AMD Radeon HD 8690M vs NVIDIA GeForce GTX 560 SE Comparison

The Verdict

The benchmark data presents a clear split between these two end-of-life mobile and desktop graphics solutions. The NVIDIA GeForce GTX 560 SE is the outright performance winner, taking the single recorded head-to-head benchmark, Geekbench OpenCL, with a score of 7171 against the AMD Radeon HD 8690M's 6137. That is a 16.8% advantage. The data shows no test where the AMD card wins.

The GTX 560 SE also sits in a higher overall position, at the 39th percentile of all GPUs in the database, while the HD 8690M rests at the 36th percentile. This suggests that, in the aggregate of all recorded scores, the NVIDIA part is more capable. Its nearest rivals include the AMD Radeon Vega 8 Mobile (score 7203, 0.4% higher) and the NVIDIA GeForce GTX 750 (score 7222, 0.7% higher), placing it in the company of entry-level desktop parts. The AMD Radeon HD 8690M, by contrast, is compared to the Intel Iris Pro Graphics 6200 (score 6117, 0.3% lower) and the NVIDIA GeForce MX230 (score 6077, 1% lower), placing it in a lower performance tier.

For a user strictly prioritizing compute performance as measured by this benchmark, the NVIDIA GeForce GTX 560 SE is the only rational choice. It offers a significant lead in raw OpenCL throughput. The AMD Radeon HD 8690M, while not without merit, trails by a considerable margin. The data does not indicate that the AMD card wins in any scenario, so the verdict is one-sided: pick the GTX 560 SE for the higher score. The HD 8690M could be considered for a system where it is the only option, but there is no performance-based reason to prefer it based on the numbers.

Architecture Differences

The two GPUs come from completely different design schools. The NVIDIA GeForce GTX 560 SE uses the GF114 chip, built on the Fermi 2.0 architecture, and fabricated on a 40 nm process at TSMC. This is a comparatively older, larger design. The chip holds 1,950 million transistors on a 332 mm² die, giving it a transistor density of 5.9 million per square millimeter. The AMD Radeon HD 8690M uses the Sun chip, based on the Graphics Core Next (GCN) 1.0 architecture, and is built on a more modern 28 nm process at the same foundry. This chip is far smaller, with 690 million transistors on a 56 mm² die, resulting in a higher density of 12.3 million transistors per square millimeter.

The memory subsystems differ in capacity and bandwidth. The NVIDIA card has 1024 MB of GDDR5 on a 192-bit bus, achieving a bandwidth of 91.87 GB/s. The AMD card has 2 GB of GDDR5, but on a 64-bit bus, resulting in a bandwidth of just 32.00 GB/s. The NVIDIA card also has a higher memory clock at 3.8 Gbps effective versus the AMD's 4 Gbps effective, though the bus width is the dominant factor in the bandwidth difference.

The internal resource distribution is also distinct. The NVIDIA GTX 560 SE has 288 shading units, 48 texture mapping units (TMUs), and 24 render output units (ROPs). The AMD Radeon HD 8690M has 320 shading units, but only 20 TMUs and 8 ROPs. This means the AMD has more compute-like shader cores but fewer of the specialized units for texture and pixel output.

The feature sets reflect their respective eras. The NVIDIA card supports DirectX 12 (11_0) and OpenGL 4.6, with no Vulkan support listed. The AMD card supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The NVIDIA card is a desktop part with a dual-slot design and two 6-pin power connectors, while the AMD card is a mobile part with no listed slot width, power connectors, or display outputs. The NVIDIA connects via PCIe 2.0 x16, while the AMD uses PCIe 3.0 x8. The production status for both is end-of-life.

Where Each One Wins

The recorded data gives only one benchmark test: Geekbench OpenCL. In that test, the NVIDIA GeForce GTX 560 SE is the definitive winner. It scores 7171 points, which is 16.8% higher than the AMD Radeon HD 8690M's 6137. Consequently, for any application that relies on the OpenCL compute performance as measured by this test, the NVIDIA card is the one to pick.

The AMD Radeon HD 8690M has no wins in the head-to-head data. There are no other tests in the database where it beats the NVIDIA part. Therefore, there are no use cases where the data supports a clear AMD victory. The AMD card's only advantage in the specification list is its larger 2 GB memory capacity and its support for Vulkan 1.2.170, which the NVIDIA card lacks. The data does not show a benchmark for Vulkan, so this feature difference cannot be quantified in terms of performance.

The NVIDIA card's strengths are its much higher fill rates. Its pixel rate is 8.832 GPixel/s and its texture rate is 35.33 GTexel/s. The AMD card is slower in both, with 7.800 GPixel/s and 19.50 GTexel/s. This suggests the NVIDIA card would be better at rendering operations that are heavily dependent on pixel and texture throughput. The AMD card's higher FP32 throughput of 624.0 GFLOPS is actually lower than the NVIDIA's 847.9 GFLOPS, so even in raw compute, the NVIDIA part leads.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GeForce GTX 560 SE has a much higher score of 7171, while the AMD Radeon HD 8690M scores 6137.

