NVIDIA GeForce GTX 660 vs NVIDIA Quadro 6000 Comparison

FAQ

Q: Which GPU wins the only head-to-head benchmark available?

A: The NVIDIA GeForce GTX 660 wins the sole head-to-head test, Geekbench OpenCL, with a score of 11,347 versus the Quadro 6000's 9,846. That is a 13.2% advantage for the GTX 660.

Q: How do the two GPUs compare in overall benchmark percentile ranking?

A: The Quadro 6000 sits at the 47th percentile of all GPUs, while the GTX 660 is at the 45th percentile. Despite the GTX 660 winning the direct comparison, its average benchmark score (9,022) is lower than the Quadro's (9,846).

Q: What are the closest rivals to each GPU according to the data?

A: The Quadro 6000's nearest rival is the NVIDIA Quadro M2000M (average score 9,832, just 0.1% behind). The GTX 660's nearest rival is the NVIDIA TITAN V CEO Edition (average score 9,037, only 0.2% ahead of the GTX 660).

Q: Which GPU has the higher memory capacity?

A: The Quadro 6000 has 6 GB of GDDR5 memory, triple the 2 GB found on the GTX 660. Both use GDDR5, and their memory bandwidth is nearly identical: 143.4 GB/s for the Quadro versus 144.2 GB/s for the GTX 660.

Q: What is the difference in power requirements between the two cards?

A: The Quadro 6000 has a TDP of 204 W and requires a 550 W power supply with 1x 6-pin plus 1x 8-pin connectors. The GTX 660 has a 140 W TDP, needs only a 300 W power supply, and uses a single 6-pin connector.

Q: Do both cards support the same graphics APIs?

A: Both support DirectX 12 (11_0) and OpenGL 4.6. However, the GTX 660 additionally supports Vulkan 1.2.175, while the Quadro 6000 has no Vulkan support listed.

Architecture Differences

The two GPUs come from entirely different NVIDIA architectures and process nodes. The Quadro 6000 is built on the Fermi architecture using the GF100 chip, manufactured on a 40 nm process at TSMC. The GTX 660 uses the Kepler architecture with the GK106 chip, fabricated on a 28 nm process, also at TSMC. This process shrink is substantial: the GTX 660 packs 2,540 million transistors into a 221 mm² die, yielding a transistor density of 11.5M per mm². The Quadro 6000, by contrast, has 3,100 million transistors on a much larger 529 mm² die, giving a density of just 5.9M per mm².

The generational divide is clear. The Quadro 6000 belongs to the Quadro Fermi (x000) generation, released in December 2010, while the GTX 660 is part of the GeForce 600 series, launched in September 2012. Their predecessors and successors also differ: the Quadro succeeded the Quadro FX Tesla line and was followed by Quadro Kepler, whereas the GTX 660 succeeded GeForce 500 and was followed by GeForce 700.

Shader organization is fundamentally different. The Quadro 6000 has 448 shading units, 56 texture mapping units, and 48 ROPs. The GTX 660 has 960 shading units, 80 TMUs, and only 24 ROPs. That is more than double the shading units and more TMUs on the GTX 660, but half the ROP count. Neither card has ray tracing cores or tensor cores.

The memory subsystem also differs in bus width. The Quadro 6000 uses a 384-bit memory bus, while the GTX 660 uses a 192-bit bus. Despite this, their bandwidth is nearly equal because the GTX 660 runs its memory at 1,502 MHz (6 Gbps effective) versus 747 MHz (3 Gbps effective) on the Quadro. The Quadro compensates with its wider bus and higher capacity.

Clock behavior differs too. The GTX 660 has a base clock of 980 MHz and a boost clock of 1,032 MHz; the Quadro 6000 has no base or boost clock listed. Interface support also differs: the Quadro uses PCIe 2.0 x16, while the GTX 660 uses PCIe 3.0 x16.

Where Each One Wins

The GTX 660 wins the only direct benchmark, Geekbench OpenCL, by 13.2%. That is the clearest performance statement in the data. Its compute-oriented strengths are also visible in raw throughput figures: the GTX 660 delivers 1.981 TFLOPS of FP32 performance versus 1,027.7 GFLOPS for the Quadro 6000 — nearly double the compute throughput. The GTX 660 also has far higher texture rate (82.56 GTexel/s versus 32.14 GTexel/s) and a higher pixel rate (20.64 GPixel/s versus 16.07 GPixel/s).

The Quadro 6000 wins on memory capacity and ROP count. Its 6 GB frame buffer is three times larger than the GTX 660's 2 GB, and its 48 ROPs are double the GTX 660's 24. The wider 384-bit memory bus also favors the Quadro, although the bandwidth figures are nearly identical. The Quadro's higher average benchmark score (9,846) and higher percentile ranking (47th versus 45th) indicate it performs better across a broader set of workloads than the GTX 660, even though it loses the single direct OpenCL test.

The GTX 660 is the more efficient card. Its TDP of 140 W is 64 W lower than the Quadro's 204 W, and it requires a 300 W power supply versus 550 W. The GTX 660 also has more modern API support, including Vulkan, and a newer PCIe interface.

