AMD Radeon RX 5700 vs NVIDIA GeForce GTX 780 Comparison

AMD Radeon RX 5700 vs NVIDIA GeForce GTX 780

The AMD Radeon RX 5700 is decisively ahead of the NVIDIA GeForce GTX 780 across every recorded benchmark in the database. The RX 5700 wins all three head-to-head tests, with the largest margin reaching 751.1% in Geekbench Metal. The GTX 780, while still a functioning GPU, trails by margins that range from 161.8% to 204.4% in the other two tests. The average benchmark score of the RX 5700 is 22170, compared to 19164 for the GTX 780, a difference of 3006 points. The RX 5700 sits at the 67th percentile among all GPUs, while the GTX 780 sits at the 64th percentile. For any workload measured here, the RX 5700 is the clear choice.

The Verdict

The data points to one conclusion: the AMD Radeon RX 5700 is the superior product for every benchmark category recorded in the database. Its average benchmark score of 22170 places it nearly 16% above the GTX 780's 19164. The RX 5700 wins all three head-to-head comparisons, and no test shows the GTX 780 taking a single victory. Users seeking maximum performance in Geekbench Metal, OpenCL, or Vulkan should choose the RX 5700.

The GTX 780, however, is not without a niche. Its percentile rank of 64 places it close to the RX 5700's 67th percentile, indicating that in the broader GPU landscape, both cards occupy similar tiers. The GTX 780's nearest rivals include the NVIDIA TITAN Xp with a delta of -0.1%, the Tesla K20m at 0.4%, and the RTX 4050 Mobile at 0.6%. These are all cards with average scores within 1% of the GTX 780, suggesting that the GTX 780 remains competitive against a specific set of older or mobile parts. But against the RX 5700, the margin is not close.

For professionals or enthusiasts who rely on compute workloads via OpenCL or Vulkan, the RX 5700 delivers 204.4% and 161.8% higher scores respectively. For those using Apple's Metal API, the RX 5700's 86079 score dwarfs the GTX 780's 10114. The RX 5700 is the pick for anyone prioritizing these modern APIs. The GTX 780, with its 2013 release date and Kepler architecture, offers no advantage in any measured test. The verdict is unambiguous: the RX 5700 is the faster card, and the GTX 780 should only be considered if it is already owned or available at a low cost, though the database does not track pricing.

Architecture Differences

The RX 5700 and GTX 780 differ fundamentally in their underlying designs. The RX 5700 uses the Navi 10 chip based on the RDNA 1.0 architecture, manufactured on a 7 nm process at TSMC. The GTX 780 uses the GK110 chip based on the Kepler architecture, manufactured on a 28 nm process, also at TSMC. This process difference is stark: 7 nm versus 28 nm. The RX 5700 packs 10,300 million transistors into a 251 mm² die, yielding a transistor density of 41.0 million per mm². The GTX 780 contains 7,080 million transistors on a much larger 561 mm² die, with a density of only 12.6 million per mm². The RX 5700 achieves higher density and smaller physical size simultaneously.

Memory configurations also diverge sharply. The RX 5700 features 8 GB of GDDR6 memory on a 256-bit bus, providing 448.0 GB/s of bandwidth. The GTX 780 has 3 GB of GDDR5 memory on a 384-bit bus, delivering 288.4 GB/s. Despite the wider bus on the GTX 780, the RX 5700's faster GDDR6 memory yields 159.6 GB/s more bandwidth. Clock speeds reflect the architectural leap: the RX 5700 runs at a base of 1465 MHz and a boost of 1725 MHz, with a game clock of 1625 MHz. The GTX 780 operates at 863 MHz base and 902 MHz boost. The RX 5700's memory clock is 1750 MHz (14 Gbps effective) versus 1502 MHz (6 Gbps effective) on the GTX 780.

