AMD Radeon HD 7970M vs NVIDIA GeForce GTX 690 Comparison

The NVIDIA GeForce GTX 690 and AMD Radeon HD 7970M are two very different products from the same era, yet benchmark data places them in a statistical tie. The GeForce GTX 690 is a dual-GPU desktop behemoth, while the Radeon HD 7970M is a mobile chip for laptops. Despite their different physical natures, their average benchmark scores are nearly identical, making for a fascinating comparison of desktop power versus mobile efficiency.

Head-to-Head Benchmarks

The single head-to-head benchmark available is Geekbench OpenCL, and the results are remarkably close. The NVIDIA GeForce GTX 690 scores 17,399, while the AMD Radeon HD 7970M scores 17,019. This gives the GTX 690 a 2.2% lead in this specific test. While the GTX 690 wins this particular matchup, the margin is so slim that it is effectively a tie in real-world terms.

Looking at the broader average benchmark scores, the picture becomes even murkier. The GTX 690 has an average benchmark score of 17,037, while the HD 7970M is just behind at 17,019. The deltaPct between them is only 0.1% in favor of the GTX 690. For context, the nearest rivals list for the GTX 690 shows the AMD Radeon RX 7600 XT is 0.3% ahead, and the NVIDIA GeForce RTX 3070 is 1% ahead. The HD 7970M’s rival list shows the same pattern: it is 0.4% behind the RX 7600 XT and 1.1% behind the RTX 3070. Both cards sit in the 60th percentile of all GPUs, confirming they are closely matched in the overall hierarchy.

The data suggests that in OpenCL compute workloads, these two chips are indistinguishable in practice. The GTX 690’s 2.2% lead is well within the margin of error for such benchmarks. Neither card has a decisive victory here; the numbers point to a dead heat.

Architecture Differences

The architectural divide between these two is substantial. The NVIDIA GeForce GTX 690 uses the GK104 chip, based on the Kepler architecture. It is built on a 28 nm process at TSMC, with 3,540 million transistors on a 294 mm² die. This results in a transistor density of 12.0M per mm². The card features 1,536 shading units, 128 texture mapping units (TMUs), and 32 ROPs.

In contrast, the AMD Radeon HD 7970M uses the Wimbledon chip, based on GCN 1.0. It is also built on a 28 nm process at TSMC, but with fewer transistors: 2,800 million on a smaller 212 mm² die. This gives it a higher transistor density of 13.2M per mm². The HD 7970M has 1,280 shading units, 80 TMUs, and 32 ROPs. The GTX 690 thus has 20% more shading units and 60% more TMUs, which typically translates to higher texture throughput.

Clock speeds tell the rest of the story. The GTX 690 has a base clock of 915 MHz and a boost clock of 1019 MHz. The HD 7970M has no listed base or boost clocks, relying instead on its memory clock of 1200 MHz (4.8 Gbps effective). The GTX 690’s memory runs at 1502 MHz (6 Gbps effective), giving it a memory bandwidth of 192.3 GB/s versus the HD 7970M’s 153.6 GB/s. Both cards use 2 GB of GDDR5 memory on a 256-bit bus.

The compute rates reflect the core differences. The GTX 690 achieves a pixel rate of 32.61 GPixel/s and a texture rate of 130.4 GTexel/s, while the HD 7970M manages 27.20 GPixel/s and 68.00 GTexel/s. In FP32 compute, the GTX 690 delivers 3.130 TFLOPS, significantly ahead of the HD 7970M’s 2.176 TFLOPS. The GTX 690 is clearly the more powerful silicon on paper, yet the benchmark results show they perform nearly identically.

Where Each One Wins

Based on the benchmark data, there is only one test to rely on, and the GTX 690 wins it. However, the margins are so small that the practical conclusion is that neither card has a meaningful advantage in raw compute performance. The GTX 690 wins the Geekbench OpenCL test with 17,399 versus 17,019, a 2.2% difference. It also has a higher average benchmark score of 17,037 versus 17,019, though that 0.1% gap is negligible.

Where the GTX 690 theoretically wins is in raw specifications. Its higher shading unit count, TMU count, clock speeds, and memory bandwidth all point to superior performance in texture-heavy workloads. The GPU’s 130.4 GTexel/s texture rate is nearly double the HD 7970M’s 68.00 GTexel/s, which would suggest a decisive advantage in games that rely heavily on texture fills. Similarly, the GTX 690’s 3.130 TFLOPS FP32 compute is 44% higher than the HD 7970M’s 2.176 TFLOPS.

