NVIDIA GeForce GTX 1660 vs NVIDIA GeForce GTX 960A Comparison

# NVIDIA GeForce GTX 960A vs NVIDIA GeForce GTX 1660

The GeForce GTX 960A is an end-of-life MXM module from 2015 built on NVIDIA's Maxwell architecture, while the GeForce GTX 1660 is a 2019 desktop card from the GeForce 16 series using Turing architecture. The data shows a single head-to-head benchmark, Geekbench OpenCL, where the GTX 1660 scores 47,850 against the GTX 960A's 11,998 — a 74.9% deficit for the older card. The GTX 1660 wins the only direct comparison, but the GTX 960A holds a narrow edge in average benchmark score across all recorded tests, at 11,998 versus 11,680, a 2.7% difference that flips the expected order.

Where Each One Wins

The GTX 960A wins on average benchmark score, but this is a statistical artifact rather than a performance victory. Its only recorded benchmark, Geekbench OpenCL at 11,998, places it 0.3% ahead of the GTX 1080 and 1.3% ahead of the RX 6500 XT. The GTX 1660's average of 11,680 is dragged down by its broader test suite, which includes lower-scoring DirectX 9 and DirectX 10 Passmark tests. The GTX 1660 wins the only apples-to-apples comparison decisively, and its benchmark portfolio shows strength across modern APIs: Geekbench Vulkan at 50,137, Geekbench OpenCL at 47,850, and Passmark G3D at 11,646.

For modern gaming and compute workloads, the GTX 1660 is the clear winner. Its 6 GB VRAM and 192-bit memory bus provide headroom that the GTX 960A's 2 GB and 128-bit bus cannot match. The GTX 960A's only meaningful "win" is in the synthetic average, where it benefits from having a single high-scoring test rather than a diverse set. In real-world terms, the GTX 1660 dominates every workload category where both cards have data.

Architecture Differences

The GTX 960A uses the GM107 chip on a 28 nm TSMC process, packing 1,870 million transistors into a 148 mm² die — a transistor density of 12.6 million per square millimeter. The GTX 1660 uses the TU116 chip on a 12 nm TSMC process, with 6,600 million transistors across a 284 mm² die, yielding 23.2 million transistors per square millimeter. The GTX 1660's process advantage is stark: nearly double the density, which translates into far more compute resources.

Shader counts tell the story. The GTX 960A has 640 shading units, 40 texture mapping units, and 16 raster operations pipelines. The GTX 1660 has 1,408 shading units, 88 TMUs, and 48 ROPs — more than double the shading units, more than double the TMUs, and triple the ROPs. Clock speeds also favor the newer card: the GTX 1660 boosts to 1785 MHz versus 1176 MHz for the GTX 960A. The GTX 1660 supports DirectX 12 (12_1) while the GTX 960A only reaches DirectX 12 (11_0); both support OpenGL 4.6 and Vulkan 1.4.

Memory architecture diverges sharply. The GTX 960A has 2 GB GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The GTX 1660 has 6 GB GDDR5 on a 192-bit bus with 192.1 GB/s bandwidth — 2.4 times the bandwidth and triple the capacity. The GTX 1660 also supports FP16 compute at 10.05 TFLOPS via a 2:1 ratio, while the GTX 960A has no FP16 data. Power delivery differs: the GTX 960A is a 75 W MXM module with no power connectors, while the GTX 1660 is a 120 W dual-slot card requiring a single 8-pin connector and a 300 W PSU.

Head-to-Head Benchmarks

The only direct comparison is Geekbench OpenCL, and it is not close. The GTX 1660 scores 47,850 against the GTX 960A's 11,998, a 74.9% margin. This fourfold difference reflects the architectural gap: the GTX 1660's 1,408 shading units and 5.027 TFLOPS FP32 throughput dwarf the GTX 960A's 640 shading units and 1.505 TFLOPS. The memory bandwidth difference — 192.1 GB/s versus 80.19 GB/s — further amplifies the gap in OpenCL workloads that stress memory access.

Beyond the direct comparison, the GTX 1660's other benchmarks show where its strengths lie. Its Geekbench Vulkan score of 50,137 is 4.8% higher than its OpenCL score, indicating strong driver optimization for modern graphics APIs. Passmark G3D at 11,646 nearly matches the GTX 960A's entire average benchmark score. The GTX 1660 also shows balanced performance across DirectX versions: DirectX 9 at 177, DirectX 11 at 79, DirectX 10 at 61, and DirectX 12 at 49. These numbers suggest the GTX 1660 handles legacy titles well but scales even better with newer APIs.

For the GTX 960A, the single Geekbench OpenCL score of 11,998 places it in the 51st percentile of all GPUs — the same percentile as the GTX 1660's 51st percentile. This is the only metric where the two cards appear comparable. The deltaPct values in the rival lists tell the same story: the GTX 960A is 0.3% ahead of the GTX 1080, while the GTX 1660 is 0.5% ahead of the RX 7800 XT. Both cards sit in a crowded mid-pack, but the GTX 1660's absolute scores are far higher.

