AMD FirePro W5170M vs NVIDIA GeForce GTX 1660 Ti Comparison

FAQ

Q: Which GPU wins in Geekbench OpenCL performance?

A: The NVIDIA GeForce GTX 1660 Ti dominates this test with a score of 51,029, which is 84% higher than the AMD FirePro W5170M's 8,140. The FirePro trails by a massive margin in compute workloads.

Q: How do the two compare in Vulkan performance?

A: The AMD FirePro W5170M wins the Geekbench Vulkan test with 9,050 points versus 5,827 for the GeForce GTX 1660 Ti, a delta of 55.3% in AMD's favor. This is the only benchmark where the FirePro comes out ahead.

Q: What are the average benchmark scores for each GPU?

A: The AMD FirePro W5170M has an average benchmark score of 8,595, placing it in the 44th percentile of all GPUs. The NVIDIA GeForce GTX 1660 Ti averages 7,723, which lands it in the 41st percentile.

Q: Which GPU has higher memory bandwidth?

A: The NVIDIA GeForce GTX 1660 Ti offers 288.0 GB/s of bandwidth from 6 GB of GDDR6 memory on a 192-bit bus. The AMD FirePro W5170M provides 72.00 GB/s from 2 GB of GDDR5 on a 128-bit bus.

Q: What is the transistor count difference between the two chips?

A: The NVIDIA GeForce GTX 1660 Ti packs 6,600 million transistors on its TU116 chip, while the AMD FirePro W5170M has 1,500 million on the Tropo die. That is a 4.4x difference in transistor count.

Q: Which GPU has a higher boost clock?

A: The NVIDIA GeForce GTX 1660 Ti boosts to 1,770 MHz, compared to 925 MHz for the AMD FirePro W5170M. The NVIDIA part also runs at a 1,500 MHz base clock versus 900 MHz for the AMD chip.

Architecture Differences

The two GPUs come from completely different architectural generations. The AMD FirePro W5170M is built on GCN 1.0, using the Tropo chip manufactured on a 28 nm process at TSMC. The NVIDIA GeForce GTX 1660 Ti uses the Turing architecture, specifically the TU116 die, fabricated on a 12 nm process also at TSMC. This process difference alone gives NVIDIA a significant density advantage: 23.2 million transistors per mm² versus 12.2 million for AMD.

The transistor counts reflect the generational gap. AMD's Tropo die contains 1,500 million transistors on a 123 mm² die. NVIDIA's TU116 packs 6,600 million transistors onto a 284 mm² die. That is 4.4 times more transistors in a package that is only 2.3 times larger by area.

Shader resources diverge sharply. The FirePro W5170M has 640 shading units, 40 texture mapping units, and 16 ROPs. The GTX 1660 Ti scales up to 1,536 shading units, 96 TMUs, and 48 ROPs. These are 2.4x, 2.4x, and 3x increases respectively, which explains the large performance gap in most workloads.

Memory architecture also differs fundamentally. The AMD card ships with 2 GB of GDDR5 on a 128-bit bus, delivering 72.00 GB/s of bandwidth. The NVIDIA card carries 6 GB of GDDR6 on a 192-bit bus, reaching 288.0 GB/s. That is 4x the capacity and 4x the bandwidth.

Clock speeds favor NVIDIA heavily. The FirePro runs at 900 MHz base and 925 MHz boost, while the GTX 1660 Ti operates at 1,500 MHz base and 1,770 MHz boost. Combined with more cores, NVIDIA's compute throughput is far higher. The GTX 1660 Ti delivers 5.437 TFLOPS of FP32 performance and 10.87 TFLOPS of FP16 via a 2:1 ratio, while the FirePro manages 1,184.0 GFLOPS of FP32 with no listed FP16 support.

Pixel and texture rates tell the same story. NVIDIA achieves 84.96 GPixel/s and 169.9 GTexel/s, while AMD manages 14.80 GPixel/s and 37.00 GTexel/s. These are 5.7x and 4.6x differences, respectively.

API support shows a subtle but meaningful difference. Both support OpenGL 4.6 and Vulkan, but AMD lists Vulkan 1.2.170 while NVIDIA lists Vulkan 1.4. For DirectX, AMD supports 12 (11_1) while NVIDIA supports 12 (12_1), meaning NVIDIA has full DirectX 12 feature level support.

Form factors and power delivery differ as well. The FirePro is an MXM module with a MXM-A (3.0) bus interface and no power connectors, designed for portable devices. The GTX 1660 Ti is a dual-slot card measuring 229 mm by 111 mm by 35 mm, using PCIe 3.0 x16, a single 8-pin power connector, and a 300 W suggested PSU. NVIDIA's TDP is listed at 120 W, while AMD's TDP is not specified.

Head-to-Head Benchmarks

The database records two direct head-to-head comparisons between these GPUs. Each side wins one test, but the magnitudes of victory are wildly asymmetric.

In Geekbench OpenCL, the NVIDIA GeForce GTX 1660 Ti produces a score of 51,029 against the AMD FirePro W5170M's 8,140. The delta is 84% in NVIDIA's favor. This is a decisive victory that reflects the massive differences in shader count, clock speed, and memory bandwidth. The GTX 1660 Ti has more than 6 times the OpenCL score of the FirePro. For compute-heavy workloads, OpenCL performance is often the primary metric, and here NVIDIA is in a different league entirely.