Q: What is the performance difference between the two in the head-to-head test?

A: The NVIDIA GeForce GTX 560 SE is 16.8% ahead of the AMD Radeon HD 8690M in the Geekbench OpenCL test.

Q: Which GPU has more shading units?

A: The AMD Radeon HD 8690M has more shading units, with 320, while the NVIDIA GeForce GTX 560 SE has 288.

Q: Which GPU offers a wider memory bus?

A: The NVIDIA GeForce GTX 560 SE has a 192-bit memory bus, while the AMD Radeon HD 8690M has a 64-bit bus.

Q: Does either GPU support Vulkan?

A: The AMD Radeon HD 8690M supports Vulkan 1.2.170, while the NVIDIA GeForce GTX 560 SE has no Vulkan support listed in the database.

Q: Which GPU has a higher transistor density?

A: The AMD Radeon HD 8690M has a higher transistor density at 12.3 million per square millimeter, compared to the NVIDIA GeForce GTX 560 SE's 5.9 million per square millimeter.

Head-to-Head Benchmarks

The single recorded benchmark, Geekbench OpenCL, provides the only direct comparison. The NVIDIA GeForce GTX 560 SE scores 7171, and the AMD Radeon HD 8690M scores 6137. The delta is 16.8% in favor of the NVIDIA card. This is a substantial margin, placing the two cards in different performance classes.

The NVIDIA card's victory is not a small one. A 16.8% lead in OpenCL compute is a significant advantage. The database shows its nearest rivals are the AMD Radeon Vega 8 Mobile, which is just 0.4% higher, and the NVIDIA GeForce GTX 750, which is 0.7% higher. This indicates the GTX 560 SE is competing with a popular set of entry-level desktop and integrated-class parts. The AMD card, trailing at a lower score, is closer to the Intel Iris Pro Graphics 6200 (0.3% lower) and the NVIDIA RTX A400 (1% higher), as well as the NVIDIA GeForce MX230 (1% higher). This shows the AMD card is in a lower performance band.

In practical terms, the fill rate numbers back up the benchmark result. The NVIDIA card's texture rate of 35.33 GTexel/s is roughly 81% higher than the AMD's 19.50 GTexel/s. Similarly, the NVIDIA card's pixel rate is 8.832 GPixel/s versus the AMD's 7.800 GPixel/s. The NVIDIA card also has more raw FP32 compute, at 447.9 GFLOPS, which is about 36% higher than the AMD's 624.0 GFLOPS. These architectural advantages all contribute to the winning OpenCL score.

Specification Differences

The two cards differ in almost every specification category, reflecting their different design targets and releases.

• Process Node: The NVIDIA GeForce GTX 560 SE is on a 40 nm process, while the AMD Radeon HD 8690M is on a 28 nm process.
• Transistors: The NVIDIA card has 1,950 million transistors, whereas the AMD card has 690 million.
• Die Size: The NVIDIA card's die is 332 mm², while the AMD card's is 56 mm².
• Memory Size: The NVIDIA card has 1024 MB, and the AMD card has 2 GB.
• Memory Bus Width: The NVIDIA card uses a 192-bit bus, while the AMD card uses a 64-bit bus.
• Memory Bandwidth: The NVIDIA card has 91.87 GB/s, compared to the AMD card's 32.00 GB/s.
• Memory Clock: The NVIDIA card's memory is rated at 957 MHz (3.8 Gbps effective), while the AMD card's is at 1000 MHz (4 Gbps effective).
• Shading Units: The NVIDIA card has 288, and the AMD card has 320.
• Texture Mapping Units: The NVIDIA card has 48, and the AMD card has 20.
• Render Output Units: The NVIDIA card has 24, and the AMD card has 8.
• Pixel Rate: The NVIDIA card has 8.832 GPixel/s, and the AMD card has 7.800 GPixel/s.
• Texture Rate: The NVIDIA card has 35.33 GTexel/s, and the AMD card has 19.50 GTexel/s.
• FP32 Performance: The NVIDIA card has 847.9 GFLOPS, and the AMD card has 624.0 GFLOPS.
• TDP: The NVIDIA card is rated at 150 W, while the AMD card has no TDP listed.
• Slot Width: The NVIDIA card is a dual-slot design, while the AMD card has no slot width listed.
• Power Connectors: The NVIDIA card requires 2x 6-pin connectors, while the AMD card has none listed.
• Bus Interface: The NVIDIA card uses PCIe 2.0 x16, and the AMD card uses PCIe 3.0 x8.
• Vulkan Support: The NVIDIA card has no Vulkan support listed, while the AMD card supports Vulkan 1.2.170.
• DirectX Support: The NVIDIA card supports DirectX 12 (11_0), while the AMD card supports DirectX 12 (11_1).
• Release Date: The NVIDIA card was released in February 2012, and the AMD card was released in February 2013.