NVIDIA GeForce GTX 690 vs NVIDIA GeForce RTX 2060 Comparison

Head-to-Head Benchmarks

The recorded data shows a clear and decisive performance gap between these two generations. In the two shared benchmark tests, the GeForce RTX 2060 wins both, and it does so by a staggering margin. In Geekbench OpenCL, the RTX 2060 scores 65,014 points against the GTX 690's 17,399. This represents a 73.2% higher score for the newer card, meaning the RTX 2060 delivers roughly four times the compute performance in this workload. The Vulkan test tells a similar story, with the RTX 2060 scoring 65,846 versus 16,675 for the GTX 690, a 74.7% advantage.

These are not marginal wins; they are generational leaps. The GTX 690's best result, its OpenCL score of 17,399, still falls short of the RTX 2060's lowest recorded benchmark, which is the Passmark DirectX 12 score of 53. This suggests that even in API-specific legacy tests, the newer architecture's raw capabilities translate into a fundamentally different performance tier. The head-to-head data records zero wins for the GTX 690 and two for the RTX 2060.

The average benchmark score in the database further illustrates the divide. The GTX 690 averages 17,037 points, while the RTX 2060 averages 15,290. Interestingly, the GTX 690 has a higher average score despite losing both direct comparisons. This is because the GTX 690's average is derived from only two high-scoring Geekbench tests, whereas the RTX 2060's average includes several Passmark tests with lower scores, such as 98 for DirectX 10 and 190 for DirectX 9. When looking strictly at the head-to-head, the RTX 2060 is the victor, but the aggregate data reveals that the GTX 690's limited test suite skews its overall standing.

Architecture Differences

The fundamental architectural split explains the benchmark results. The GTX 690 is built on the Kepler architecture with the GK104 chip, manufactured on a 28 nm process at TSMC. It packs 3,540 million transistors into a 294 mm² die, yielding a transistor density of 12.0M per mm². The RTX 2060, by contrast, uses the Turing architecture with the TU106 chip, built on a 12 nm process. It contains 10,800 million transistors across a 445 mm² die, achieving a density of 24.3M per mm². The newer process allows for nearly double the transistor density, which is foundational to the performance difference.

Memory configurations also diverge sharply. The GTX 690 comes with 2 GB of GDDR5 memory on a 256-bit bus, delivering 192.3 GB/s of bandwidth. The RTX 2060 offers 6 GB of GDDR6 memory on a 192-bit bus, but its bandwidth reaches 336.0 GB/s. The newer memory type and higher effective clock speed, 14 Gbps versus 6 Gbps, more than compensate for the narrower bus. This gives the RTX 2060 significantly more memory headroom and bandwidth, which is critical for modern workloads and higher resolutions.

Compute resources tell a similar story. The GTX 690 has 1,536 shading units, 128 texture mapping units, and 32 ROPs. The RTX 2060 has 1,920 shading units, 120 TMUs, and 48 ROPs. While the GTX 690 has more TMUs, the RTX 2060 compensates with higher clocks and more ROPs. The pixel rate for the RTX 2060 is 80.64 GPixel/s versus 32.61 GPixel/s for the GTX 690, and the texture rate is 201.6 GTexel/s versus 130.4 GTexel/s. The FP32 performance is 6.451 TFLOPS for the RTX 2060 versus 3.130 TFLOPS for the GTX 690, more than doubling the raw compute throughput.

The RTX 2060 also introduces hardware features that did not exist in the Kepler generation. It includes 30 RT cores and 240 tensor cores, enabling real-time ray tracing and AI-accelerated workloads. The GTX 690 has no such dedicated hardware. The RTX 2060 also supports DirectX 12 Ultimate (12_2), while the GTX 690 is limited to DirectX 12 (11_0). Vulkan support is newer on the RTX 2060, at version 1.4 versus 1.2.175. These are not just clock speed differences; they represent entirely different feature sets.

FAQ

Q: Which card has more shading units?

A: The RTX 2060 has 1,920 shading units, while the GTX 690 has 1,536. This gives the RTX 2060 a 25% advantage in raw shader count.

Q: How do their memory bandwidths compare?

A: The RTX 2060 achieves 336.0 GB/s with 6 GB of GDDR6, whereas the GTX 690 provides 192.3 GB/s with 2 GB of GDDR5. The RTX 2060 has 75% more bandwidth.

Q: Does the GTX 690 support ray tracing?

A: No. The GTX 690 has no RT cores. The RTX 2060 includes 30 RT cores specifically for ray tracing workloads.

