the older card. In Geekbench Vulkan, the RTX 2080 scores 107,797 against 27,238, a -74.7% delta. These are not incremental gains; they represent a generational leap in raw compute throughput, memory bandwidth, and API support. The RTX 2080's wins are absolute in this comparison, with 2 wins and 0 losses in the head-to-head test set.

The GTX 780 Ti, however, is not without merit in its own benchmark suite. Its Geekbench Metal score of 18,144 and OpenCL score of 27,326 show it remains functional in legacy compute contexts, but it lacks the RTX 2080's breadth of modern API support. The RTX 2080's PassMark scores span DirectX 9 through DirectX 12, with a G3D score of 18,720 and a GPU compute score of 7,872, indicating balanced performance across older and newer graphics APIs. The GTX 780 Ti, by contrast, has no PassMark entries in the provided data, limiting its measurable scope. For users targeting modern workloads—ray tracing, DX12 Ultimate, Vulkan 1.4—the RTX 2080 is the only viable choice. For legacy DirectX 9 or 10 applications, the GTX 780 Ti's architecture may still suffice, though no direct comparison data exists for those tests.

Architecture Differences

The fundamental architectural split is node and transistor count. The GTX 780 Ti uses a 28 nm process at TSMC, packing 7,080 million transistors on a 561 mm² die, yielding a transistor density of 12.6M per mm². The RTX 2080 moves to a 12 nm process, also TSMC, with 13,600 million transistors on a slightly smaller 545 mm² die, achieving a density of 25.0M per mm². This near-doubling of transistor density is the primary driver of performance differences. The GTX 780 Ti's chip is GK110B, part of the Kepler architecture, while the RTX 2080 uses TU104, a Turing chip. Turing introduces two feature classes absent from Kepler: 46 RT cores and 368 tensor cores. The RTX 2080 also supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the GTX 780 Ti is limited to DirectX 12 (11_1) and Vulkan 1.2.175. Both cards support OpenGL 4.6.

Memory architecture diverges sharply. The GTX 780 Ti has 3 GB of GDDR5 on a 384-bit bus, delivering 336.6 GB/s of bandwidth. The RTX 2080 has 8 GB of GDDR6 on a 256-bit bus, but achieves 448.0 GB/s due to higher effective clock speeds (14 Gbps vs 7 Gbps). The RTX 2080's smaller bus is compensated by faster memory, and the larger capacity is critical for modern texture-heavy workloads. Clock speeds also differ: the GTX 780 Ti runs at 875 MHz base and 928 MHz boost, while the RTX 2080 starts at 1515 MHz base and 1710 MHz boost. Pixel rate jumps from 55.68 GPixel/s to 109.4 GPixel/s, and texture rate from 222.7 GTexel/s to 314.6 GTexel/s. FP32 compute nearly doubles from 5.345 TFLOPS to 10.07 TFLOPS, with the RTX 2080 also offering FP16 at 20.14 TFLOPS (2:1), a capability the GTX 780 Ti lacks entirely.

Head-to-Head Benchmarks

The two head-to-head tests in the database are both Geekbench compute workloads, and both are landslides. In Geekbench OpenCL, the RTX 2080's score of 91,313 is 3.34 times higher than the GTX 780 Ti's 27,326. The delta percentage of -70.1% indicates the GTX 780 Ti trails by over two-thirds. This is not a subtle difference; it suggests fundamental throughput advantages in the RTX 2080's shading units (2,944 vs 2,880), texture units (184 vs 240, actually lower), and ROPs (64 vs 48). The RTX 2080 has fewer TMUs but compensates with higher clocks and memory bandwidth. The OpenCL test likely stresses memory bandwidth and compute, where the RTX 2080's 448 GB/s versus 336.6 GB/s provides a clear edge.

In Geekbench Vulkan, the margin widens further. The RTX 2080 scores 107,797 versus 27,238, a -74.7% delta. Vulkan's low-level API access favors the newer architecture's driver maturity and hardware features. The RTX 2080's 368 tensor cores may also be leveraged in Vulkan compute workloads, though the database does not specify whether the test uses them. The GTX 780 Ti's Vulkan 1.2.175 support is older, and its Kepler architecture lacks the asynchronous compute optimizations present in Turing. The result is a 3.96x performance ratio, making the RTX 2080 nearly four times faster in this specific workload. There are no other head-to-head tests in the data, so these two benchmarks define the entire comparison: the RTX 2080 wins both, and wins them decisively.

