AMD Radeon Pro 570 vs NVIDIA GeForce GTX 980 Ti Comparison

# AMD Radeon Pro 570 vs NVIDIA GeForce GTX 980 Ti

The Radeon Pro 570 and the GeForce GTX 980 Ti represent two very different design philosophies from the same era. The AMD part is a professional mobile workstation GPU, while the NVIDIA card is a desktop enthusiast-class part. Benchmark data places them within 1.1% of each other in average score, making direct comparisons meaningful. This analysis is based on recorded database results, covering seven distinct benchmark tests.

Where Each One Wins - use-case split based on benchmark wins

The recorded benchmarks split exactly along architectural lines. The AMD Radeon Pro 570 wins one out of three direct head-to-head comparisons. The NVIDIA GeForce GTX 980 Ti wins the remaining two.

The AMD card’s single win comes in the Geekbench Metal test, a graphics and compute API test that is heavily optimized for AMD’s GCN architecture. In this specific test, the Radeon Pro 570 scores 39,945, while the GTX 980 Ti scores 19,520. That is a 104.6% advantage for the AMD part, its largest margin of victory in any test.

The NVIDIA card dominates in the other two recorded tests. In the Geekbench OpenCL test, which is a more general compute workload, the GTX 980 Ti scores 43,513 versus 27,702 for the Radeon Pro 570, a 36.3% difference. In the Geekbench Vulkan test, the NVIDIA part scores 47,724 against 31,974, a 33% advantage. Vulkan is a cross-platform graphics API, and here the Maxwell-based NVIDIA card shows a clear throughput advantage.

The wins are defined by the underlying architecture. AMD’s card leads in a single Apple-specific API, while NVIDIA’s card leads in two cross-platform compute standards. The use case is therefore simple: for macOS-specific compute, the Radeon Pro 570 is better; for general compute or Vulkan workloads, the GTX 980 Ti is better.

Architecture Differences

The two cards come from different fabs and design nodes. The AMD Radeon Pro 570 is built on the Ellesmere chip, using the GCN 4.0 architecture. It is manufactured on a 14 nm process at GlobalFoundries, with 5,700 million transistors on a 232 mm² die. This yields a transistor density of 24.6 million per square millimeter.

The NVIDIA GeForce GTX 980 Ti uses the GM200 chip, based on the Maxwell 2.0 architecture. It is made on a 28 nm process at TSMC, packing 8,000 million transistors into a 601 mm² die. This results in a transistor density of 13.3 million per square millimeter. The AMD is the denser chip, but the NVIDIA has a larger absolute transistor count and a significantly larger die.

There are major differences in core compute resources. The Radeon Pro 570 has 1,792 shading units, 112 texture mapping units, and 32 render output units (ROPs). The GTX 980 Ti, by contrast, has 2,816 shading units, 176 TMUs, and 96 ROPs. This is a 57.1% increase in shading units, a 57.1% increase in texture mapping units, and a 200% increase in ROP count. The NVIDIA card has a higher absolute count of every key fixed-function resource.

Memory also differs. The AMD card ships with 4 GB of GDDR5 on a 256-bit bus, providing 217.0 GB/s of bandwidth. The NVIDIA card has 6 GB of GDDR5 on a 384-bit bus, delivering 336.6 GB/s. The GTX 980 Ti therefore has a 55.1% memory capacity advantage and a similar 55.1% bandwidth advantage over the AMD.

Clock speeds are closer than the resource counts. The Radeon Pro 570 runs with a 1,000 MHz base clock and a 1,105 MHz boost. Its memory operates at 1,695 MHz. The NVIDIA card has a 1,000 MHz base and a 1,076 MHz boost. Its memory runs at 1,753 MHz. The AMD has the higher boost clock by 2.6%, but NVIDIA has the higher memory clock by 3.6%.

The result is a substantial compute gap. The AMD card is rated for 3.960 TFLOPS of FP32 compute, and NVIDIA for 6.060 TFLOPS, a 53.0% advantage. The AMD supports FP16 at a 1:1 ratio (3.960 TFLOPS), while the NVIDIA does not list a FP16 rate. Pixel fill and texture rate also favor the GTX 980 Ti: 103.3 GPixel/s and 189.4 GTexel/s versus 35.36 GPixel/s and 123.8 GTexel/s for the AMD. This is a 2.9x advantage in pixel throughput and a 1.5x advantage in texture rate.

