AMD Radeon RX 6800S vs NVIDIA GeForce RTX 3080 Comparison

The AMD Radeon RX 6800S and NVIDIA GeForce RTX 3080 occupy opposite ends of the mobile and desktop GPU spectrum, yet the benchmark data reveals a more nuanced rivalry than their respective form factors suggest. The RTX 3080, a high-power desktop part from the GeForce 30-series, dominates the overall average score with 23,172 points against the RX 6800S’s 24,063, though the AMD part actually holds a higher aggregate mark. This inversion is explained by a single, dramatic benchmark win for AMD, while NVIDIA claims victory in the remaining nine head-to-head tests. The following analysis breaks down where each GPU excels, what their architectures dictate, and who should choose which.

Head-to-Head Benchmarks

The most startling result in this comparison is the Geekbench Vulkan test, where the AMD Radeon RX 6800S scores 81,272 against the RTX 3080’s 33,620. That is a delta of 141.7% in AMD’s favor, a margin so large it flips the entire average-score narrative. In every other head-to-head benchmark, the RTX 3080 wins outright, often by substantial margins. For instance, in 3DMark Steel Nomad DX12, the RTX 3080 scores 4,407 versus 1,689 for the RX 6800S, a 61.7% deficit for AMD. This is the single largest performance gap in the entire dataset, indicating that in modern, compute-heavy DX12 workloads, the NVIDIA part is in a different league.

The compute-oriented benchmarks reinforce this trend. In Geekbench OpenCL, the RTX 3080 posts 152,423 against 73,202 for the RX 6800S, a 52% lead. PassMark GPU Compute shows a similar story: 14,397 for NVIDIA versus 6,144 for AMD, a 57.3% delta. These results align with the raw FP32 throughput figures, where the RTX 3080 delivers 29.77 TFLOPS compared to 8.602 TFLOPS for the RX 6800S. The PassMark suite also shows consistent NVIDIA superiority across legacy APIs: DirectX 10 (170 vs 91, a 46.5% delta), DirectX 11 (207 vs 139, a 32.9% delta), DirectX 12 (100 vs 65, a 35% delta), and DirectX 9 (258 vs 200, a 22.5% delta). Even in 2D performance, the RTX 3080 leads with 1,054 in PassMark G2D against 732, a 30.6% delta.

The only other benchmark where AMD comes close is PassMark G3D, where the RTX 3080 scores 25,086 versus 15,908 — still a 36.6% deficit for AMD. Meanwhile, the RTX 3080’s nearest rivals in the database include the AMD Radeon RX 6600M (deltaPct -0.4) and the NVIDIA P106-100 (deltaPct -0.3), while the RX 6800S sits near the NVIDIA GeForce RTX 2080 SUPER (deltaPct -0.4) and the AMD Radeon RX 9070 (deltaPct 0.8). This places the two GPUs in similar percentile ranges — 68th for NVIDIA and 69th for AMD — despite the stark per-test differences.

Where Each One Wins

The data splits cleanly into two use cases. The AMD Radeon RX 6800S wins decisively in Vulkan-based workloads, as evidenced by its 141.7% advantage in Geekbench Vulkan. This suggests that applications optimized for Vulkan’s low-level API will see outsized performance from the RDNA 2.0 architecture, potentially making it a strong choice for Linux gaming or Vulkan-native engines. Beyond that single benchmark, the RX 6800S does not lead in any other test, so its utility is narrowly defined by API preference rather than raw throughput.

The NVIDIA GeForce RTX 3080, by contrast, is the clear winner for everything else. Its dominance spans DX9 through DX12, OpenCL, and compute workloads, making it the superior choice for legacy DirectX titles, modern DX12 games, and GPU-accelerated compute tasks like rendering or machine learning inference. The 52% lead in OpenCL and 57.3% lead in PassMark GPU Compute are particularly telling for productivity use cases. The RTX 3080 also wins the 2D test by 30.6%, which, while less impactful for gaming, indicates better overall desktop compositing and window management performance. In short, the RTX 3080 is the all-rounder, while the RX 6800S is a specialist with a single, albeit massive, strength.

Architecture Differences

The architectural divide between these two GPUs is foundational. The AMD Radeon RX 6800S uses the Navi 23 chip on RDNA 2.0 architecture, fabricated on TSMC’s 7 nm process with 11,060 million transistors on a 237 mm² die. The NVIDIA GeForce RTX 3080 uses the GA102 chip on Ampere architecture, built on Samsung’s 8 nm process with 28,300 million transistors on a 628 mm² die. The transistor density is nearly identical — 46.7M/mm² for AMD versus 45.1M/mm² for NVIDIA — but the RTX 3080’s physically larger die packs more than 2.5 times the transistors.

