NVIDIA GeForce GTX 780 Ti vs NVIDIA GeForce RTX 3070 Mobile Comparison

Where Each One Wins

The benchmark data splits these two NVIDIA parts along generational and workload lines. The GTX 780 Ti, a desktop flagship from the Kepler era, posts competitive scores in compute-oriented OpenCL and Vulkan tests, but it does not win a single head-to-head matchup against the RTX 3070 Mobile. The database records zero wins for the 780 Ti and two wins for the mobile Ampere part. That outcome is not close: the RTX 3070 Mobile leads by 70.6% in Geekbench OpenCL and by 68.6% in Geekbench Vulkan. For any user prioritizing raw compute throughput in cross-API workloads, the mobile chip is the clear choice. The GTX 780 Ti, however, remains relevant in legacy contexts. Its average benchmark score of 24,236 places it in the 70th percentile of all GPUs, which is actually higher than the RTX 3070 Mobile's 65th percentile, despite the mobile part's superior absolute scores in the two shared tests. This discrepancy reflects the fact that the 780 Ti has a much smaller set of benchmark records, and its rivals in the database include desktop cards like the GTX 1630 and RX 6600 XT, whereas the RTX 3070 Mobile competes against other mobile and older workstation parts.

The RTX 3070 Mobile wins decisively in every shared test, but the GTX 780 Ti shows strength in its own recorded metrics. The 780 Ti achieves a Geekbench Metal score of 18,144, a Geekbench OpenCL score of 27,326, and a Geekbench Vulkan score of 27,238. These are not negligible numbers for a 2013 card. The RTX 3070 Mobile, by contrast, reaches 92,939 in OpenCL and 86,768 in Vulkan, which are roughly 3.4x and 3.2x higher respectively. The mobile part also has a Passmark G3D score of 15,309, a Passmark GPU compute score of 6,827, and a 3DMark Steel Nomad DX12 score of 2,380, none of which the 780 Ti has a recorded result for. So the use case split is stark: the RTX 3070 Mobile is the compute and modern API champion, while the GTX 780 Ti only makes sense for older workloads or as a legacy desktop card, and even then its average score is lower than its nearest rival, the GTX 1630, by 0.2%.

Architecture Differences

The two cards come from different foundries, nodes, and design philosophies. The GTX 780 Ti uses the GK110B chip, built on Kepler architecture, fabricated by TSMC on a 28 nm process. The die measures 561 mm² and packs 7,080 million transistors, which gives a transistor density of 12.6 million per square millimeter. The RTX 3070 Mobile uses the GA104 chip, based on Ampere, made by Samsung on an 8 nm process. Its die is smaller at 392 mm², but it contains 17,400 million transistors, resulting in a density of 44.4 million per square millimeter. That is more than 3.5x the density of the Kepler part, reflecting the generational leap in manufacturing.

Memory is another major differentiator. The GTX 780 Ti has 3 GB of GDDR5 on a 384-bit bus, delivering 336.6 GB/s of bandwidth. The RTX 3070 Mobile has 8 GB of GDDR6 on a 256-bit bus, with bandwidth of 448.0 GB/s. Although the bus is narrower, the higher effective memory speed of 14 Gbps versus 7 Gbps gives the newer part more bandwidth. The clock speeds also tell a story: the 780 Ti runs at 875 MHz base and 928 MHz boost, while the RTX 3070 Mobile runs at 1110 MHz base and 1560 MHz boost. The mobile part is over 600 MHz higher at boost.

Compute resources are massively different. The 780 Ti has 2,880 shading units, 240 texture mapping units, and 48 ROPs. The RTX 3070 Mobile has 5,120 shading units, 160 TMUs, and 80 ROPs. The newer card has more shaders and ROPs but fewer TMUs. Critically, the RTX 3070 Mobile adds 40 RT cores and 160 tensor cores, which the Kepler card lacks entirely. The FP32 throughput is 5.345 TFLOPS for the 780 Ti versus 15.97 TFLOPS for the RTX 3070 Mobile, and the Ampere card also supports FP16 at the same 15.97 TFLOPS rate, while the 780 Ti has no listed FP16 figure. Pixel rate jumps from 55.68 GPixel/s to 124.8 GPixel/s, though texture rate is closer: 222.7 GTexel/s versus 249.6 GTexel/s.

Power draw differs, with the 780 Ti rated at 250 W TDP, a dual-slot card requiring one 6-pin and one 8-pin connector plus a 600 W PSU, while the RTX 3070 Mobile is 115 W TDP with no power connectors listed, as it is a portable device. The bus interface also changes from PCIe 3.0 x16 to PCIe 4.0 x16. API support is another differentiator: the 780 Ti supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.175, whereas the RTX 3070 Mobile supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The newer card adds hardware ray tracing and tensor cores for AI workloads, which the Kepler card cannot do.

Head-to-Head Benchmarks

The only two shared benchmark results are Geekbench OpenCL and Geekbench Vulkan, and the RTX 3070 Mobile dominates both. In OpenCL, the 780 Ti scores 27,326, and the RTX 3070 Mobile scores 92,939, a delta of -70.6%, meaning the mobile card is 3.4 times faster. In Vulkan, the 780 Ti scores 27,238, and the RTX 3070 Mobile scores 86,768, a delta of -68.6%, or about 3.2 times faster. These are massive gaps, but they should be contextualized. The 780 Ti is a 2013 desktop card with 3 GB VRAM and no ray tracing, while the RTX 3070 Mobile is a 2021 mobile chip with 8 GB VRAM and the full Ampere feature set.