Specification Differences

The following fields differ between the two cards, based solely on the FACT PACK:

| Specification | GTX 780 Ti | RTX 2080 |

|---|---|---|

| Architecture | Kepler | Turing |

| Process Node | 28 nm | 12 nm |

| Transistors | 7,080 million | 13,600 million |

| Die Size | 561 mm² | 545 mm² |

| Transistor Density | 12.6M / mm² | 25.0M / mm² |

| Base Clock | 875 MHz | 1515 MHz |

| Boost Clock | 928 MHz | 1710 MHz |

| Memory Size | 3 GB | 8 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus | 384 bit | 256 bit |

| Memory Bandwidth | 336.6 GB/s | 448.0 GB/s |

| Memory Effective Speed | 7 Gbps | 14 Gbps |

| Shading Units | 2880 | 2944 |

| TMUs | 240 | 184 |

| ROPs | 48 | 64 |

| RT Cores | None | 46 |

| Tensor Cores | None | 368 |

| Pixel Rate | 55.68 GPixel/s | 109.4 GPixel/s |

| Texture Rate | 222.7 GTexel/s | 314.6 GTexel/s |

| FP32 | 5.345 TFLOPS | 10.07 TFLOPS |

| FP16 | Not specified | 20.14 TFLOPS (2:1) |

| TDP | 250 W | 215 W |

| Suggested PSU | 600 W | 550 W |

| Display Outputs | 2x DVI, 1x HDMI 1.4a, 1x DP 1.2 | 1x HDMI 2.0, 3x DP 1.4a, 1x USB Type-C |

| DirectX | 12 (11_1) | 12 Ultimate (12_2) |

| Vulkan | 1.2.175 | 1.4 |

| Height | 111 mm (4.4 inches) | 116 mm (4.6 inches) |

| Width | 38 mm (1.5 inches) | 35 mm (1.4 inches) |

Notably, the GTX 780 Ti has a higher TDP (250 W vs 215 W) and larger physical footprint in width, despite being older. The RTX 2080 is more power-efficient per transistor, a direct result of the 12 nm node. Both cards use the same power connector configuration (1x 6-pin + 1x 8-pin) and have identical length (267 mm, 10.5 inches) and bus interface (PCIe 3.0 x16). The GTX 780 Ti's launch MSRP was 699 USD, and the RTX 2080's launch MSRP was also 699 USD.

FAQ

Q: Which card has better compute performance in OpenCL?

A: The RTX 2080 dominates, scoring 91,313 in Geekbench OpenCL versus the GTX 780 Ti's 27,326, a delta of -70.1% for the older card. The RTX 2080's higher FP32 throughput (10.07 TFLOPS vs 5.345 TFLOPS) and memory bandwidth (448.0 GB/s vs 336.6 GB/s) explain the gap.

Q: Does the GTX 780 Ti support ray tracing or tensor cores?

A: No. The GTX 780 Ti has no RT cores and no tensor cores. The RTX 2080 includes 46 RT cores and 368 tensor cores, which are exclusive to the Turing architecture.

Q: What is the memory capacity difference?

A: The GTX 780 Ti has 3 GB of GDDR5 on a 384-bit bus, while the RTX 2080 has 8 GB of GDDR6 on a 256-bit bus. Despite the narrower bus, the RTX 2080 has higher bandwidth (448.0 GB/s vs 336.6 GB/s) due to faster 14 Gbps effective memory.

Q: Which card has a higher transistor density?

A: The RTX 2080 achieves 25.0M transistors per mm² on a 545 mm² die, versus the GTX 780 Ti's 12.6M per mm² on a 561 mm² die. The RTX 2080 packs nearly double the density despite a smaller physical die.

Q: Are there any benchmark tests where the GTX 780 Ti wins?

A: In the provided head-to-head data, the GTX 780 Ti wins zero tests. The RTX 2080 wins both Geekbench OpenCL and Geekbench Vulkan. The GTX 780 Ti's own benchmarks (Geekbench Metal: 18,144, OpenCL: 27,326, Vulkan: 27,238) show it is competitive with other cards in its percentile range, but not against the RTX 2080.

Q: How do the API supports compare?

A: The RTX 2080 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the GTX 780 Ti is limited to DirectX 12 (11_1) and Vulkan 1.2.175. Both support OpenGL 4.6. The RTX 2080's newer API support enables features like hardware ray tracing and mesh shaders that the GTX 780 Ti cannot access.