Core counts differ dramatically. The RX 6800S has 2,048 shading units, 128 TMUs, 64 ROPs, and 32 ray-tracing cores, with no tensor cores. The RTX 3080 has 8,704 shading units, 272 TMUs, 96 ROPs, 68 RT cores, and 272 tensor cores. This explains the FP32 throughput gap: 8.602 TFLOPS for AMD versus 29.77 TFLOPS for NVIDIA, with the latter also offering 1:1 FP16 performance (29.77 TFLOPS) versus AMD’s 2:1 ratio (17.20 TFLOPS). The RTX 3080’s tensor cores enable features like DLSS, which the RX 6800S lacks entirely.

Memory subsystems also diverge sharply. The RX 6800S uses 8 GB of GDDR6 on a 128-bit bus, yielding 256.0 GB/s of bandwidth. The RTX 3080 uses 10 GB of GDDR6X on a 320-bit bus, delivering 760.3 GB/s — three times the bandwidth. Clock speeds tell a different story: AMD runs higher at 1800 MHz base and 2100 MHz boost, versus NVIDIA’s 1440 MHz base and 1710 MHz boost, but the NVIDIA part’s wider bus and faster memory (19 Gbps effective versus 16 Gbps) compensate. Power and physical design reflect their intended use: the RX 6800S is an IGP with a 100 W TDP and no power connectors, while the RTX 3080 is a dual-slot card with a 320 W TDP, a single 12-pin connector, and a suggested 700 W PSU. The NVIDIA card measures 285 mm long, 112 mm tall, and 40 mm wide.

FAQ

Q: Why does the AMD Radeon RX 6800S have a higher average benchmark score than the RTX 3080 despite losing 9 of 10 head-to-head tests?

A: The RX 6800S’s average of 24,063 is boosted by its massive 141.7% win in Geekbench Vulkan (81,272 vs 33,620), which outweighs its losses in other tests. The RTX 3080’s average is 23,172, despite winning most tests, because its Vulkan score is disproportionately low.

Q: Which GPU is better for DirectX 12 gaming?

A: The RTX 3080 is significantly better. In 3DMark Steel Nomad DX12, it scores 4,407 versus 1,689 for the RX 6800S, a 61.7% delta. PassMark DirectX 12 also favors NVIDIA, with a 35% lead (100 vs 65).

Q: Does the RX 6800S have any advantage outside of Vulkan?

A: No. The data shows the RX 6800S wins only the Geekbench Vulkan test. In every other head-to-head benchmark — including DX9, DX10, DX11, DX12, OpenCL, G2D, G3D, and GPU Compute — the RTX 3080 leads by margins ranging from 22.5% to 61.7%.

Q: How do the memory configurations compare?

A: The RTX 3080 offers 10 GB of GDDR6X on a 320-bit bus with 760.3 GB/s bandwidth, while the RX 6800S has 8 GB of GDDR6 on a 128-bit bus with 256.0 GB/s. The RTX 3080’s bandwidth advantage is roughly threefold.

Q: What is the power consumption difference?

A: The RX 6800S has a 100 W TDP and is an IGP with no power connectors, making it suitable for portable devices. The RTX 3080 has a 320 W TDP, requires a 12-pin power connector, and recommends a 700 W PSU, reflecting its desktop dual-slot design.

Q: Which GPU has a higher transistor density?

A: The RX 6800S has a slightly higher density at 46.7M/mm², versus 45.1M/mm² for the RTX 3080. However, the RTX 3080’s die is nearly three times larger (628 mm² vs 237 mm²), giving it far more total transistors (28,300 million vs 11,060 million).

The Verdict

The data is unambiguous for most users: the NVIDIA GeForce RTX 3080 is the superior GPU for the vast majority of workloads. It wins 9 of 10 head-to-head benchmarks, including all DirectX variants, OpenCL, and compute tests, with margins ranging from 22.5% to 61.7%. Its 29.77 TFLOPS FP32 performance, 760.3 GB/s memory bandwidth, and 272 tensor cores make it the clear choice for high-resolution gaming, ray tracing, and GPU-accelerated productivity tasks. The launch MSRP of 699 USD reflects its position as a high-end desktop part, and its 320 W TDP demands a robust power supply.

The AMD Radeon RX 6800S, however, is not without merit. Its 141.7% Vulkan win is a genuine outlier that cannot be ignored for Vulkan-centric use cases, and its 100 W TDP and IGP form factor make it the only viable option for thin-and-light laptops. Its higher base and boost clocks (1800/2100 MHz) and slightly higher transistor density show AMD’s efficiency at the process node level. But for anyone building a desktop or prioritizing raw performance across APIs, the RTX 3080’s consistent dominance — underscored by its 68th percentile ranking versus the RX 6800S’s 69th, despite the latter’s average-score quirk — makes it the data-backed recommendation. The RX 6800S is a niche specialist; the RTX 3080 is the all-around winner.