The GTX 780 Ti does have other benchmark records in the database, but they are not shared with the RTX 3070 Mobile. Its Geekbench Metal score is 18,144, which is lower than its OpenCL and Vulkan scores, indicating that Metal may not be a preferred API for this architecture. The RTX 3070 Mobile, meanwhile, has a wide range of Passmark results: DirectX 9 at 160, DirectX 10 at 113, DirectX 11 at 138, DirectX 12 at 64, G2D at 641, G3D at 15,309, and GPU compute at 6,827. It also has a 3DMark Steel Nomad DX12 score of 2,380. The low DirectX 12 score in Passmark (64) is notably lower than the DirectX 11 score (138), but the 3DMark DX12 result is separate and shows the card handling modern workloads. No comparable numbers exist for the 780 Ti, so a direct comparison in these tests is impossible.

Looking at the nearest rivals for each card provides additional context. The 780 Ti's average score of 24,236 is nearly identical to the GTX 1630 (24,277, a -0.2% difference) and the RX 6600 XT (24,442, a -0.8% difference), but it outperforms the RTX 2080 SUPER (24,170, +0.3%) and the RX 6800S (24,063, +0.7%) by small margins. The RTX 3070 Mobile's average score of 20,534 is close to the Intel Arc B570 (20,556, -0.1%), the Arc A750 (20,582, -0.2%), the Quadro M4000M (20,480, +0.3%), and the R9 M390X (20,662, -0.6%). These rival comparisons show that the 780 Ti sits in a higher performance tier in the database, despite being older, because it is a desktop card with a higher power budget and a 561 mm² die.

FAQ

Q: Which GPU has a higher average benchmark score, and by how much?

A: The GTX 780 Ti has an average benchmark score of 24,236, while the RTX 3070 Mobile has an average of 20,534. The 780 Ti is 15% higher, which is noteworthy given the RTX 3070 Mobile is a newer architecture.

Q: Does the RTX 3070 Mobile support ray tracing?

A: Yes, it has 40 RT cores. The GTX 780 Ti does not have any RT cores, as it predates the ray tracing feature set.

Q: What is the memory capacity and type for each card?

A: The GTX 780 Ti has 3 GB of GDDR5 on a 384-bit bus with 336.6 GB/s bandwidth. The RTX 3070 Mobile has 8 GB of GDDR6 on a 256-bit bus with 448.0 GB/s bandwidth.

Q: Which card has higher boost clock?

A: The RTX 3070 Mobile has a boost clock of 1560 MHz, which is 632 MHz higher than the GTX 780 Ti's 928 MHz boost.

Q: In the OpenCL benchmark, what is the exact score gap?

A: The RTX 3070 Mobile scores 92,939, and the GTX 780 Ti scores 27,326. The RTX 3070 Mobile is 70.6% faster.

Q: Which card has a higher transistor density?

A: The RTX 3070 Mobile has a density of 44.4 million transistors per square millimeter, versus 12.6 million for the GTX 780 Ti, due to the 8 nm process versus 28 nm.

The Verdict

The data does not favor the GTX 780 Ti in any direct comparison. In the two shared benchmarks, the RTX 3070 Mobile wins by 70.6% and 68.6%. The mobile card has more than double the FP32 throughput (15.97 TFLOPS versus 5.345 TFLOPS), more than double the shading units (5,120 versus 2,880), and three times the memory capacity (8 GB versus 3 GB). It also has hardware ray tracing and tensor cores, which the Kepler card lacks entirely. For anyone running modern APIs, larger VRAM, or ray-traced titles, the RTX 3070 Mobile is the only sensible choice from these two.

However, the GTX 780 Ti is not without context. Its average benchmark score is 24,236, which places it at the 70th percentile of all GPUs, compared to the RTX 3070 Mobile's 65th percentile. The 780 Ti also outperforms the RTX 2080 SUPER and RX 6800S in average score, by 0.3% and 0.7% respectively, which is surprising for a 2013 card. But those are average scores across a different set of tests. The recorded data shows that the GTX 780 Ti only has three benchmark entries, all from Geekbench, while the RTX 3070 Mobile has ten, including 3DMark and Passmark. So the percentile and average figures are not directly comparable.

The verdict is simple: if you need a GPU for a legacy desktop system, the GTX 780 Ti has a unique profile, but it is end-of-life, with a launch MSRP of 699 USD. If you need any modern feature, raw compute, or efficiency, the RTX 3070 Mobile wins. The 70.6% OpenCL lead and 68.6% Vulkan lead are decisive. The RTX 3070 Mobile is also a mobile part with lower TDP (115 W versus 250 W), making it viable for laptops, whereas the 780 Ti is a dual-slot desktop card. In short, for any practical use, the RTX 3070 Mobile is the superior product, and the only reason to pick the GTX 780 Ti is if you are specifically collecting older desktop hardware.