AMD Radeon Pro 560 vs NVIDIA GeForce GTX 660 Ti Comparison

The AMD Radeon Pro 560 and NVIDIA GeForce GTX 660 Ti represent two very different eras of graphics hardware, separated by nearly five years of architectural evolution. The data shows the Radeon Pro 560 winning both recorded head-to-head benchmarks, yet the GTX 660 Ti still holds notable advantages in raw throughput metrics. This analysis examines what those numbers mean for real-world performance, where each card excels, and which user should consider which option based strictly on the recorded specifications and benchmark results.

Head-to-Head Benchmarks

The database records two direct comparisons between these GPUs: Geekbench OpenCL and Geekbench Vulkan. In OpenCL, the AMD Radeon Pro 560 scores 15504 against the GTX 660 Ti’s 15113, a delta of 2.6% in AMD’s favor. That is a modest margin, close enough that driver optimizations or workload characteristics could shift the outcome. The Vulkan result is more decisive: the Radeon Pro 560 posts 16232 versus 15012 for the NVIDIA card, an 8.1% advantage. Vulkan’s lower-level API overhead tends to favor newer architectures, and the data reflects that.

The average benchmark scores tell a similar story. The Radeon Pro 560 averages 17551 across all recorded tests, while the GTX 660 Ti averages 15063. That is a 16.5% gap in favor of AMD. The percentile rankings reinforce this: the Radeon Pro 560 sits at the 61st percentile of all GPUs, while the GTX 660 Ti sits at the 57th. Neither card is a top-tier performer, but the AMD part clearly sits above its rival in the database’s overall ranking.

However, the head-to-head table only includes two tests, and both are compute-oriented APIs. The GTX 660 Ti’s higher pixel rate (27.44 GPixel/s versus 14.51 GPixel/s) and texture rate (109.8 GTexel/s versus 58.05 GTexel/s) suggest that in older rasterization-heavy workloads, the NVIDIA card could potentially close or even reverse the gap. The benchmark data does not include DirectX or OpenGL gaming tests, so those scenarios remain speculative based on the specification differences.

Architecture Differences

The Radeon Pro 560 uses the Polaris 21 chip built on GCN 4.0 architecture, fabricated on a 14 nm process by GlobalFoundries. The GTX 660 Ti uses the GK104 chip with Kepler architecture, fabricated on a 28 nm process by TSMC. This process node difference is significant: 14 nm versus 28 nm allows the AMD chip to pack 3,000 million transistors into a 123 mm² die, yielding a transistor density of 24.4 million per square millimeter. The NVIDIA chip has more transistors overall (3,540 million) but on a much larger 294 mm² die, resulting in only 12.0 million per square millimeter. The AMD design is nearly twice as dense, which explains how it achieves competitive compute performance at a fraction of the power draw.

Memory configurations differ sharply. The Radeon Pro 560 has 4 GB of GDDR5 on a 128-bit bus, delivering 81.28 GB/s of bandwidth. The GTX 660 Ti has 2 GB of GDDR5 on a 192-bit bus, delivering 144.2 GB/s. The NVIDIA card has 77% more bandwidth, which matters for texture-heavy scenes and large framebuffers. However, the AMD card’s larger capacity (4 GB versus 2 GB) becomes more relevant in modern workloads where 2 GB is a hard constraint.

The compute resources also diverge. The Radeon Pro 560 has 1024 shading units, 64 texture mapping units, and 16 ROPs. The GTX 660 Ti has 1344 shading units, 112 TMUs, and 24 ROPs. In raw numbers, NVIDIA leads in every category. The GTX 660 Ti’s FP32 throughput is 2.634 TFLOPS versus 1.858 TFLOPS for the AMD card, a 41.8% advantage. Yet the benchmark results show AMD winning, which implies that architectural efficiency, driver maturity, or API support compensates for the raw throughput deficit. The Radeon Pro 560’s FP16 throughput matches its FP32 at 1.858 TFLOPS (1:1 ratio), whereas the GTX 660 Ti has no recorded FP16 capability.

Power and physical design are another major divide. The Radeon Pro 560 has a 75 W TDP, requires no power connectors, and is listed as an integrated graphics processor (IGP) with a slot width of IGP. The GTX 660 Ti has a 150 W TDP, needs two 6-pin power connectors, and is a dual-slot card with a 241 mm length. The AMD part is clearly designed for mobile or compact systems, while the NVIDIA card is a traditional desktop add-in board.

API support also differs. The Radeon Pro 560 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.3. The GTX 660 Ti supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The AMD card has a more recent Vulkan version and a higher DirectX feature level, which may explain its better Vulkan benchmark showing.

Where Each One Wins

The Radeon Pro 560 wins in compute-oriented benchmarks. Its OpenCL score of 15504 and Vulkan score of 16232 both exceed the GTX 660 Ti’s corresponding figures. The 8.1% Vulkan margin is particularly notable, as Vulkan is designed to reduce CPU overhead and expose more parallel work to the GPU. The AMD card’s GCN architecture, with its emphasis on asynchronous compute and explicit command queues, likely contributes to this result. The 4 GB memory capacity also gives it an advantage in workloads that exceed 2 GB, such as large datasets or high-resolution textures.

The GTX 660 Ti wins in raw throughput metrics. Its pixel rate of 27.44 GPixel/s is nearly double the Radeon Pro 560’s 14.51 GPixel/s, and its texture rate of 109.8 GTexel/s is almost double the AMD card’s 58.05 GTexel/s. These figures suggest that in fill-rate-bound scenarios, such as early DirectX 11 games or legacy OpenGL titles, the NVIDIA card could outperform its newer rival despite losing in modern API benchmarks. The GTX 660 Ti’s higher memory bandwidth (144.2 GB/s versus 81.28 GB/s) further supports this, as bandwidth is often a limiting factor in rasterization-heavy rendering.

