AMD Radeon Pro 560X vs NVIDIA GeForce GTX 1070 Ti Comparison

Head-to-Head Benchmarks

The benchmark data shows a decisive NVIDIA victory across all three comparable tests, with the GeForce GTX 1070 Ti winning every head-to-head matchup. The largest margin appears in Geekbench OpenCL, where the GTX 1070 Ti scores 49,219 against the Radeon Pro 560X's 9,663 — a delta of 80.4%. That is not a marginal gap; it represents a fundamental throughput difference between the two cards, one that persists across every API tested.

Vulkan results tell a similar story. The GTX 1070 Ti posts 61,006 in Geekbench Vulkan, while the Radeon Pro 560X manages 16,819. The delta here is 72.4%, meaning NVIDIA's card delivers roughly three and a half times the raw Vulkan performance. Metal is the closest contest of the three, yet even there the GTX 1070 Ti leads with 22,180 versus 18,763 — a 15.4% advantage. For a card designed primarily for macOS-driven Metal workloads, losing by that margin in its home API is notable.

The average benchmark scores reinforce the same hierarchy, though with less drama. The Radeon Pro 560X averages 15,082 across its benchmark suite, while the GTX 1070 Ti averages 14,277. That puts the AMD card slightly ahead in aggregate scoring — but the head-to-head tests are more specific and more telling. The discrepancy arises because the GTX 1070 Ti's benchmark list includes several PassMark sub-tests (DirectX 9 at 207, DirectX 10 at 90, DirectX 11 at 107, DirectX 12 at 51) that drag its average down, while the Radeon's shorter list of three Geekbench tests keeps its average higher. When the same test runs on both cards, NVIDIA wins every time.

Percentile rankings place both cards in the middle of the GPU distribution: the Radeon Pro 560X sits at the 57th percentile, the GTX 1070 Ti at the 56th. This indicates that while the GTX 1070 Ti dominates this specific head-to-head, neither card is a top-tier performer by current standards. The nearest rivals for the Radeon Pro 560X include the NVIDIA GeForce GTX 660 Ti (average 15,063, 0.1% behind), the AMD Radeon RX 7600 (15,171, 0.6% ahead), and the NVIDIA GeForce RTX 3050 OEM (15,199, 0.8% ahead). For the GTX 1070 Ti, the nearest rivals are the Intel Iris Xe MAX Graphics (14,315, 0.3% ahead), the AMD Radeon Vega 11 (14,352, 0.5% ahead), and the NVIDIA GeForce GTX TITAN (14,373, 0.7% ahead). These rival deltas are all under 1%, suggesting that both cards sit in crowded performance neighborhoods.

Architecture Differences

The two GPUs come from different foundries and process nodes. AMD's Radeon Pro 560X uses a Polaris 21 chip built on GlobalFoundries' 14 nm process, containing 3,000 million transistors on a 123 mm² die. NVIDIA's GeForce GTX 1070 Ti uses a GP104 chip fabricated by TSMC at 16 nm, with 7,200 million transistors spread across a 314 mm² die. The transistor density is nearly identical — 24.4 million per mm² for AMD versus 22.9 million per mm² for NVIDIA — but the absolute scale differs massively. The GTX 1070 Ti packs more than twice the transistors into more than twice the die area.

Architecturally, the Radeon Pro 560X runs GCN 4.0, while the GTX 1070 Ti runs Pascal. These are fundamentally different designs. GCN 4.0 emphasizes compute uniformity with a 1:1 FP16 to FP32 ratio — both sit at 2.056 TFLOPS. Pascal, by contrast, heavily deprioritizes FP16 at 127.9 GFLOPS, a 1:64 ratio. This makes the Radeon Pro 560X far more capable in half-precision compute tasks per unit of throughput, though the GTX 1070 Ti's raw FP32 power overwhelms that advantage in practice: 8.186 TFLOPS versus 2.056 TFLOPS.

The memory subsystems diverge sharply. The Radeon Pro 560X offers 4 GB of GDDR5 on a 128-bit bus, yielding 94.08 GB/s of bandwidth. The GTX 1070 Ti provides 8 GB of GDDR5 on a 256-bit bus, delivering 256.3 GB/s — roughly 2.7 times the bandwidth. Clock speeds reflect this too: the Radeon's memory runs at 1470 MHz (5.9 Gbps effective), while the GTX 1070 Ti's memory runs at 2002 MHz (8 Gbps effective). The NVIDIA card also has explicit base and boost clocks of 1607 MHz and 1683 MHz, whereas the Radeon Pro 560X lists no base or boost clock in the data.

Shader resources are similarly lopsided. The GTX 1070 Ti has 2,432 shading units, 152 texture mapping units, and 64 ROPs. The Radeon Pro 560X has 1,024 shading units, 64 TMUs, and 16 ROPs. This translates to pixel rates of 107.7 GPixel/s and texture rates of 255.8 GTexel/s for NVIDIA, versus 16.06 GPixel/s and 64.26 GTexel/s for AMD. The API support also differs: NVIDIA lists DirectX 12 (12_1) and Vulkan 1.4, while AMD lists DirectX 12 (12_0) and Vulkan 1.3. Both support OpenGL 4.6.

The Verdict

The data is unambiguous. The GeForce GTX 1070 Ti wins all three head-to-head benchmarks, with margins ranging from 15.4% in Metal to 80.4% in OpenCL. Any workload that can use the GTX 1070 Ti's resources will see a substantial performance advantage. The Radeon Pro 560X does not win a single comparative test, and its only claim to superiority lies in the aggregate average benchmark score, which is a statistical artifact of different test lists rather than a genuine performance edge.

