AMD Radeon RX 6800M vs NVIDIA GeForce GTX 980 Ti Comparison

# AMD Radeon RX 6800M vs NVIDIA GeForce GTX 980 Ti

The AMD Radeon RX 6800M is a mobile-first GPU built on RDNA 2.0, while the NVIDIA GeForce GTX 980 Ti is a desktop flagship from the Maxwell 2.0 era. The benchmark data shows a decisive sweep: the RX 6800M wins all four head-to-head tests, with deltas ranging from 69.4% to 482.6%. Despite the GTX 980 Ti's higher raw shading unit count (2816 vs 2560), the RX 6800M's architectural advantages and memory configuration deliver consistently superior results across every measured workload.

Where Each One Wins

The RX 6800M wins every benchmark in the head-to-head comparison, but the magnitude of its advantage varies dramatically by workload type. In the 3DMark Steel Nomad DX12 test, the RX 6800M scores 2238 against the GTX 980 Ti's 1321, a 69.4% lead. This is its narrowest victory, suggesting that in traditional rasterization-heavy DX12 workloads, the older NVIDIA card remains relatively competitive despite its age.

The RX 6800M's dominance becomes overwhelming in compute-oriented tests. In Geekbench OpenCL, it scores 87621 versus 43513, more than doubling the GTX 980 Ti's output with a 101.4% delta. The Vulkan test shows a similar pattern: 94766 versus 47724, a 98.6% advantage. These results indicate that RDNA 2.0's compute architecture is far more efficient per clock than Maxwell 2.0, particularly in API-level parallel workloads.

The most extreme gap appears in Geekbench Metal, where the RX 6800M scores 113721 against the GTX 980 Ti's 19520 — a staggering 482.6% difference. This is partly explained by the GTX 980 Ti's age and the fact that Metal is a modern API optimized for newer hardware, but the sheer scale of the gap underscores how far GPU compute has advanced since 2015. The GTX 980 Ti has no wins in any category, making this a one-sided comparison where the RX 6800M is the clear choice for any modern workload.

Architecture Differences

The two GPUs represent fundamentally different design philosophies separated by six years of silicon evolution. The RX 6800M uses TSMC's 7 nm process with 17,200 million transistors on a 335 mm² die, achieving a transistor density of 51.3M per mm². The GTX 980 Ti relies on TSMC's 28 nm process with just 8,000 million transistors on a much larger 601 mm² die, yielding only 13.3M transistors per mm². This density advantage is the primary driver of the RX 6800M's superior performance-per-watt and overall throughput.

Memory configuration further separates the two. The RX 6800M features 12 GB of GDDR6 on a 192-bit bus, delivering 384.0 GB/s of bandwidth. The GTX 980 Ti has 6 GB of GDDR5 on a wider 384-bit bus, but its bandwidth is only 336.6 GB/s due to slower memory clocks. The RX 6800M's memory runs at 2000 MHz (16 Gbps effective), while the GTX 980 Ti's memory operates at 1753 MHz (7 Gbps effective). This means the newer card offers double the capacity and 14% more bandwidth despite a narrower bus.

Feature support diverges sharply. The RX 6800M includes 40 dedicated ray tracing cores and supports DirectX 12 Ultimate (12_2), while the GTX 980 Ti has no ray tracing hardware and is limited to DirectX 12 (12_1). Both cards support OpenGL 4.6 and Vulkan 1.4, but the RX 6800M's RDNA 2.0 architecture enables hardware-accelerated ray tracing and variable rate shading, features entirely absent from Maxwell 2.0. The RX 6800M also delivers 12.24 TFLOPS of FP32 compute versus the GTX 980 Ti's 6.060 TFLOPS, and supports FP16 at 24.47 TFLOPS (2:1 ratio), which the GTX 980 Ti lacks entirely.

The GTX 980 Ti is a dual-slot desktop card requiring 1x 6-pin and 1x 8-pin power connectors with a 250 W TDP and 600 W suggested PSU. The RX 6800M is an IGP (integrated graphics processor) with no power connectors and a 145 W TDP, reflecting its mobile design intent. The GTX 980 Ti measures 267 mm in length, while the RX 6800M's dimensions are listed as portable device dependent. The GTX 980 Ti uses PCIe 3.0 x16, while the RX 6800M uses PCIe 4.0 x16, doubling the available bandwidth for data transfer.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test provides the most balanced comparison. The RX 6800M's 2238 score beats the GTX 980 Ti's 1321 by 69.4%. This test represents a modern DX12 rasterization workload, and while the RX 6800M wins clearly, the delta is smaller than in compute tests, indicating that Maxwell 2.0's traditional rendering pipeline retains some relevance in pure graphics tasks.

Geekbench OpenCL shows a 101.4% advantage for the RX 6800M, with scores of 87621 versus 43513. This doubling of compute performance aligns with the FP32 throughput gap: the RX 6800M's 12.24 TFLOPS is roughly double the GTX 980 Ti's 6.060 TFLOPS. The OpenCL benchmark exercises general-purpose compute, where RDNA 2.0's architecture excels due to its higher clock speeds (2116 MHz base, 2390 MHz boost versus 1000 MHz base, 1076 MHz boost) and superior instruction efficiency.

