AMD Radeon HD 8690M vs NVIDIA GeForce GT 555M Comparison

The NVIDIA GeForce GT 555M and the AMD Radeon HD 8690M are two end-of-life mobile graphics processors from different architectural eras. The benchmark data presents a close contest in raw compute, with the GT 555M edging ahead in the sole Geekbench OpenCL test, yet the underlying specifications tell a story of two very different design philosophies. This analysis breaks down where each GPU holds an advantage, the architectural chasm between them, and what the numbers ultimately imply for a potential user.

Where Each One Wins

The head-to-head benchmark data is decisively one-sided in terms of victory count, with the NVIDIA GeForce GT 555M taking the single available test. In the Geekbench OpenCL workload, the GT 555M scores 6493 against the HD 8690M's 6137, a delta of 5.8%. This makes the NVIDIA part the winner in the only directly comparable metric, suggesting a lead in general-purpose compute tasks that leverage OpenCL. However, this single-data-point victory should be viewed in context; the GT 555M's score places it at the 37th percentile of all GPUs, while the HD 8690M sits just one point lower at the 36th percentile. The performance gap is real but narrow, placing both firmly in the same performance tier for this workload.

The AMD Radeon HD 8690M, while losing the benchmark, wins on architectural efficiency and feature set. Its 28 nm process node and higher transistor density (12.3M / mm² vs. 4.9M / mm²) indicate a more modern, power-savvy design. The HD 8690M also wins on memory technology, employing faster GDDR5 memory compared to the GT 555M's DDR3. While the GT 555M has a wider 128-bit memory bus, the HD 8690M's effective 4 Gbps memory speed results in a higher total bandwidth of 32.00 GB/s versus 28.80 GB/s. For tasks sensitive to memory bandwidth, the AMD part holds a theoretical edge, even if the OpenCL score doesn't reflect it. The HD 8690M also counters with superior pixel and texture fill rates (7.800 GPixel/s and 19.50 GTexel/s, respectively) compared to the GT 555M's 3.540 GPixel/s and 14.16 GTexel/s, which could translate to better performance in specific rasterization-heavy scenarios.

Architecture Differences

The most profound difference lies in the underlying architectures. The NVIDIA GeForce GT 555M is built on the Fermi architecture using the GF106 chip, manufactured on a 40 nm process at TSMC. This is a larger, older design with 1,170 million transistors on a 238 mm² die. In contrast, the AMD Radeon HD 8690M utilizes the Graphics Core Next (GCN) 1.0 architecture with the "Sun" chip, built on a more advanced 28 nm process. This allows AMD to pack 690 million transistors onto a much smaller 56 mm² die, resulting in superior transistor density.

These architectural choices lead to different compute resource allocations. The GT 555M features 144 shading units, 24 texture mapping units (TMUs), and 16 raster output units (ROPs). The HD 8690M, while having fewer TMUs (20) and ROPs (8), boasts a significantly higher number of shading units at 320. This massive disparity in shader count explains the AMD part's higher FP32 performance of 624.0 GFLOPS, nearly double the GT 555M's 339.8 GFLOPS. The NVIDIA GPU's higher ROP count and wider memory bus suggest a design focused on traditional rendering throughput, whereas AMD's GCN architecture leans towards parallel compute. The memory subsystems also differ fundamentally: the GT 555M uses 1024 MB of DDR3 on a 128-bit bus, while the HD 8690M uses 2 GB of GDDR5 on a 64-bit bus. Finally, the bus interface differs, with NVIDIA using PCIe 2.0 x16 and AMD using PCIe 3.0 x8.

FAQ

Q: Which GPU has the higher memory bandwidth?

A: The AMD Radeon HD 8690M, with a bandwidth of 32.00 GB/s, outperforms the NVIDIA GeForce GT 555M's 28.80 GB/s, despite the NVIDIA part having a wider 128-bit memory bus. This is due to the HD 8690M's use of faster GDDR5 memory running at an effective 4 Gbps.

Q: How does the shading unit count compare between the two?

A: The AMD Radeon HD 8690M has over double the shading units of the NVIDIA GeForce GT 555M, with 320 compared to 144. This contributes to the HD 8690M's higher theoretical FP32 performance of 624.0 GFLOPS versus 339.8 GFLOPS.

Q: Which GPU is based on a more recent manufacturing process?

A: The AMD Radeon HD 8690M is built on a 28 nm process, while the NVIDIA GeForce GT 555M uses a 40 nm process. This makes the AMD part the more modern design, allowing for a smaller die size of 56 mm² compared to 238 mm².

