NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce GTX 560 Comparison

# NVIDIA GeForce GTX 560 vs NVIDIA GeForce GTX 1660 Ti

The GeForce GTX 560 and the GeForce GTX 1660 Ti belong to different eras of NVIDIA's product stack. The GTX 560 is a Fermi 2.0 architecture part from the GeForce 500 series, built on a 40 nm process at TSMC. The GTX 1660 Ti, meanwhile, is a Turing architecture chip from the GeForce 16 series, fabricated on a 12 nm node. The database records a substantial generational gap between them: the older card ships with 1,950 million transistors on a 332 mm² die, while the newer card packs 6,600 million transistors into a 284 mm² die. That difference in transistor density (5.9M per mm² versus 23.2M per mm²) is the first clear signal that these two products solve different problems.

In aggregate benchmark scoring, the GTX 560 lands at the 45th percentile of all GPUs in the database, with an average score of 9,058. The GTX 1660 Ti sits at the 41st percentile, with an average score of 7,723. The newer card wins the single head-to-head benchmark, the OpenCL test, by a wide margin: 51,029 against 9,058, a delta of 82.2% in favor of the 1660 Ti. That head-to-head result is the strongest single piece of evidence in this comparison.

FAQ

Q: Which card has the higher transistor count?

A: The GeForce GTX 1660 Ti. It packs 6,600 million transistors, whereas the GTX 560 has 1,910 million.

Q: What is the memory configuration of each card?

A: The GTX 560 has 1024 MB of GDDR5 on a 256-bit bus, delivering 128.0 GB/s of bandwidth. The GTX 1660 Ti comes with 6 GB of GDDR6 on a 192-bit bus, for 288.0 GB/s.

Q: How do their shading and rasterization resources compare?

A: The GTX 560 has 336 shading units and 32 ROPs. The GTX 1660 Ti scales up to 1,536 shading units and 48 ROPs.

Q: Which card supports newer DirectX and OpenGL versions?

A: The GTX 1660 Ti supports DirectX 12 (12_1) and OpenGL 4.6. The GTX 560 supports DirectX 12 (11_0) and OpenGL 4.6.

Q: What is the thermal design power of each card?

A: The GTX 560 is rated at 170 W, while the GTX 1660 Ti is rated at 120 W.

Architecture Differences

The GTX 560 is built on the Fermi 2.0 architecture, a design NVIDIA used across the GeForce 500 family. Its chip is the GF114, a 40 nm part with a 332 mm² die. The GTX 1660 Ti is a Turing chip, TU116, built on a 12 nm TSMC process. The die area is slightly smaller at 284 mm², but the transistor count grows to 6,600 million. The newer process node is the main reason the 1660 Ti can fit more compute despite the smaller physical footprint.

The two cards also differ in their memory subsystems. The GTX 560 uses a 256-bit GDDR5 bus with a memory clock of 1000 MHz, producing an effective 4 Gbps data rate and 128.0 GB/s of bandwidth. The GTX 1660 Ti uses a 192-bit GDDR6 interface running at 1500 MHz, with an effective 12 Gbps, resulting in 288.0 GB/s. The bandwidth difference is among the largest quantitative gaps in the comparison.

The rendering resources follow the same story. The GTX 560 has 336 shading units and 32 ROPs. The GTX 1660 Ti has 1,536 shading units and 48 ROPs, roughly 4.6x and 1.5x the older card. Pixel throughput on the GTX 560 is 11.34 GPixel/s and texture rate is 45.36 GTexel/s. The GTX 1660 Ti reaches 84.96 GPixel/s and 169.9 GTexel/s.

Output connectivity also changes generationally. The GTX 560 uses two DVI outputs plus a mini-HDMI 1.3a port, while the GTX 1660 Ti provides one DVI, one HDMI 2.0, and one DisplayPort 1.4a.

Head-to-Head Benchmarks

The only direct benchmark in the database is the OpenCL test, and it is not close. The GTX 1660 Ti scores 51,029 versus 9,058 for the GTX 560, a difference of 82.2%. That result is consistent with the architectural distance between the two products: the newer card has more shading units, more memory bandwidth, and a much younger process node.

The pixel and texture rates gap is also decisive. The GTX 1660 Ti's 84.96 GPixel/s is 7.5 times the GTX 560's 11.34 GPixel/s. Texture rate goes from 45.36 GTexel/s to 169.9 GTexel/s, a 3.7x jump. The GTX 560 does keep a lead in none of the recorded measurements; every recorded number favors the 1660 Ti.

Specification Differences

| Field | NVIDIA GeForce GTX 560 | NVIDIA GeForce GTX 1660 Ti |

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

| Chip | GF114 | TU116 |

| Architecture | Fermi 2.0 | Turing |

| Process node | 40 nm | 12 nm |

| Transistors | 1,950 million | 6,600 million |

| Die size | 332 mm² | 284 mm² |

| Memory bus | 256 bit | 192 bit |

| Memory type | GDDR5 | GDDR6 |

| ROPs | 32 | 48 |

| Shading units | 336 | 1,536 |

| Power connectors | 2x 6-pin | 1x 8-pin |

| Bus interface | PCIe 2.0 x16 | PCIe 3.0 x16 |

| APIs | DirectX 12 (11_0), OpenGL 4.6 | DirectX 12 (12_1), OpenGL 4.6, Vulkan 1.4 |

Where Each One Wins

The GeForce GTX 1660 Ti wins the only direct head-to-head benchmark, the OpenCL compute test, by 82.2%. It also wins every architectural comparison in the database: newer process, more transistors, higher memory bandwidth, more shading units, and higher pixel and texture rates.

The GeForce GTX 560 does not win any recorded category. Its percentile rank (45th) is lower than the 1660 Ti's 41st percentile, but that is a relative placement, not a category win. In the one-on-one comparison, the 1660 Ti posts a 82.2% score advantage.

For workloads where raw single-thread OpenCL throughput is the limiting factor, the GTX 1660 Ti has a clear edge. The 560's older Fermi architecture, with its lower shading unit count and much lower memory bandwidth, places it at a disadvantage in any modern compute-heavy task. The GTX 1660 Ti is also the more power-efficient part in the database records, rated at 120 W versus 150 W.

The Verdict

The database points one direction. The GeForce GTX 1660 Ti leads the GTX 560 in every recorded measure: the OpenCL test, the pixel throughput, the texture rate, shading units, memory bandwidth, and transistor density. The GTX 560 remains a representative example of its 2011 generation, but the GTX 1660 Ti is the faster card in every recorded metric.

Pick the GeForce GTX 1660 Ti if the task is modern compute-heavy workloads, since the data shows a 82.2% lead in the specific OpenCL test. Pick the GeForce GTX 560 only if a system requires its older PCIe 2.0 host interface or its specific 2x 6-pin power input, as the newer card uses an 8-pin connector.