Q: What is the transistor density difference?

A: The RTX 2060 has a density of 24.3M transistors per mm², while the GTX 690 has 12.0M per mm². The newer 12 nm process enables this doubling.

Q: Which card has a higher boost clock?

A: The RTX 2060 boosts to 1680 MHz, while the GTX 690 boosts to 1019 MHz. The RTX 2060 runs 65% faster at peak boost.

Q: What are their average benchmark scores?

A: The GTX 690 averages 17,037 points, and the RTX 2060 averages 15,290 points. However, in head-to-head tests, the RTX 2060 wins both by margins exceeding 73%.

The Verdict

The data indicates that the RTX 2060 is the superior card for virtually any modern workload. Its 73.2% lead in OpenCL and 74.7% lead in Vulkan are decisive. The RTX 2060 also offers more memory, higher bandwidth, and dedicated RT and tensor cores. The GTX 690's only statistical advantage is its higher average benchmark score, but this is an artifact of its limited two-test benchmark suite. When directly compared, the GTX 690 loses by a wide margin in both recorded tests.

The RTX 2060's percentile rank of 58 places it slightly below the GTX 690's rank of 60 among all GPUs, but this does not reflect their head-to-head performance. The percentile is based on average scores, where the GTX 690's two strong Geekbench results inflate its standing. In practical terms, the RTX 2060 is the card to choose for anyone running modern games or compute tasks. The GTX 690 remains an interesting piece of 2012 hardware, but its performance is firmly in the past.

Specification Differences

| Specification | NVIDIA GeForce GTX 690 | NVIDIA GeForce RTX 2060 |

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

| Architecture | Kepler | Turing |

| Chip | GK104 | TU106 |

| Process Node | 28 nm | 12 nm |

| Transistors | 3,540 million | 10,800 million |

| Die Size | 294 mm² | 445 mm² |

| Transistor Density | 12.0M / mm² | 24.3M / mm² |

| Base Clock | 915 MHz | 1365 MHz |

| Boost Clock | 1019 MHz | 1680 MHz |

| Memory Size | 2 GB | 6 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus | 256 bit | 192 bit |

| Memory Bandwidth | 192.3 GB/s | 336.0 GB/s |

| Shading Units | 1536 | 1920 |

| TMUs | 128 | 120 |

| ROPs | 32 | 48 |

| RT Cores | None | 30 |

| Tensor Cores | None | 240 |

| Pixel Rate | 32.61 GPixel/s | 80.64 GPixel/s |

| Texture Rate | 130.4 GTexel/s | 201.6 GTexel/s |

| FP32 Performance | 3.130 TFLOPS | 6.451 TFLOPS |

| FP16 Performance | None | 12.90 TFLOPS (2:1) |

| TDP | 300 W | 160 W |

| Power Connectors | 2x 8-pin | 1x 8-pin |

| Suggested PSU | 700 W | 450 W |

| DirectX Support | 12 (11_0) | 12 Ultimate (12_2) |

| Vulkan Support | 1.2.175 | 1.4 |

| Display Outputs | 3x DVI, 1x mini-DisplayPort 1.2 | 1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, 1x USB Type-C |

| Release Date | 2012-05-02 | 2019-01-06 |

| Launch MSRP | 999 USD | 349 USD |

Where Each One Wins

The RTX 2060 wins in every direct benchmark comparison, but the GTX 690 still holds a few statistical claims. The GTX 690's average benchmark score of 17,037 is higher than the RTX 2060's 15,290. This is a quirk of the data, but it does indicate that in the specific OpenCL and Vulkan tests it ran, the older card performed competently for its era. The GTX 690 also has a higher percentile rank at 60 versus 58, meaning it sits above more GPUs in the overall distribution. For legacy applications that rely on its specific feature set, such as older DirectX 11 workloads, the GTX 690 remains functional.

The RTX 2060, however, is the clear choice for modern use. Its OpenCL score of 65,014 and Vulkan score of 65,846 are more than three times higher than the GTX 690's corresponding results. The RTX 2060's Passmark scores, while lower than its Geekbench results, still show strength in compute with a GPU compute score of 5,488. The RTX 2060's 6 GB memory capacity is triple the GTX 690's 2 GB, and its 336.0 GB/s bandwidth is 75% higher. For ray tracing, AI workloads, and high-resolution gaming, the RTX 2060 is the only viable option between these two. The GTX 690's dual-GPU design and 300 W power draw are relics of a different era, while the RTX 2060 delivers more performance at nearly half the power consumption.