AMD Radeon Pro WX 3100 vs NVIDIA GeForce GTX 660 Comparison

Head-to-Head Benchmarks

The recorded data leaves little ambiguity: in the two shared workloads, the NVIDIA GeForce GTX 660 wins outright. The GTX 660 posts a Geekbench OpenCL score of 11,347, which is 54.7% ahead of the AMD Radeon Pro WX 3100’s 7,333. In the Vulkan test, the margin narrows but remains decisive: 11,415 versus 7,827, a 45.8% lead for the NVIDIA card. The head-to-head tally stands at 2 wins for the GTX 660 and 0 for the Radeon Pro WX 3100.

These are not marginal differences. A 54.7% OpenCL advantage means the GTX 660 delivers roughly half again as much compute throughput in that API. The Vulkan gap, while smaller, still places the GTX 660 nearly 46% ahead, indicating that the NVIDIA card’s advantage is consistent across different driver stacks and workload types. For any application that relies on OpenCL or Vulkan, the GTX 660 is the stronger performer in this pairing.

Looking at broader positioning, the GTX 660’s average benchmark score is 9,022, which places it at the 45th percentile of all GPUs in the database. The Radeon Pro WX 3100 averages 7,580, sitting at the 41st percentile. That is a 1,442-point gap in average score, or roughly 19% higher for the NVIDIA part. The nearest rivals for the GTX 660 include the NVIDIA TITAN V CEO Edition (average 9,037, only 0.2% higher), the GeForce GTX 560 (9,058, 0.4% higher), the AMD Radeon 550X (8,918, 1.2% lower), and the AMD Radeon Pro WX 5100 (8,863, 1.8% lower). So the GTX 660 sits in a tight cluster near the 9,000 mark, trading blows with cards from both vendors.

For the Radeon Pro WX 3100, its nearest rivals are the AMD Radeon R7 250 (7,557, 0.3% lower), Intel Arc A310 (7,550, 0.4% lower), AMD Radeon 540 (7,673, 1.2% higher), and NVIDIA GeForce GTX 1650 (7,472, 1.4% higher). The WX 3100 is essentially bracketed by other entry-level cards, with nothing in its immediate vicinity that it dominates by more than 1.4%. This reinforces the picture: the WX 3100 is a modest performer, while the GTX 660, despite being the older architecture, commands a clear performance tier above it.

The data is unambiguous on the compute front. The GTX 660’s FP32 throughput is 1.981 TFLOPS, while the WX 3100 manages 1,248.3 GFLOPS. That is a 58.7% raw compute advantage for the NVIDIA card, closely mirroring the OpenCL benchmark delta. Even in pixel fill rate, the GTX 660’s 20.64 GPixel/s edges out the WX 3100’s 19.50 GPixel/s, a smaller but still real lead. Texture rate is where the NVIDIA card runs away: 82.56 GTexel/s versus 39.01 GTexel/s, a 111.7% advantage that reflects the GTX 660’s far larger TMU count (80 versus 32).

FAQ

Q: Which GPU wins the head-to-head benchmark comparison?

A: The NVIDIA GeForce GTX 660 wins both shared tests. It scores 11,347 in Geekbench OpenCL versus 7,333 for the AMD Radeon Pro WX 3100, a 54.7% lead, and 11,415 in Geekbench Vulkan versus 7,827, a 45.8% lead.

Q: How do their average benchmark scores compare?

A: The GTX 660 has an average benchmark score of 9,022, placing it at the 45th percentile of all GPUs. The Radeon Pro WX 3100 averages 7,580, at the 41st percentile. That is a 19% higher average score for the NVIDIA card.

Q: Are these cards close in performance to their own rivals?

A: No. The GTX 660’s nearest rivals, like the NVIDIA TITAN V CEO Edition and GeForce GTX 560, are within 0.4% of its average score. The Radeon Pro WX 3100’s nearest rivals, including the AMD Radeon R7 250 and Intel Arc A310, are within 0.4% as well. Both cards sit in tight clusters, but the GTX 660’s cluster is roughly 1,400 points higher.

Q: What is the difference in memory configuration?