Compute resources show a mixed picture. Both cards have 2304 shading units, but the RX 5700 has 144 texture mapping units and 64 ROPs, while the GTX 780 has 192 TMUs and 48 ROPs. The RX 5700's pixel rate is 110.4 GPixel/s versus 43.30 GPixel/s on the GTX 780, and its texture rate is 248.4 GTexel/s versus 173.2 GTexel/s. The RX 5700's FP32 throughput is 7.949 TFLOPS, nearly double the GTX 780's 4.156 TFLOPS. The RX 5700 also supports FP16 at 15.90 TFLOPS with a 2:1 ratio, while the GTX 780 has no recorded FP16 capability. The RX 5700 uses PCIe 4.0 x16, while the GTX 780 uses PCIe 3.0 x16. Power consumption reflects the efficiency gain: the RX 5700 has a TDP of 180 W and a suggested PSU of 450 W, while the GTX 780 has a TDP of 250 W and a suggested PSU of 600 W.

Where Each One Wins

Based on the recorded benchmarks, the RX 5700 wins in every category. The Geekbench Metal test shows the largest gap, with the RX 5700 scoring 86079 against the GTX 780's 10114, a delta of 751.1%. This indicates that the RX 5700 is exceptionally strong in Apple's Metal API, likely due to its newer architecture and higher memory bandwidth. The Geekbench OpenCL test shows the RX 5700 at 69603 versus 22863, a delta of 204.4%. This test measures general-purpose GPU compute, and the RX 5700's advantage here is substantial but smaller than Metal. The Geekbench Vulkan test shows the RX 5700 at 64166 versus 24514, a delta of 161.8%. Vulkan is a cross-platform graphics and compute API, and the RX 5700 still leads by a wide margin.

The GTX 780 has no benchmark wins in the database. Its best relative performance comes in Vulkan, where its score of 24514 is about 38% of the RX 5700's 64166. In OpenCL, the GTX 780 achieves 22863, which is just under 33% of the RX 5700's 69603. In Metal, the GTX 780's 10114 is only 11.7% of the RX 5700's 86079. The GTX 780's nearest rivals, the TITAN Xp and Tesla K20m, have average scores within 0.4% of its own, indicating that the GTX 780 is comparable to those older high-end parts. But against the RX 5700, the GTX 780 has no competitive segment.

For users running modern compute workloads, the RX 5700 is the only rational choice. For legacy applications that rely on older APIs not tested here, the GTX 780 may still function, but the database provides no evidence of any such advantage. The RX 5700's higher FP32 throughput, faster memory, and lower power draw (180 W versus 250 W) make it the better option for sustained workloads. The GTX 780's only potential role is as a low-cost secondary card for systems already equipped with adequate power supplies, but that is outside the scope of benchmark data.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The AMD Radeon RX 5700 has an average benchmark score of 22170, while the NVIDIA GeForce GTX 780 has an average score of 19164. The RX 5700 leads by 3006 points.

Q: How large is the performance gap in Geekbench Metal?

A: The RX 5700 scores 86079 in Geekbench Metal, compared to the GTX 780's 10114. This is a delta of 751.1%, the largest margin in any recorded test.

Q: Does the GTX 780 win any benchmark against the RX 5700?

A: No. The RX 5700 wins all three head-to-head benchmarks: Geekbench Metal, Geekbench OpenCL, and Geekbench Vulkan. The GTX 780 has zero wins in the database.

Q: What are the memory sizes and types for each card?

A: The RX 5700 has 8 GB of GDDR6 memory with a 256-bit bus and 448.0 GB/s bandwidth. The GTX 780 has 3 GB of GDDR5 memory with a 384-bit bus and 288.4 GB/s bandwidth.

Q: How do the TDP ratings compare?

A: The RX 5700 has a TDP of 180 W with a suggested PSU of 450 W. The GTX 780 has a TDP of 250 W with a suggested PSU of 600 W. The RX 5700 draws less power and requires a less powerful power supply.

Q: Which card has higher shading unit count?

A: Both cards have 2304 shading units. However, the RX 5700 has 144 TMUs and 64 ROPs, while the GTX 780 has 192 TMUs and 48 ROPs. The RX 5700 still achieves higher pixel and texture rates.

Head-to-Head Benchmarks

The most striking result is Geekbench Metal, where the RX 5700 scores 86079 and the GTX 780 scores 10114. This 751.1% delta is not a marginal improvement; it is a near-total dominance. The RX 5700's score is more than eight times higher. This test likely benefits from the RX 5700's modern RDNA architecture and its 448.0 GB/s memory bandwidth, which allows for faster data movement in Metal workloads. The GTX 780's Kepler architecture, with its 288.4 GB/s bandwidth and 3 GB memory capacity, cannot keep up.

