AMD Radeon RX 5700 vs NVIDIA GeForce RTX 3080 Comparison

FAQ

Q: Which GPU wins the majority of benchmark comparisons?

A: The NVIDIA GeForce RTX 3080 wins 9 out of 10 head-to-head benchmark comparisons against the AMD Radeon RX 5700. The only test where AMD comes out ahead is Geekbench Vulkan, where the RX 5700 scores 64,166 versus the RTX 3080's 33,620 — a 47.6% advantage for AMD.

Q: How large is the performance gap in the most demanding modern test?

A: In the 3DMark Steel Nomad DX12 test, the RTX 3080 scores 4,407 points against the RX 5700's 1,862 points. That represents a 136.7% delta — the RTX 3080 more than doubles the RX 5700's result in this particular workload.

Q: What do the average benchmark scores indicate about overall performance?

A: The RTX 3080 has an average benchmark score of 23,172, while the RX 5700 averages 22,170. The RTX 3080 sits at the 68th percentile of all GPUs, and the RX 5700 sits at the 67th percentile — placing both cards in a similar overall tier despite the large deltas in individual tests.

Q: Which GPU has a higher transistor count and what is the difference?

A: The RTX 3080 uses 28,300 million transistors on a 628 mm² die, while the RX 5700 has 10,300 million transistors on a 251 mm² die. The RTX 3080's transistor density is 45.1M per mm², compared to the RX 5700's 41.0M per mm².

Q: Do both cards support the same DirectX version?

A: No. The RTX 3080 supports DirectX 12 Ultimate (12_2), while the RX 5700 supports DirectX 12 (12_1). Both cards support OpenGL 4.6 and Vulkan 1.4.

Q: What is the difference in memory bandwidth between the two cards?

A: The RTX 3080 delivers 760.3 GB/s of memory bandwidth from its 10 GB of GDDR6X on a 320-bit bus. The RX 5700 provides 448.0 GB/s from 8 GB of GDDR6 on a 256-bit bus.

The Verdict

The data points to a clear overall winner: the NVIDIA GeForce RTX 3080. It dominates in 9 of 10 head-to-head benchmarks, with deltas ranging from 20.2% in Passmark G2D to 136.7% in 3DMark Steel Nomad DX12. The RTX 3080's average score of 23,172 versus 22,170 for the RX 5700 puts it ahead, and its 68th percentile ranking versus 67th confirms the edge.

However, the verdict is not a blanket recommendation for every scenario. The RX 5700's single win in Geekbench Vulkan by 47.6% is substantial. For users whose primary workload relies heavily on Vulkan compute or rendering, the RX 5700 could be the more appropriate choice despite losing most other comparisons. The RTX 3080 also carries a 1x 12-pin power connector and a suggested PSU of 700 W, versus the RX 5700's 1x 6-pin + 1x 8-pin and 450 W suggested PSU — a meaningful system-level difference.

The RTX 3080 is the pick for users prioritizing raw rasterization, DX12 performance, or compute throughput. The RX 5700 is the pick for users on a tighter power budget or those whose software stack is heavily optimized for Vulkan. Neither card is a bad choice; the data simply shows they excel in different domains.

Head-to-Head Benchmarks

The largest victory for the RTX 3080 comes in 3DMark Steel Nomad DX12, where its score of 4,407 versus 1,862 yields a 136.7% delta. This is not a marginal win — the RTX 3080 achieves more than double the RX 5700's result. Similarly, in Geekbench OpenCL, the RTX 3080 posts 152,423 against 69,603, a 119% advantage. Passmark GPU Compute shows 14,397 versus 6,517, a 120.9% delta, indicating the RTX 3080 is far stronger in general-purpose compute workloads.

In the Passmark DirectX suite, the RTX 3080 wins every tier. The DX11 test shows 207 versus 101 (105% delta), DX10 shows 170 versus 87 (95.4% delta), and DX12 shows 100 versus 54 (85.2% delta). Even the legacy DX9 test results in a 23.4% win for the RTX 3080, with scores of 258 versus 209. The Passmark G3D score of 25,086 versus 14,314 represents a 75.3% lead, while G2D shows a more modest 20.2% delta (1,054 versus 877).