A: The GTX 660 has 2 GB of GDDR5 on a 192-bit bus with 144.2 GB/s of bandwidth. The Radeon Pro WX 3100 has 4 GB of GDDR5 on a 128-bit bus with 96.00 GB/s of bandwidth. The WX 3100 has double the capacity, but the GTX 660 has 50% more bandwidth.

Q: Which card has the higher boost clock?

A: The AMD Radeon Pro WX 3100 boosts to 1,219 MHz, while the NVIDIA GeForce GTX 660 boosts to 1,032 MHz. The WX 3100’s base clock is 925 MHz versus 980 MHz for the GTX 660.

Q: Do either of these cards support modern APIs?

A: Both support DirectX 12 and OpenGL 4.6. The GTX 660 supports Vulkan 1.2.175, while the WX 3100 supports Vulkan 1.3. The WX 3100 also supports DirectX 12 (12_0), whereas the GTX 660 is limited to DirectX 12 (11_0).

Architecture Differences

The two GPUs come from different architectural generations and foundries. The NVIDIA GeForce GTX 660 uses the GK106 chip on the Kepler architecture, built on a 28 nm process at TSMC. The die measures 221 mm² and contains 2,540 million transistors, yielding a transistor density of 11.5M per mm². The AMD Radeon Pro WX 3100 uses the Lexa chip on GCN 4.0, built on a 14 nm process at GlobalFoundries. Its die is much smaller at 103 mm², with 2,200 million transistors, giving a higher density of 21.4M per mm². So the AMD part is newer, denser, and more power-efficient per transistor, but the NVIDIA chip packs more silicon area and raw resources.

The compute configuration differs sharply. The GTX 660 has 960 shading units, 80 texture mapping units, and 24 ROPs. The WX 3100 has 512 shading units, 32 TMUs, and 16 ROPs. The NVIDIA card has nearly double the shading units, 2.5 times the TMUs, and 1.5 times the ROPs. This explains the large texture rate and FP32 advantages. The GTX 660’s FP32 output is 1.981 TFLOPS; the WX 3100’s is 1,248.3 GFLOPS. Notably, the WX 3100 supports FP16 at a 1:1 ratio with FP32 (1,248.3 GFLOPS), while the GTX 660 has no listed FP16 capability. That makes the AMD card more flexible for mixed-precision workloads, even though it loses on absolute FP32.

Memory architecture also diverges. The GTX 660 uses a 192-bit bus with 2 GB of GDDR5 at 6 Gbps effective, yielding 144.2 GB/s of bandwidth. The WX 3100 uses a 128-bit bus with 4 GB of GDDR5 at 6 Gbps effective, yielding 96.00 GB/s. The NVIDIA card has higher bandwidth, the AMD card has more capacity. The GTX 660’s pixel rate is 20.64 GPixel/s, slightly above the WX 3100’s 19.50 GPixel/s, despite the AMD card having a higher boost clock (1,219 MHz versus 1,032 MHz). The NVIDIA card compensates with more ROPs and a wider memory bus.

Power and physical design reflect their different targets. The GTX 660 has a TDP of 140 W, requires a 1x 6-pin power connector, and a suggested PSU of 300 W. It is dual-slot and measures 241 mm (9.5 inches) in length. The WX 3100 has a TDP of 65 W, requires no power connectors, and a suggested PSU of 250 W. It is single-slot and measures 168 mm (6.6 inches) long and 69 mm (2.7 inches) tall. The AMD card is clearly designed for compact, low-power professional workstations, while the NVIDIA card is a conventional dual-slot gaming-era board.

Bus interface and display outputs differ as well. The GTX 660 uses PCIe 3.0 x16, while the WX 3100 uses PCIe 3.0 x8. The GTX 660 outputs via 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The WX 3100 outputs via 1x DisplayPort 1.4a and 2x mini-DisplayPort 1.4a, offering newer display connectivity.

