AMD Radeon Vega 8 vs NVIDIA GeForce GTX 980 Comparison

FAQ

Q: How does the AMD Radeon Vega 8 compare to the NVIDIA GeForce GTX 980 in average benchmark score?

A: The AMD Radeon Vega 8 has an average benchmark score of 9221, while the NVIDIA GeForce GTX 980 averages 8167. The Vega 8 sits slightly above the GTX 980 in this aggregate metric, despite the GTX 980 winning every head-to-head test.

Q: Which GPU wins the Geekbench Metal test?

A: The NVIDIA GeForce GTX 980 wins Geekbench Metal with a score of 15163 versus 10706 for the AMD Radeon Vega 8, a delta of -29.4% for the Vega 8.

Q: What is the largest performance gap in the head-to-head benchmarks?

A: The largest gap is in Geekbench OpenCL, where the GTX 980 scores 34676 against 8822 for the Vega 8, a -74.6% delta. This is the most decisive victory for NVIDIA in the recorded data.

Q: What is the transistor count and die size for each GPU?

A: The AMD Radeon Vega 8 has 4,940 million transistors on a 210 mm² die (23.5M per mm²), while the NVIDIA GeForce GTX 980 has 5,200 million transistors on a 398 mm² die (13.1M per mm²).

Q: What memory configurations do these GPUs use?

A: The AMD Radeon Vega 8 uses system-shared memory with bandwidth described as "System Dependent." The NVIDIA GeForce GTX 980 uses 4 GB of GDDR5 on a 256-bit bus with 224.4 GB/s bandwidth.

Q: Which GPU supports Vulkan 1.4?

A: The NVIDIA GeForce GTX 980 supports Vulkan 1.4, while the AMD Radeon Vega 8 supports Vulkan 1.3. Both support DirectX 12 (12_1) and OpenGL 4.6.

Where Each One Wins

The recorded head-to-head data shows a clean sweep: the NVIDIA GeForce GTX 980 wins all three shared benchmarks. However, the AMD Radeon Vega 8 wins the aggregate average score comparison, 9221 versus 8167 for the GTX 980. This discrepancy comes from the fact that the two GPUs have different benchmark suites recorded in the database.

The Vega 8 has three recorded Geekbench tests (Metal, OpenCL, Vulkan) with scores of 10706, 8822, and 8134 respectively. The GTX 980 has a much larger test set, including 3DMark Steel Nomad DX12 (474), Passmark DirectX 9 (164), DirectX 10 (53), DirectX 11 (83), DirectX 12 (46), G2D (792), G3D (11095), GPU Compute (4753), and the same three Geekbench tests. The GTX 980's average is dragged down by its lower scores in the older DirectX Passmark tests, while its Geekbench scores are far higher.

In terms of percentile ranking, the Vega 8 sits at the 45th percentile among all GPUs, while the GTX 980 sits at the 43rd percentile. This confirms the average score ordering, but the head-to-head tests tell a different story about raw compute capability.

Architecture Differences

The AMD Radeon Vega 8 is built on GCN 5.0 architecture using the Raven chip, fabricated on a 14 nm process at GlobalFoundries. Its transistor density is 23.5 million per mm², which is notably higher than the NVIDIA part, reflecting the more modern process node.

The NVIDIA GeForce GTX 980 uses Maxwell 2.0 architecture with the GM204 chip, fabricated on a 28 nm process at TSMC. Its transistor density is 13.1 million per mm², significantly lower than the Vega 8, but its total transistor count is higher at 5,200 million versus 4,940 million.

The Vega 8 is an integrated GPU (IGP) with no power connectors and a 25 W TDP. The GTX 980 is a dual-slot discrete card with 2x 6-pin power connectors and a 165 W TDP. The GTX 980's suggested PSU is 450 W, while the Vega 8 draws power from the motherboard.

In terms of feature support, the GTX 980 has a newer Vulkan version (1.4 versus 1.3), while both support DirectX 12 (12_1) and OpenGL 4.6. The Vega 8 has no dedicated FP16 field recorded, but its FP32 output is 1,126.4 GFLOPS. The GTX 980 has FP32 of 4.981 TFLOPS, roughly 4.4 times higher.

Specification Differences

The shading unit count differs dramatically: the GTX 980 has 2048 shading units versus 512 for the Vega 8, a 4x difference. Texture mapping units are 128 versus 32, and ROPs are 64 versus 8. These are the core compute resource differences.