The HD 7970M does not win any benchmark tests in the data, but it wins in a different arena: power consumption. The GTX 690 has a TDP of 300 W and requires dual 8-pin power connectors with a suggested 700 W PSU. The HD 7970M, being a mobile MXM module, has a TDP of just 100 W and requires no external power connectors. This makes the HD 7970M the clear winner for laptops and compact systems where power delivery and heat dissipation are critical constraints.

The Verdict

The data shows a near-perfect tie in benchmark scores, but the physical reality of these cards could not be more different. The NVIDIA GeForce GTX 690 is a desktop card that is 279 mm long, 111 mm tall, and 38 mm wide, occupying a dual-slot form factor. It was launched on 2012-05-02 with a launch MSRP of 999 USD. The AMD Radeon HD 7970M is an MXM module with no listed dimensions, designed to be portable device dependent.

If you are building a desktop system with ample power and space, the GTX 690 is the pick. It offers a 2.2% benchmark lead, a higher texture rate, and more compute throughput. Its 300 W TDP is a non-issue when you have a 700 W PSU and a full tower case. The card supports PCIe 3.0 x16 and has display outputs including 3x DVI and 1x mini-DisplayPort 1.2.

If you need a mobile solution or a low-power embedded system, the HD 7970M is the only choice. Its 100 W TDP and MXM-B (3.0) interface make it suitable for laptops where the GTX 690 would be physically impossible to install. The HD 7970M’s performance is within 0.1% of the GTX 690’s average, so you sacrifice almost nothing in compute performance while gaining massive efficiency improvements.

FAQ

Q: How close are the GTX 690 and HD 7970M in benchmark performance?

A: The GTX 690 has an average benchmark score of 17,037, while the HD 7970M scores 17,019, a difference of only 0.1%. In the Geekbench OpenCL test, the GTX 690 wins 17,399 to 17,019, a 2.2% margin.

Q: Which card has more shading units?

A: The NVIDIA GeForce GTX 690 has 1,536 shading units, while the AMD Radeon HD 7970M has 1,280, giving the GTX 690 a 20% advantage.

Q: What is the memory bandwidth difference?

A: The GTX 690 has a memory bandwidth of 192.3 GB/s, while the HD 7970M has 153.6 GB/s. Both use 2 GB of GDDR5 memory on a 256-bit bus.

Q: What is the power consumption comparison?

A: The GTX 690 has a TDP of 300 W and requires dual 8-pin connectors with a suggested 700 W PSU. The HD 7970M has a TDP of 100 W and requires no power connectors.

Q: Are these cards in the same performance percentile?

A: Yes, both the GTX 690 and HD 7970M are in the 60th percentile of all GPUs, confirming they are closely matched overall.

Q: What are the process nodes for each chip?

A: Both the GTX 690 (GK104) and HD 7970M (Wimbledon) are built on a 28 nm process at TSMC.

Specification Differences

| Specification | NVIDIA GeForce GTX 690 | AMD Radeon HD 7970M |

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

| Chip | GK104 | Wimbledon |

| Architecture | Kepler | GCN 1.0 |

| Transistors | 3,540 million | 2,800 million |

| Die Size | 294 mm² | 212 mm² |

| Transistor Density | 12.0M / mm² | 13.2M / mm² |

| Base Clock | 915 MHz | None |

| Boost Clock | 1019 MHz | None |

| Memory Clock | 1502 MHz (6 Gbps) | 1200 MHz (4.8 Gbps) |

| Memory Bandwidth | 192.3 GB/s | 153.6 GB/s |

| Shading Units | 1536 | 1280 |

| TMUs | 128 | 80 |

| ROPs | 32 | 32 |

| Pixel Rate | 32.61 GPixel/s | 27.20 GPixel/s |

| Texture Rate | 130.4 GTexel/s | 68.00 GTexel/s |

| FP32 | 3.130 TFLOPS | 2.176 TFLOPS |

| TDP | 300 W | 100 W |

| Slot Width | Dual-slot | MXM Module |

| Power Connectors | 2x 8-pin | None |

| Suggested PSU | 700 W | None |

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

| Display Outputs | 3x DVI, 1x mini-DisplayPort 1.2 | Portable Device Dependent |

| Dimensions | 279 mm x 111 mm x 38 mm | None |

| Release Date | 2012-05-02 | 2012-04-23 |

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

| Vulkan Support | 1.2.175 | 1.2.170 |