In Geekbench OpenCL, the RX 5700 scores 69603 against the GTX 780's 22863, a delta of 204.4%. This is a smaller but still overwhelming lead. OpenCL is a general-purpose compute API, and the RX 5700's FP32 throughput of 7.949 TFLOPS versus the GTX 780's 4.156 TFLOPS explains part of the gap. The RX 5700 also has a higher texture rate (248.4 GTexel/s versus 173.2 GTexel/s) and pixel rate (110.4 GPixel/s versus 43.30 GPixel/s), which contribute to compute performance.

Geekbench Vulkan shows the RX 5700 scoring 64166 and the GTX 780 scoring 24514, a delta of 161.8%. Vulkan is a lower-level API that can expose hardware differences more directly. The RX 5700's PCIe 4.0 x16 interface, compared to the GTX 780's PCIe 3.0 x16, may reduce data transfer bottlenecks, though the card's internal compute capabilities are the primary driver. The GTX 780's 24514 score is its best relative showing, but it still trails by more than 2.6 times.

Across all three tests, the RX 5700 wins by margins that range from 161.8% to 751.1%. The GTX 780 has no recorded victory, and its scores are consistently lower in absolute terms. The database shows a clear hierarchy: the RX 5700 is a modern, high-performance GPU, while the GTX 780 is an older part that cannot match it in any measured workload.

Specification Differences

The two GPUs differ in nearly every specification category. The RX 5700 uses a 7 nm process node, while the GTX 780 uses 28 nm. The RX 5700 has 10,300 million transistors on a 251 mm² die, while the GTX 780 has 7,080 million transistors on a 561 mm² die. Transistor density is 41.0M per mm² for the RX 5700 versus 12.6M per mm² for the GTX 780.

Clock speeds are substantially higher on the RX 5700: base clock of 1465 MHz versus 863 MHz, boost clock of 1725 MHz versus 902 MHz, and a game clock of 1625 MHz on the RX 5700 only. Memory clocks also differ: the RX 5700 runs at 1750 MHz (14 Gbps effective), while the GTX 780 runs at 1502 MHz (6 Gbps effective). Memory capacity is 8 GB of GDDR6 on the RX 5700 versus 3 GB of GDDR5 on the GTX 780. Bus width is 256-bit on the RX 5700 and 384-bit on the GTX 780, but bandwidth favors the RX 5700 at 448.0 GB/s versus 288.4 GB/s.

Compute resources: both have 2304 shading units, but the RX 5700 has 144 TMUs and 64 ROPs, while the GTX 780 has 192 TMUs and 48 ROPs. Pixel rate is 110.4 GPixel/s for the RX 5700 versus 43.30 GPixel/s for the GTX 780. Texture rate is 248.4 GTexel/s versus 173.2 GTexel/s. FP32 performance is 7.949 TFLOPS versus 4.156 TFLOPS. The RX 5700 has FP16 at 15.90 TFLOPS (2:1), while the GTX 780 has no recorded FP16.

Power and physical dimensions: TDP is 180 W for the RX 5700 versus 250 W for the GTX 780. Suggested PSU is 450 W versus 600 W. Both are dual-slot cards with 1x 6-pin + 1x 8-pin power connectors. The RX 5700 is 268 mm long, 111 mm high, and 36 mm wide. The GTX 780 is 267 mm long, 111 mm high, and 38 mm wide. The RX 5700 uses PCIe 4.0 x16, while the GTX 780 uses PCIe 3.0 x16. Display outputs: the RX 5700 has 1x HDMI 2.0b and 3x DisplayPort 1.4a, while the GTX 780 has 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. API support: the RX 5700 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX 780 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The RX 5700 was released on 2019-07-06, while the GTX 780 was released on 2013-05-22. The RX 5700's predecessor is Vega and successor is Navi II. The GTX 780's predecessor is GeForce 600 and successor is GeForce 900. The RX 5700 has a launch MSRP of 349 USD, while the GTX 780 has a launch MSRP of 649 USD. Both are end-of-life products.