The board specifications reflect their target markets. The AMD card uses the PCIe 3.0 x16 bus, while the NVIDIA uses the same. The AMD's slot width is IGP, meaning it has no dedicated power connectors and draws all power from the slot. The NVIDIA uses a dual-slot cooler and needs one 6-pin plus one 8-pin power connector. The AMD has TDP of 150 W, the NVIDIA 250 W. Both have three display outputs: the AMD has portable device dependent outputs, the NVIDIA has 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2 outputs.

Head-to-Head Benchmarks

The most decisive result in the database is the Metal test, where the AMD Radeon Pro 570 wins outright by 104.6%. That test is a compute and graphics API with a particular strength on AMD's GCN architecture. The score of 39,945 for the Radeon is almost exactly double the 19,520 of the NVIDIA. This is not a minor optimization, it is a case of an architecture that maps to a specific API.

The NVIDIA card returns the favor in two tests. In Geekbench OpenCL, the GTX 980 Ti scores 43,513 against 27,702, a 36.3% advantage. This is the broadest test, covering general compute, and the Maxwell architecture's strength shows up here.

In the Geekbench Vulkan test, the GTX 980 Ti scores 47,724 over 31,974. The 33% margin indicates that the NVIDIA card provides a more efficient Vulkan driver and/or more compute headroom for this API.

Looking at the test scores, it is clear that the 980 Ti is the faster card in absolute terms in the two OpenCL and Vulkan tests, but the AMD card is the absolute winner in the one API it was designed to accelerate.

The Verdict

The database splits the two cards sharply. The AMD Radeon Pro 570 wins only one test, but that's a decisive win in Apple's Metal API. For users on macOS where Metal is the primary API, this card offers a 104.6% advantage over the NVIDIA, a massive performance lead in the platform's native graphics and compute environment.

The NVIDIA GeForce GTX 980 Ti wins two of three tests, and the margins are equally decisive. The 36.3% lead in OpenCL and the 33% lead in Vulkan make it the better card for cross-platform compute, for Linux and Windows gaming, or for any workload that relies on Vulkan or general purpose OpenCL compute. Its 53% greater FP32 compute rating, 200% more ROPs, and 55.1% more memory bandwidth further back the case that it is the stronger card.

Choose the AMD Radeon Pro 570 if your primary workload is Metal-based on macOS. Choose the NVIDIA GeForce GTX 980 Ti for general compute, modern cross-platform APIs, or for any performance work where the extra shading units, ROPs, and memory bandwidth are more important.

FAQ

Q: Which card has more shading units?

A: The NVIDIA GeForce GTX 980 Ti has 2,816 shading units, versus 1,792 for the AMD Radeon Pro 570. That's 57.1% more shading units.

Q: Which card has higher clock speed?

A: The AMD Radeon Pro 570 has a higher boost clock of 1,105 MHz, versus 1,076 MHz for the NVIDIA GeForce GTX 980 Ti, a 2.7% advantage.

Q: What are the memory sizes of both cards?

A: The AMD Radeon Pro 570 has 4 GB of GDDR5 memory. The NVIDIA GeForce GTX 980 Ti has 6 GB of GDDR5 memory. The NVIDIA has 50% more memory capacity.

Q: Which card is better in the Metal API?

A: The AMD Radeon Pro 570 wins the Geekbench Metal test by a 104.7% margin. It scored 39,945 versus 19,520.

Q: What is the transistor density of the two cards?

A: The AMD Radeon Pro 570 has a transistor density of 24.6 million per mm², as its die is 232 mm² and has 5,700 million transistors. The NVIDIA GeForce GTX 980 Ti has a density of 13.3 million per mm², with 8,000 million transistors on a 601 mm² die.

Q: Which card has a higher texture fill rate?

A: The NVIDIA GeForce GTX 980 Ti has a texture rate of 189.4 GTexel/s, versus the AMD Radeon Pro 570's 123.8 GTexel/s. The NVIDIA is 53.1% ahead in this metric.