The Radeon Pro 560’s lower TDP (75 W versus 150 W) makes it the clear winner for power-constrained environments. The absence of power connectors and the IGP form factor mean it can fit into systems where a dual-slot, 150 W card would not be viable. The GTX 660 Ti requires a 450 W suggested PSU, while the Radeon Pro 560 has no such requirement listed.

The GTX 660 Ti’s higher shading unit count (1344 versus 1024) and FP32 throughput (2.634 TFLOPS versus 1.858 TFLOPS) suggest it could excel in compute workloads that are not captured by the recorded Geekbench tests. However, the actual benchmark results contradict that speculation, implying that the AMD architecture extracts more useful work per FLOP in the tested scenarios.

The Verdict

The recorded data favors the AMD Radeon Pro 560 in every head-to-head benchmark. It wins OpenCL by 2.6% and Vulkan by 8.1%, and its average benchmark score is 16.5% higher than the GTX 660 Ti’s. Its percentile rank (61st versus 57th) confirms a higher standing in the database’s overall GPU hierarchy. For users who prioritize modern API performance, compute workloads, or power efficiency, the Radeon Pro 560 is the statistically superior choice.

The GTX 660 Ti is not without merit. Its higher pixel rate, texture rate, and memory bandwidth point to strengths in legacy rasterization workloads. Users running older games or applications that rely heavily on fill rate might see different results, though the database does not include such tests. The NVIDIA card’s larger shading unit count and FP32 throughput also suggest untapped potential in compute tasks that are not represented in the two recorded benchmarks.

The decisive factor is the benchmark data itself. The Radeon Pro 560 wins both tests, and the margins are not trivial, especially in Vulkan. The 4 GB memory capacity is another practical advantage, as 2 GB is insufficient for many modern workloads. The 75 W TDP and IGP form factor make the AMD card more versatile across system types. The GTX 660 Ti remains a capable desktop GPU for its era, but the data indicates that the Radeon Pro 560 is the stronger performer in the metrics that matter most today.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon Pro 560 averages 17551 across all recorded tests, while the NVIDIA GeForce GTX 660 Ti averages 15063. That is a 16.5% difference in favor of AMD.

Q: How much faster is the Radeon Pro 560 in Vulkan?

A: The Radeon Pro 560 scores 16232 in Geekbench Vulkan, while the GTX 660 Ti scores 15012. The AMD card is 8.1% faster in that test.

Q: Does the GTX 660 Ti have any specification advantages?

A: Yes. The GTX 660 Ti has a higher pixel rate (27.44 GPixel/s versus 14.51 GPixel/s), a higher texture rate (109.8 GTexel/s versus 58.05 GTexel/s), and higher memory bandwidth (144.2 GB/s versus 81.28 GB/s). It also has more shading units (1344 versus 1024) and higher FP32 throughput (2.634 TFLOPS versus 1.858 TFLOPS).

Q: What are the memory size differences?

A: The AMD Radeon Pro 560 has 4 GB of GDDR5 memory on a 128-bit bus. The NVIDIA GeForce GTX 660 Ti has 2 GB of GDDR5 memory on a 192-bit bus.

Q: Which GPU has a lower power draw?

A: The AMD Radeon Pro 560 has a 75 W TDP and requires no power connectors. The NVIDIA GeForce GTX 660 Ti has a 150 W TDP and requires two 6-pin power connectors, with a suggested PSU of 450 W.

Q: What are the DirectX and Vulkan support levels?

A: The Radeon Pro 560 supports DirectX 12 (12_0) and Vulkan 1.3. The GTX 660 Ti supports DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6.

Specification Differences

| Specification | AMD Radeon Pro 560 | NVIDIA GeForce GTX 660 Ti |

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

| Architecture | GCN 4.0 | Kepler |

| Process Node | 14 nm | 28 nm |

| Foundry | GlobalFoundries | TSMC |

| Transistors | 3,000 million | 3,540 million |

| Die Size | 123 mm² | 294 mm² |

| Transistor Density | 24.4M / mm² | 12.0M / mm² |

| Base Clock | Not listed | 915 MHz |

| Boost Clock | Not listed | 980 MHz |

| Memory Clock | 1270 MHz (5.1 Gbps effective) | 1502 MHz (6 Gbps effective) |

| Memory Size | 4 GB | 2 GB |

| Memory Bus Width | 128 bit | 192 bit |

| Memory Bandwidth | 81.28 GB/s | 144.2 GB/s |

| Shading Units | 1024 | 1344 |

| TMUs | 64 | 112 |

| ROPs | 16 | 24 |

| Pixel Rate | 14.51 GPixel/s | 27.44 GPixel/s |

| Texture Rate | 58.05 GTexel/s | 109.8 GTexel/s |

| FP32 | 1.858 TFLOPS | 2.634 TFLOPS |

| FP16 | 1.858 TFLOPS (1:1) | Not listed |

| TDP | 75 W | 150 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 2x 6-pin |

| Suggested PSU | Not listed | 450 W |

| Bus Interface | PCIe 3.0 x8 | PCIe 3.0 x16 |

| Display Outputs | Portable Device Dependent | 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| DirectX | 12 (12_0) | 12 (11_0) |

| Vulkan | 1.3 | 1.2.175 |

| Length | Not listed | 241 mm (9.5 inches) |

| Release Date | 2017-04-17 | 2012-08-15 |

| Launch MSRP | Not listed | 299 USD |