Clock speeds also differ. The Vega 8 has a base clock of 300 MHz and a boost clock of 1100 MHz. The GTX 980 has a base clock of 1127 MHz and a boost clock of 1216 MHz. The GTX 980's memory runs at 1753 MHz (7 Gbps effective), while the Vega 8 uses system-shared memory.

Pixel rate for the GTX 980 is 77.82 GPixel/s versus 8.800 GPixel/s for the Vega 8, an 8.8x difference. Texture rate is 155.6 GTexel/s versus 35.20 GTexel/s, a 4.4x difference. The GTX 980's memory bandwidth is 224.4 GB/s, while the Vega 8's bandwidth is "System Dependent."

Physical dimensions differ substantially: the GTX 980 is 267 mm long, 111 mm tall, and 40 mm wide. The Vega 8 is an IGP with no recorded dimensions. The GTX 980 uses a PCIe 3.0 x16 interface, while the Vega 8 uses an IGP bus interface. Display outputs are 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2 for the GTX 980, versus "Motherboard Dependent" for the Vega 8.

Head-to-Head Benchmarks

The Geekbench Metal test shows the GTX 980 at 15163 versus the Vega 8 at 10706, a delta of -29.4% for the AMD part. This is the closest of the three head-to-head results, but still a clear NVIDIA victory. The score gap of 4457 points suggests the GTX 980 has a meaningful advantage in Metal-accelerated workloads.

Geekbench OpenCL is the most lopsided result. The GTX 980 scores 34676, while the Vega 8 scores 8822, a delta of -74.6%. This 25854-point gap reflects the massive difference in shading units (2048 versus 512) and FP32 throughput (4.981 TFLOPS versus 1,126.4 GFLOPS). OpenCL compute tasks appear to scale strongly with these resources.

Geekbench Vulkan shows the GTX 980 at 22543 versus 8134 for the Vega 8, a delta of -63.9%. The 14409-point gap confirms that the GTX 980's advantage extends to modern graphics APIs as well. The Vega 8's support for Vulkan 1.3 does not offset its lower compute resources.

The wins tally is 0 for the AMD Radeon Vega 8 and 3 for the NVIDIA GeForce GTX 980. No benchmark in the shared set favors the Vega 8. However, the average score comparison flips this narrative: the Vega 8's 9221 average exceeds the GTX 980's 8167 average. This is because the GTX 980's Passmark DirectX 9, 10, 11, and 12 scores (164, 53, 83, and 46 respectively) are far below the Geekbench scores, pulling its average down.

The Verdict

The data supports a clear performance hierarchy in direct comparison: the NVIDIA GeForce GTX 980 dominates the AMD Radeon Vega 8 in every shared benchmark. The OpenCL gap of -74.6% and Vulkan gap of -63.9% show that the GTX 980's higher shading unit count, texture units, and memory bandwidth translate into substantial real-world compute advantages.

However, the average benchmark score tells a different story. The Vega 8's 9221 average versus the GTX 980's 8167 average means that, across the full recorded test suite, the AMD part comes out ahead. This is largely due to the GTX 980's poor showing in legacy DirectX Passmark tests, which are not part of the Vega 8's recorded benchmark set.

For users prioritizing modern compute workloads (OpenCL, Vulkan, Metal), the GTX 980 is the clear choice based on the head-to-head data. For users looking at the aggregate database ranking, the Vega 8 sits one percentile higher (45th versus 43rd) and has a higher average score. The GTX 980's nearest rivals include the GeForce GTX 950M (0.4% lower score), while the Vega 8's nearest rivals include the AMD Radeon 890M (0.1% higher score).

The GTX 980 offers 4 GB of dedicated GDDR5 memory with 224.4 GB/s bandwidth, a 256-bit bus, and 64 ROPs. The Vega 8 relies on system-shared memory with motherboard-dependent bandwidth and only 8 ROPs. For gaming at higher resolutions or compute tasks with large data sets, the GTX 980's memory subsystem provides a structural advantage that the benchmark scores confirm.

The Vega 8's 25 W TDP versus the GTX 980's 165 W TDP means the AMD part is suited for low-power integrated systems, while the GTX 980 requires a discrete power delivery setup with 2x 6-pin connectors and a 450 W suggested PSU. The Vega 8's 14 nm process and higher transistor density (23.5M per mm² versus 13.1M per mm²) show a more modern fabrication approach, but this does not translate into compute superiority in the recorded tests.

The verdict from the data: pick the GTX 980 for maximum compute performance in OpenCL, Vulkan, and Metal workloads. Pick the Vega 8 for an integrated solution with a higher aggregate benchmark average and dramatically lower power draw. The GTX 980's launch MSRP was 549 USD.