The single counterpoint is Geekbench Vulkan. The RX 5700 scores 64,166, while the RTX 3080 manages only 33,620. This 47.6% delta in AMD's favor is the only test where the RX 5700 leads, and it is a substantial one. It suggests that in Vulkan-specific scenarios, the RX 5700's architecture holds a distinct advantage that is not reflected in other workloads.

The overall pattern is consistent: the RTX 3080 wins by large margins in compute-heavy and DX12 tests, while the RX 5700 shows a single but notable strength in Vulkan. The average scores — 23,172 for the RTX 3080 and 22,170 for the RX 5700 — reflect this distribution, with the RTX 3080's average being pulled up by its extreme results in several tests.

Specification Differences

The two cards differ in nearly every major specification category. The RTX 3080 has 8,704 shading units, 272 TMUs, and 96 ROPs, compared to the RX 5700's 2,304 shading units, 144 TMUs, and 64 ROPs. The RTX 3080 also features 68 RT cores and 272 tensor cores, while the RX 5700 has neither — it lists null values for both.

Clock speeds are similar at the base and boost levels: the RTX 3080 runs at 1440 MHz base and 1710 MHz boost, while the RX 5700 runs at 1465 MHz base and 1725 MHz boost. The RX 5700 has a game clock of 1625 MHz, which the RTX 3080 does not list. Memory clocks differ significantly: the RTX 3080's memory runs at 1188 MHz (19 Gbps effective), while the RX 5700's runs at 1750 MHz (14 Gbps effective).

The process nodes differ: the RTX 3080 uses an 8 nm process from Samsung, while the RX 5700 uses a 7 nm process from TSMC. The RTX 3080's die is 628 mm² versus 251 mm² for the RX 5700, and transistor counts are 28,300 million versus 10,300 million. The RTX 3080 has a TDP of 320 W compared to 180 W for the RX 5700, and the suggested PSU is 700 W versus 450 W.

Physical dimensions also differ. The RTX 3080 is 285 mm long, 112 mm tall, and 40 mm wide; the RX 5700 is 268 mm long, 111 mm tall, and 36 mm wide. Both are dual-slot cards. Power connectors are 1x 12-pin for the RTX 3080 and 1x 6-pin + 1x 8-pin for the RX 5700. Display outputs are similar — both have 3x DisplayPort 1.4a, but the RTX 3080 has HDMI 2.1 while the RX 5700 has HDMI 2.0b. Both use PCIe 4.0 x16.

Architecture Differences

The RTX 3080 is built on NVIDIA's Ampere architecture with the GA102 chip, which is a significant generational leap from the preceding GeForce 20 series. The RX 5700 uses AMD's RDNA 1.0 architecture with the Navi 10 chip, succeeding the Vega architecture. The RTX 3080's support for DirectX 12 Ultimate (12_2) versus the RX 5700's DirectX 12 (12_1) reflects a newer feature set.

The RTX 3080's FP32 throughput is 29.77 TFLOPS, and its FP16 performance is identical at 29.77 TFLOPS with a 1:1 ratio. The RX 5700 delivers 7.949 TFLOPS FP32 and 15.90 TFLOPS FP16 with a 2:1 ratio. This means the RTX 3080 has nearly four times the FP32 throughput and about 1.87 times the FP16 throughput of the RX 5700.

The pixel rate and texture rate follow the same pattern. The RTX 3080 achieves 164.2 GPixel/s and 465.1 GTexel/s, while the RX 5700 achieves 110.4 GPixel/s and 248.4 GTexel/s. The RTX 3080's pixel rate is 48.7% higher, and its texture rate is 87.2% higher. The RTX 3080's memory architecture — GDDR6X on a 320-bit bus — provides 760.3 GB/s versus the RX 5700's GDDR6 on a 256-bit bus at 448.0 GB/s.

The presence of RT cores and tensor cores is the most fundamental architectural difference. The RTX 3080's 68 RT cores enable hardware-accelerated ray tracing, and its 272 tensor cores accelerate AI workloads. The RX 5700 has no such dedicated hardware. This is a categorical difference, not just a numerical one — the RTX 3080 can offload ray tracing and tensor operations to specialized silicon, while the RX 5700 must handle these tasks through its general-purpose shading units.

The release dates also differ: the RTX 3080 was released on 2020-08-31, while the RX 5700 was released on 2019-07-06. Both are end-of-life products. The RTX 3080's predecessor is the GeForce 20 series and its successor is the GeForce 40 series; the RX 5700's predecessor is Vega and its successor is Navi II.