FAQ

Q: Which card has better raw compute performance?

A: The GTX 1660. It delivers 5.027 TFLOPS FP32 versus the GTX 960A's 1.505 TFLOPS, a 3.3x advantage. In Geekbench OpenCL, the GTX 1660 scores 47,850 versus 11,998, a 74.9% lead.

Q: How do memory capacities and bandwidth compare?

A: The GTX 1660 has 6 GB GDDR5 on a 192-bit bus with 192.1 GB/s bandwidth. The GTX 960A has 2 GB GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The GTX 1660 offers triple the capacity and 2.4 times the bandwidth.

Q: Do both cards support the same DirectX version?

A: No. The GTX 1660 supports DirectX 12 (12_1), while the GTX 960A only supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Q: What are the power requirements?

A: The GTX 960A is a 75 W MXM module with no power connectors. The GTX 1660 is a 120 W dual-slot card with one 8-pin connector and a 300 W suggested PSU.

Q: Which card has higher clock speeds?

A: The GTX 1660 runs at 1530 MHz base and 1785 MHz boost. The GTX 960A runs at 1097 MHz base and 1176 MHz boost. The GTX 1660's boost clock is 609 MHz higher.

Q: Are both cards end-of-life?

A: Yes. The GTX 960A ended production with a 2015 release date, and the GTX 1660 ended production with a 2019 release date. The GTX 1660's launch MSRP was 219 USD.

The Verdict

The data is unambiguous for anyone building or upgrading a desktop system: the GTX 1660 is the superior card in every measured dimension except average benchmark score, and that exception is misleading. The GTX 960A's 11,998 average comes from a single test, while the GTX 1660's 11,680 average spans ten tests including legacy DirectX 9 and compute workloads. In the only direct benchmark, the GTX 1660 wins by 74.9%.

The GTX 960A is a mobile-oriented MXM module from the Maxwell era, designed for portable devices with a 75 W power envelope and no external power connectors. It is end-of-life, has 2 GB VRAM that modern games will exhaust quickly, and its 80.19 GB/s bandwidth is a bottleneck. The GTX 1660, despite also being end-of-life, offers 6 GB VRAM, 192.1 GB/s bandwidth, and 5.027 TFLOPS — enough for 1080p gaming across most titles at medium-to-high settings.

Choose the GTX 1660 if you need desktop performance, modern API support, or any kind of compute workload. Its Vulkan score of 50,137 and DirectX 12 (12_1) support make it future-proof for a 2019 card. Choose the GTX 960A only if you have a specific MXM-based portable system that requires a 75 W, connector-free card — and even then, the data shows it will perform at roughly a quarter of the GTX 1660's level in OpenCL. The percentile ranking of 51 for both cards masks a yawning absolute gap that only grows with workload complexity.

Specification Differences

| Specification | GTX 960A | GTX 1660 |

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

| Chip | GM107 | TU116 |

| Architecture | Maxwell | Turing |

| Process Node | 28 nm | 12 nm |

| Transistors | 1,870 million | 6,600 million |

| Die Size | 148 mm² | 284 mm² |

| Transistor Density | 12.6M / mm² | 23.2M / mm² |

| Base Clock | 1097 MHz | 1530 MHz |

| Boost Clock | 1176 MHz | 1785 MHz |

| Memory Clock | 1253 MHz (5 Gbps effective) | 2001 MHz (8 Gbps effective) |

| Memory Size | 2 GB | 6 GB |

| Memory Bus Width | 128 bit | 192 bit |

| Memory Bandwidth | 80.19 GB/s | 192.1 GB/s |

| Shading Units | 640 | 1408 |

| TMUs | 40 | 88 |

| ROPs | 16 | 48 |

| Pixel Rate | 18.82 GPixel/s | 85.68 GPixel/s |

| Texture Rate | 47.04 GTexel/s | 157.1 GTexel/s |

| FP32 | 1.505 TFLOPS | 5.027 TFLOPS |

| FP16 | — | 10.05 TFLOPS (2:1) |

| TDP | 75 W | 120 W |

| Slot Width | MXM Module | Dual-slot |

| Power Connectors | None | 1x 8-pin |

| Suggested PSU | — | 300 W |

| Bus Interface | MXM-B (3.0) | PCIe 3.0 x16 |

| Display Outputs | Portable Device Dependent | 1x DVI, 1x HDMI 2.0, 1x DisplayPort 1.4a |

| DirectX | 12 (11_0) | 12 (12_1) |

| Dimensions | — | 229 mm x 111 mm x 35 mm |

| Release Date | 2015-03-12 | 2019-03-13 |

| Predecessor | GeForce 800A | GeForce 10 |

| Successor | — | GeForce 20 |

| Launch MSRP | — | 219 USD |