Q: What is the difference in their DirectX support?

A: The AMD Radeon HD 8690M supports DirectX 12 (11_1), while the NVIDIA GeForce GT 555M supports DirectX 12 (11_0). The HD 8690M also has Vulkan support at version 1.2.170, whereas the GT 555M has no listed Vulkan support.

Q: Which GPU has a higher pixel fill rate?

A: The AMD Radeon HD 8690M has a higher pixel fill rate of 7.800 GPixel/s, which is more than double the NVIDIA GeForce GT 555M's 3.540 GPixel/s. This is despite the GT 555M having double the ROPs (16 vs. 8).

Specification Differences

The following table highlights the key specifications where the NVIDIA GeForce GT 555M and AMD Radeon HD 8690M differ:

| Specification | NVIDIA GeForce GT 555M | AMD Radeon HD 8690M |

| :--- | :--- | :--- |

| Architecture | Fermi | GCN 1.0 |

| Process Node | 40 nm | 28 nm |

| Transistors | 1,170 million | 690 million |

| Die Size | 238 mm² | 56 mm² |

| Transistor Density | 4.9M / mm² | 12.3M / mm² |

| Base Clock | N/A | 925 MHz |

| Boost Clock | N/A | 975 MHz |

| Memory Clock | 1800 Mbps effective | 4 Gbps effective |

| Memory Size | 1024 MB | 2 GB |

| Memory Type | DDR3 | GDDR5 |

| Memory Bus Width | 128 bit | 64 bit |

| Memory Bandwidth | 28.80 GB/s | 32.00 GB/s |

| Shading Units | 144 | 320 |

| TMUs | 24 | 20 |

| ROPs | 16 | 8 |

| Pixel Rate | 3.540 GPixel/s | 7.800 GPixel/s |

| Texture Rate | 14.16 GTexel/s | 19.50 GTexel/s |

| FP32 Performance | 339.8 GFLOPS | 624.0 GFLOPS |

| TDP | 35 W | N/A |

| Bus Interface | PCIe 2.0 x16 | PCIe 3.0 x8 |

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

| Vulkan Support | N/A | 1.2.170 |

| Release Date | 2011-07-01 | 2013-02-28 |

Head-to-Head Benchmarks

The single head-to-head benchmark result is the Geekbench OpenCL test. Here, the NVIDIA GeForce GT 555M achieves a score of 6493, while the AMD Radeon HD 8690M scores 6137. This gives NVIDIA the win with a delta of 5.8%. This result is interesting because the HD 8690M has a significantly higher theoretical FP32 performance (624.0 GFLOPS vs. 339.8 GFLOPS). The data suggests that the GT 555M's wider 128-bit memory bus and higher ROP count (16 vs. 8) may provide a tangible advantage in this specific workload, or that the OpenCL test is not purely a measure of raw shader throughput. The GT 555M's nearest rivals include the NVIDIA Quadro M5000M (6481) and AMD Radeon Vega 10 Mobile (6476), with deltas of 0.2% and 0.3%, respectively, placing it in a tight cluster. The HD 8690M's closest competitors are the Intel Iris Pro Graphics 6200 (6117) and NVIDIA RTX A400 (6078), with deltas of 0.3% and 1.0%. This puts the two GPUs in different competitive neighborhoods despite their proximity to each other.

The Verdict

From the data, the NVIDIA GeForce GT 555M is the choice for users prioritizing raw OpenCL compute performance as measured by Geekbench. Its 5.8% lead over the HD 8690M, while modest, is a clear victory in the only direct comparison available. Its wider memory bus and higher ROP count suggest a design that may be better suited for standard graphics rendering tasks that rely on these features. The GT 555M also has a clearly defined TDP of 35 W, allowing for predictable power budgeting in a laptop design.

The AMD Radeon HD 8690M, however, presents a compelling case for those looking at the broader feature set and theoretical capabilities. Its GCN architecture, built on a smaller 28 nm process, is more modern and efficient. The data shows it possesses a 2x advantage in shading units and FP32 performance, along with higher fill rates and memory bandwidth. For compute-heavy applications that can leverage its massive shader array, the HD 8690M has a higher ceiling. Its support for Vulkan 1.2.170 and a newer DirectX feature level (11_1) also provides better forward-looking API compatibility. Ultimately, the GT 555M wins the benchmark, but the HD 8690M wins on architectural merit and specification superiority, making the choice dependent on whether the user values the measured OpenCL score or the theoretical and feature-based advantages of the AMD part.