Geekbench Vulkan produces a 98.6% delta, with the RX 6800M scoring 94766 against 47724. Vulkan's low-overhead design benefits both cards, but the RX 6800M's architectural improvements — including async compute and better shader occupancy — yield nearly double the performance. The RX 6800M's game clock of 2300 MHz further amplifies this advantage in sustained workloads.

The Geekbench Metal test is the outlier in magnitude. The RX 6800M's 113721 score dwarfs the GTX 980 Ti's 19520, a 482.6% difference. Metal is Apple's graphics API, and its performance characteristics favor newer hardware with better driver optimization and modern feature sets. The GTX 980 Ti's Maxwell architecture, released in 2015, was not designed with Metal in mind, and its 28 nm process limits its ability to execute the parallel workloads Metal prioritizes. This result should be interpreted cautiously, as Metal performance may not directly translate to gaming or other real-world scenarios, but it does indicate the RX 6800M's massive compute headroom.

FAQ

Q: Which GPU has higher raw FP32 compute performance?

A: The AMD Radeon RX 6800M delivers 12.24 TFLOPS of FP32 performance, which is more than double the NVIDIA GeForce GTX 980 Ti's 6.060 TFLOPS. This directly correlates with the RX 6800M's 101.4% lead in Geekbench OpenCL and 98.6% lead in Geekbench Vulkan.

Q: Does the GTX 980 Ti support ray tracing?

A: No. The GTX 980 Ti has no ray tracing cores and is limited to DirectX 12 (12_1). The RX 6800M includes 40 ray tracing cores and supports DirectX 12 Ultimate (12_2), enabling hardware-accelerated ray tracing that the GTX 980 Ti cannot perform.

Q: How do their memory configurations compare?

A: The RX 6800M has 12 GB of GDDR6 on a 192-bit bus with 384.0 GB/s bandwidth. The GTX 980 Ti has 6 GB of GDDR5 on a 384-bit bus with 336.6 GB/s bandwidth. The RX 6800M offers twice the capacity and 14% more bandwidth despite a narrower bus.

Q: What is the power consumption difference?

A: The RX 6800M has a TDP of 145 W and requires no power connectors, while the GTX 980 Ti has a TDP of 250 W and requires 1x 6-pin and 1x 8-pin power connectors with a 600 W suggested PSU. This makes the RX 6800M significantly more power-efficient.

Q: Which GPU has better driver support for modern APIs?

A: Both cards support OpenGL 4.6 and Vulkan 1.4, but the RX 6800M supports DirectX 12 Ultimate (12_2) while the GTX 980 Ti is limited to DirectX 12 (12_1). The RX 6800M's newer architecture also enables features like variable rate shading that the GTX 980 Ti lacks.

Q: How do their average benchmark scores compare?

A: The RX 6800M has an average benchmark score of 28874, slightly higher than the GTX 980 Ti's 28020. Both occupy similar percentile positions (74th versus 73rd), indicating they are broadly comparable in overall performance, though the RX 6800M wins every direct head-to-head test.

Specification Differences

| Specification | AMD Radeon RX 6800M | NVIDIA GeForce GTX 980 Ti |

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

| Architecture | RDNA 2.0 | Maxwell 2.0 |

| Process Node | 7 nm | 28 nm |

| Transistors | 17,200 million | 8,000 million |

| Die Size | 335 mm² | 601 mm² |

| Transistor Density | 51.3M / mm² | 13.3M / mm² |

| Base Clock | 2116 MHz | 1000 MHz |

| Boost Clock | 2390 MHz | 1076 MHz |

| Game Clock | 2300 MHz | N/A |

| Memory Clock | 2000 MHz (16 Gbps effective) | 1753 MHz (7 Gbps effective) |

| Memory Size | 12 GB GDDR6 | 6 GB GDDR5 |

| Memory Bus Width | 192 bit | 384 bit |

| Memory Bandwidth | 384.0 GB/s | 336.6 GB/s |

| Shading Units | 2560 | 2816 |

| TMUs | 160 | 176 |

| ROPs | 64 | 96 |

| RT Cores | 40 | N/A |

| Pixel Rate | 153.0 GPixel/s | 103.3 GPixel/s |

| Texture Rate | 382.4 GTexel/s | 189.4 GTexel/s |

| FP32 Performance | 12.24 TFLOPS | 6.060 TFLOPS |

| FP16 Performance | 24.47 TFLOPS (2:1) | N/A |

| TDP | 145 W | 250 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 6-pin + 1x 8-pin |

| Suggested PSU | N/A | 600 W |

| Bus Interface | PCIe 4.0 x16 | PCIe 3.0 x16 |

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

| Release Date | 2021-05-30 | 2015-06-01 |

| Predecessor | Polaris Mobile | GeForce 700 |

| Successor | N/A | GeForce 10 |

| Launch MSRP | N/A | 649 USD |