Specification Differences

| Specification | NVIDIA GeForce GTX 660 | AMD Radeon Pro WX 3100 |

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

| Architecture | Kepler | GCN 4.0 |

| Process node | 28 nm | 14 nm |

| Foundry | TSMC | GlobalFoundries |

| Transistors | 2,540 million | 2,200 million |

| Die size | 221 mm² | 103 mm² |

| Transistor density | 11.5M / mm² | 21.4M / mm² |

| Base clock | 980 MHz | 925 MHz |

| Boost clock | 1032 MHz | 1219 MHz |

| Memory size | 2 GB | 4 GB |

| Memory bus width | 192 bit | 128 bit |

| Memory bandwidth | 144.2 GB/s | 96.00 GB/s |

| Shading units | 960 | 512 |

| TMUs | 80 | 32 |

| ROPs | 24 | 16 |

| Pixel rate | 20.64 GPixel/s | 19.50 GPixel/s |

| Texture rate | 82.56 GTexel/s | 39.01 GTexel/s |

| FP32 | 1.981 TFLOPS | 1,248.3 GFLOPS |

| FP16 | None listed | 1,248.3 GFLOPS (1:1) |

| TDP | 140 W | 65 W |

| Slot width | Dual-slot | Single-slot |

| Power connectors | 1x 6-pin | None |

| Suggested PSU | 300 W | 250 W |

| Bus interface | PCIe 3.0 x16 | PCIe 3.0 x8 |

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

| Vulkan | 1.2.175 | 1.3 |

| Length | 241 mm (9.5 inches) | 168 mm (6.6 inches) |

| Height | Not listed | 69 mm (2.7 inches) |

| Launch MSRP | 229 USD | 199 USD |

The Verdict

The data points to a clear performance winner: the NVIDIA GeForce GTX 660. It wins both head-to-head benchmarks decisively, with a 54.7% OpenCL lead and a 45.8% Vulkan lead. Its average benchmark score of 9,022 is 19% higher than the Radeon Pro WX 3100’s 7,580. The GTX 660 also holds advantages in texture rate (82.56 GTexel/s versus 39.01 GTexel/s), FP32 compute (1.981 TFLOPS versus 1,248.3 GFLOPS), and memory bandwidth (144.2 GB/s versus 96.00 GB/s). For raw compute throughput, the older NVIDIA card is simply in a different class.

However, the Radeon Pro WX 3100 is not without reasons to exist. It has double the memory capacity (4 GB versus 2 GB), a much lower TDP (65 W versus 140 W), a single-slot design, and no external power connector. It supports newer display outputs (DisplayPort 1.4a versus 1.2), a newer Vulkan version (1.3 versus 1.2.175), and FP16 compute at a 1:1 ratio, which the GTX 660 lacks entirely. Its smaller die (103 mm² versus 221 mm²) and higher transistor density (21.4M / mm² versus 11.5M / mm²) reflect a more modern manufacturing process.

The choice depends on the use case. If the priority is maximum compute performance in OpenCL or Vulkan, the GTX 660 is the correct pick, and the margin is substantial. If the priority is low power draw, compact size, modern display outputs, or double the VRAM, the WX 3100 fits that niche. The GTX 660’s launch MSRP was 229 USD, while the WX 3100’s was 199 USD, but neither card is currently in production. Benchmark data places the GTX 660 at the 45th percentile of all GPUs and the WX 3100 at the 41st, so the NVIDIA card also ranks higher in the overall distribution.

Where Each One Wins

NVIDIA GeForce GTX 660 wins on: all head-to-head compute benchmarks (OpenCL and Vulkan), FP32 throughput, texture fill rate, pixel fill rate, memory bandwidth, average benchmark score, and overall GPU percentile ranking.

AMD Radeon Pro WX 3100 wins on: memory capacity (4 GB versus 2 GB), power efficiency (65 W TDP versus 140 W), physical footprint (single-slot versus dual-slot, shorter length), no external power requirement, newer display outputs (DisplayPort 1.4a), newer Vulkan support (1.3), DirectX 12 (12_0) feature level, FP16 compute support, and higher boost clock (1,219 MHz versus 1,032 MHz).

For compute-heavy tasks such as OpenCL rendering or Vulkan-based workloads, the GTX 660 is the stronger card by a wide margin. For compact workstation builds, low-noise systems, or applications that need more than 2 GB of VRAM, the WX 3100 provides a viable alternative with modern I/O. The recorded data does not show any workload where the WX 3100 outperforms the GTX 660, but its feature set targets a different set of constraints: power, size, and memory capacity over raw speed.