Who should pick the Radeon Pro 560X? Strictly from the data, only someone constrained by form factor or power envelope. The Radeon Pro 560X is an integrated-class part (IGP slot width) with no power connectors and a 75 W TDP, versus the GTX 1070 Ti's dual-slot design, single 8-pin connector, and 180 W TDP. If the system cannot physically accommodate a dual-slot card with external power, the Radeon is the only choice. The data also shows the Radeon's 1:1 FP16 ratio, which could matter for specific compute workloads that emphasize half-precision — though its FP32 is only a quarter of NVIDIA's.

Who should pick the GTX 1070 Ti? Anyone who needs performance. It offers 8 GB of VRAM versus 4 GB, nearly three times the memory bandwidth, four times the FP32 throughput, and dominant scores across every API tested. The Geekbench OpenCL score of 49,219 versus 9,663 is the most striking number in the comparison — it suggests the GTX 1070 Ti is in a completely different performance class, despite the similar percentile rankings. The GTX 1070 Ti also has a documented launch MSRP of 399 USD, which can be stated once for reference.

Specification Differences

The two cards differ on nearly every specification field. Process node: 14 nm (AMD) versus 16 nm (NVIDIA). Transistor count: 3,000 million versus 7,200 million. Die size: 123 mm² versus 314 mm². Memory: 4 GB GDDR5 on a 128-bit bus versus 8 GB GDDR5 on a 256-bit bus. Bandwidth: 94.08 GB/s versus 256.3 GB/s. Shading units: 1,024 versus 2,432. TMUs: 64 versus 152. ROPs: 16 versus 64. Pixel rate: 16.06 GPixel/s versus 107.7 GPixel/s. Texture rate: 64.26 GTexel/s versus 255.8 GTexel/s. FP32: 2.056 TFLOPS versus 8.186 TFLOPS. FP16: 2.056 TFLOPS (1:1) versus 127.9 GFLOPS (1:64). TDP: 75 W versus 180 W. Slot width: IGP versus dual-slot. Power connectors: none versus 1x 8-pin. Bus interface: PCIe 3.0 x8 versus PCIe 3.0 x16. Display outputs: portable device dependent versus 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a. The GTX 1070 Ti also has physical dimensions (267 mm length, 112 mm height, 40 mm width) and a suggested PSU of 450 W, while the Radeon Pro 560X lists none.

FAQ

Q: Which card wins in Geekbench Metal?

A: The NVIDIA GeForce GTX 1070 Ti wins with a score of 22,180, a 15.4% advantage over the AMD Radeon Pro 560X's 18,763.

Q: How large is the OpenCL performance gap?

A: The GTX 1070 Ti scores 49,219 in Geekbench OpenCL versus 9,663 for the Radeon Pro 560X, a delta of 80.4% in NVIDIA's favor.

Q: Does the Radeon Pro 560X have any benchmark win?

A: No. In the three head-to-head tests (Metal, OpenCL, Vulkan), the Radeon Pro 560X wins zero and the GTX 1070 Ti wins all three.

Q: What is the memory bandwidth difference?

A: The GTX 1070 Ti has 256.3 GB/s of bandwidth, while the Radeon Pro 560X has 94.08 GB/s — a difference of roughly 2.7 times.

Q: Do both cards support the same DirectX version?

A: No. The GTX 1070 Ti supports DirectX 12 (12_1), while the Radeon Pro 560X supports DirectX 12 (12_0). Vulkan support also differs: 1.4 for NVIDIA, 1.3 for AMD.

Q: What is the FP16 capability of each card?

A: The Radeon Pro 560X delivers 2.056 TFLOPS FP16 at a 1:1 ratio with FP32. The GTX 1070 Ti delivers 127.9 GFLOPS FP16 at a 1:64 ratio, making the AMD card substantially better at half-precision compute per unit of throughput.

Where Each One Wins

The GTX 1070 Ti wins everywhere that raw rasterization or compute throughput matters. Its 8.186 TFLOPS FP32, 107.7 GPixel/s pixel rate, and 255.8 GTexel/s texture rate place it far above the Radeon Pro 560X's 2.056 TFLOPS, 16.06 GPixel/s, and 64.26 GTexel/s. For gaming, 3D rendering, video processing, or any general-purpose GPU workload, the GTX 1070 Ti is the superior choice by every measured metric. The Vulkan score of 61,006 versus 16,819 and the OpenCL score of 49,219 versus 9,663 leave no ambiguity.

The Radeon Pro 560X has narrower but real niches. Its 1:1 FP16 ratio makes it suitable for workloads that operate primarily in half precision, where its FP16 throughput equals its FP32 throughput — a feature the GTX 1070 Ti lacks with its 1:64 ratio. Its 75 W TDP and absence of power connectors mean it can function in systems with minimal power delivery, and its IGP slot width allows installation in chassis that cannot fit a dual-slot card. The Radeon's PCIe 3.0 x8 interface versus the GTX 1070 Ti's x16 is a further indicator of its lower bandwidth footprint.

For users who need Metal performance on a Mac-compatible part, the Radeon Pro 560X's 18,763 score is competitive in absolute terms — it sits near the 57th percentile of all GPUs — but it is still 15.4% behind the GTX 1070 Ti even in that favorable test. The data does not support any scenario where the Radeon Pro 560X outperforms the GTX 1070 Ti in a directly comparable workload. The only reasons to choose the AMD card are physical constraints (power, space, bus width) or the specific need for balanced FP16 compute. For everything else, the GTX 1070 Ti is the data-backed pick.