Specification Differences

| Specification | NVIDIA Quadro 6000 | NVIDIA GeForce GTX 660 |

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

| Chip | GF100 | GK106 |

| Architecture | Fermi | Kepler |

| Generation | Quadro Fermi (x000) | GeForce 600 |

| Process Node | 40 nm | 28 nm |

| Transistors | 3,100 million | 2,540 million |

| Die Size | 529 mm² | 221 mm² |

| Transistor Density | 5.9M / mm² | 11.5M / mm² |

| Base Clock | — | 980 MHz |

| Boost Clock | — | 1,032 MHz |

| Memory Clock | 747 MHz (3 Gbps effective) | 1,502 MHz (6 Gbps effective) |

| Memory Size | 6 GB | 2 GB |

| Memory Bus Width | 384 bit | 192 bit |

| Memory Bandwidth | 143.4 GB/s | 144.2 GB/s |

| Shading Units | 448 | 960 |

| TMUs | 56 | 80 |

| ROPs | 48 | 24 |

| Pixel Rate | 16.07 GPixel/s | 20.64 GPixel/s |

| Texture Rate | 32.14 GTexel/s | 82.56 GTexel/s |

| FP32 | 1,027.7 GFLOPS | 1.981 TFLOPS |

| TDP | 204 W | 140 W |

| Power Connectors | 1x 6-pin + 1x 8-pin | 1x 6-pin |

| Suggested PSU | 550 W | 300 W |

| Bus Interface | PCIe 2.0 x16 | PCIe 3.0 x16 |

| Display Outputs | 1x DVI, 2x DisplayPort, 1x S-Video | 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| Vulkan Support | — | 1.2.175 |

| Release Date | 2010-12-09 | 2012-09-05 |

| Launch MSRP | 4,399 USD | 229 USD |

Head-to-Head Benchmarks

The only direct benchmark comparison is Geekbench OpenCL. The GTX 660 scores 11,347, while the Quadro 6000 scores 9,846. That is a 13.2% margin in favor of the GTX 660. This is a decisive win for the Kepler card in a compute-oriented OpenCL workload.

The broader benchmark context complicates the picture. The Quadro 6000's average benchmark score is 9,846, which is its single Geekbench OpenCL result. The GTX 660's average is 9,022, pulled down by its Geekbench Metal score of 4,305 and Geekbench Vulkan score of 11,415. The OpenCL result is the only test both cards share, and the GTX 660 wins it clearly.

Looking at nearest rivals reinforces the Quadro's standing. The Quadro 6000 is 0.1% behind the Quadro M2000M (9,832) and 0.4% ahead of the AMD FirePro W5000 (9,803). It is 0.7% ahead of the GeForce GTX 1070 (9,780). The GTX 660, meanwhile, is 0.2% behind the TITAN V CEO Edition (9,037) and 0.4% behind the GeForce GTX 560 (9,058). It is 1.2% ahead of the AMD Radeon 550X (8,918) and 1.8% ahead of the Radeon Pro WX 5100 (8,863).

The raw throughput numbers amplify the GTX 660's advantage in certain workloads. Its FP32 output of 1.981 TFLOPS is roughly 93% higher than the Quadro's 1,027.7 GFLOPS. Its texture rate of 82.56 GTexel/s is more than 2.5 times the Quadro's 32.14 GTexel/s. The pixel rate also favors the GTX 660, at 20.64 GPixel/s versus 16.07 GPixel/s.

However, the Quadro counters with its ROP advantage. With 48 ROPs versus 24, the Quadro has twice the raster operation throughput capability, which typically benefits fill-rate-bound workloads at high resolutions. Its 6 GB memory capacity also gives it a clear edge in memory-hungry scenarios, even though bandwidth is essentially a tie.

The Verdict

The data points to a clear compute-performance winner: the NVIDIA GeForce GTX 660. It wins the only direct head-to-head benchmark by 13.2%, delivers nearly double the FP32 throughput, and has more than twice the texture rate. Its higher shading unit count (960 versus 448) and TMU count (80 versus 56) support this conclusion. For OpenCL compute workloads, the GTX 660 is the better card.

The Quadro 6000 is the better choice for memory-capacity-sensitive tasks. Its 6 GB frame buffer is triple the GTX 660's 2 GB, and its 48 ROPs are double. The wider 384-bit memory bus is another structural advantage, even though real-world bandwidth is nearly identical. The Quadro also has a higher average benchmark score (9,846 versus 9,022) and a higher percentile ranking (47th versus 45th), indicating more consistent performance across diverse workloads.

The efficiency argument strongly favors the GTX 660. It draws 140 W versus 204 W, needs a 300 W PSU versus 550 W, and uses a single 6-pin connector instead of 6-pin plus 8-pin. It also supports Vulkan, which the Quadro does not, and uses the faster PCIe 3.0 interface.

The release timeline matters. The Quadro 6000 launched in December 2010 with a launch MSRP of 4,399 USD. The GTX 660 launched nearly two years later, in September 2012, at a launch MSRP of 229 USD. The GTX 660 is the newer, more efficient, and faster card in the direct comparison. The Quadro 6000 retains value only in scenarios where 6 GB of memory and 48 ROPs are decisive — a narrow niche in the face of the GTX 660's comprehensive compute and efficiency advantages.

For general-purpose compute and modern API support, the GTX 660 is the data-backed pick. For memory-bound professional workloads requiring large frame buffers and high ROP counts, the Quadro 6000 has a specific but limited case. The benchmark results otherwise favor the GTX 660 decisively.