In Geekbench Vulkan, the tables turn. The AMD FirePro W5170M scores 9,050, beating the GTX 1660 Ti's 5,827 by 55.3%. This is a notable upset given NVIDIA's hardware advantages. Vulkan performance can depend heavily on driver optimization and architecture-specific scheduling, and the GCN 1.0 design appears to handle this particular workload more efficiently relative to its raw specs. The FirePro's Vulkan score is actually higher than its OpenCL score, while NVIDIA's Vulkan score is drastically lower than its OpenCL score.

Looking at the broader benchmark context, the GTX 1660 Ti shows a wide range of results across different tests. In Passmark tests, it scores 12,871 in G3D, 5,226 in GPU compute, 804 in G2D, 190 in DirectX 9, 103 in DirectX 11, 66 in DirectX 10, and 51 in DirectX 12. It also records 1,061 in 3DMark Steel Nomad DX12. The FirePro has only OpenCL and Vulkan scores in the database, so direct comparison is limited to those two tests.

The average benchmark scores are somewhat counterintuitive. The FirePro averages 8,595 across its two recorded benchmarks, while the GTX 1660 Ti averages 7,723 across ten benchmarks. This is because NVIDIA's low Vulkan score drags down its average, while AMD's two scores are both moderately high. The percentile rankings reflect this: the FirePro sits at the 44th percentile, slightly above the GTX 1660 Ti's 41st percentile.

Rival comparisons add context. The FirePro's nearest rival is the NVIDIA Quadro P2200 at 8,686, which is 1.1% faster, and the Intel Arc A380 at 8,558, which is 0.4% slower. The GTX 1660 Ti's nearest rivals include the AMD Radeon 540 at 7,673 (0.7% slower), the AMD Radeon Pro WX 3100 at 7,580 (1.9% slower), the AMD Radeon R7 250 at 7,557 (2.2% slower), and the Intel Arc A310 at 7,550 (2.3% slower). These deltas are all small, suggesting the GTX 1660 Ti's average score is tightly clustered with its peers.

The Verdict

The data paints a clear picture for different use cases. The NVIDIA GeForce GTX 1660 Ti is the obvious choice for compute-heavy applications. Its OpenCL score of 51,029 is more than 6 times the FirePro's 8,140, and it delivers 5.437 TFLOPS of FP32 performance versus 1,184.0 GFLOPS. The 4x bandwidth advantage (288.0 GB/s vs 72.00 GB/s) and 4x memory capacity (6 GB vs 2 GB) make it far more capable for large datasets and modern workloads.

The AMD FirePro W5170M should be considered for Vulkan-specific scenarios. Its 55.3% lead in Geekbench Vulkan suggests that for applications built around this API, the FirePro can outperform the GTX 1660 Ti despite having far less raw hardware. The FirePro also has a higher average benchmark score (8,595 vs 7,723) and a higher percentile ranking (44 vs 41), though this is largely an artifact of having only two benchmarks recorded.

For mobile or embedded systems, the FirePro's MXM form factor and lack of power connectors make it appropriate for portable device integration. The GTX 1660 Ti requires a dual-slot footprint, a 120 W TDP, a 300 W suggested PSU, and an 8-pin power connector, which limits it to desktop builds with adequate power delivery.

The GTX 1660 Ti is end-of-life, as is the FirePro, but NVIDIA's part is backed by a much larger benchmark suite in the database, including ten tests across Passmark, 3DMark, and Geekbench. This gives a fuller picture of its capabilities. The FirePro's data is limited to two Geekbench tests, which constrains confidence in its overall standing.

In practical terms, choose the GTX 1660 Ti for general-purpose GPU workloads, gaming, and compute tasks. Its launch MSRP is 279 USD. Choose the FirePro only if Vulkan performance is the priority and the MXM form factor is required. The GTX 1660 Ti is the stronger all-around product.

Specification Differences

| Field | AMD FirePro W5170M | NVIDIA GeForce GTX 1660 Ti |

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

| Architecture | GCN 1.0 | Turing |

| Chip | Tropo | TU116 |

| Process Node | 28 nm | 12 nm |

| Transistors | 1,500 million | 6,600 million |

| Die Size | 123 mm² | 284 mm² |

| Transistor Density | 12.2M / mm² | 23.2M / mm² |

| Base Clock | 900 MHz | 1,500 MHz |

| Boost Clock | 925 MHz | 1,770 MHz |

| Memory Clock | 1,125 MHz (4.5 Gbps effective) | 1,500 MHz (12 Gbps effective) |

| Memory Size | 2 GB | 6 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus Width | 128 bit | 192 bit |

| Memory Bandwidth | 72.00 GB/s | 288.0 GB/s |

| Shading Units | 640 | 1,536 |

| TMUs | 40 | 96 |

| ROPs | 16 | 48 |

| Pixel Rate | 14.80 GPixel/s | 84.96 GPixel/s |

| Texture Rate | 37.00 GTexel/s | 169.9 GTexel/s |

| FP32 Performance | 1,184.0 GFLOPS | 5.437 TFLOPS |

| FP16 Performance | Not listed | 10.87 TFLOPS (2:1) |

| TDP | Not listed | 120 W |

| Slot Width | MXM Module | Dual-slot |

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

| Suggested PSU | Not listed | 300 W |

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

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

| DirectX Support | 12 (11_1) | 12 (12_1) |

| OpenGL Support | 4.6 | 4.6 |

| Vulkan Support | 1.2.170 | 1.4 |

| Length | Not listed | 229 mm (9 inches) |

| Height | Not listed | 111 mm (4.4 inches) |

| Width | Not listed | 35 mm (1.4 inches) |