GPU Comparison

The AMD Radeon HD 8690M and the NVIDIA GeForce 930A are both end-of-life mobile graphics processors aimed at thin-and-light laptops, but they deliver measurably different performance profiles. The recorded benchmark data shows a clear winner in raw compute, though each GPU exhibits distinct strengths in memory technology, feature support, and architectural efficiency. This analysis draws exclusively from the database measurements to break down where these two chips stand.

Head-to-Head Benchmarks

The only directly comparable benchmark in the database is the Geekbench OpenCL test, and it is decisively won by the AMD Radeon HD 8690M. The AMD part scores 6,137 points, while the NVIDIA GeForce 930A trails with 5,317 points. That is a 15.4% delta, meaning the AMD card is roughly 15% faster in this synthetic compute workload. In practical terms, the HD 8690M completes the same OpenCL tasks noticeably quicker, which matters for applications that offload parallel math to the GPU.

The percentile rankings reinforce this gap. The AMD Radeon HD 8690M sits at the 36th percentile among all GPUs in the database, while the NVIDIA GeForce 930A lands at the 31st percentile. That five-point percentile difference indicates the AMD part outranks a larger share of the overall GPU landscape, not just its direct competitor here.

Looking at the nearest rivals for each card clarifies the competitive picture. The AMD Radeon HD 8690M is nearly tied with the Intel Iris Pro Graphics 6200, which averages 6,117 points, a mere 0.3% difference. It is also 1% ahead of the NVIDIA RTX A400 and the NVIDIA GeForce MX230, both scoring around 6,077 to 6,078. However, it is 1.2% slower than the AMD Radeon HD 6950, which scores 6,210. This places the HD 8690M in a tight cluster where small margins separate it from several contemporaries.

The NVIDIA GeForce 930A has a different rival profile. It is virtually identical to the GeForce 840M, which scores 5,322, a -0.1% delta. It is 0.2% ahead of the GeForce GTX 980M (5,308) and 0.6% ahead of the GeForce 940M (5,284). However, it is 0.8% behind the AMD Radeon R7 M445, which scores 5,358. The 930A thus sits near the bottom of its peer group, while the HD 8690M occupies a slightly higher performance tier.

The head-to-head result is unambiguous. The AMD card wins the only recorded benchmark, and the margin is not trivial. A 15.4% advantage in OpenCL compute translates to noticeably faster performance in any workload that leverages the GPU for general-purpose calculations. This is the largest single win in the comparison.

Where Each One Wins

The AMD Radeon HD 8690M wins in raw compute performance. Its Geekbench OpenCL score of 6,137 is 15.4% higher than the NVIDIA's 5,317. This makes the AMD the better choice for tasks like video encoding, physics simulations, or any software that offloads parallel arithmetic to the GPU. The higher pixel rate also favors the AMD: 7.800 GPixel/s versus 7.528 GPixel/s, a small but real fill-rate advantage for higher-resolution 2D workloads.

However, the NVIDIA GeForce 930A does not come away empty-handed. It wins in texture throughput, posting a texture rate of 22.58 GTexel/s versus the AMD's 19.50 GTexel/s. This is a 15.8% advantage, which can matter in games or applications that are heavily texture-bound, where fetching and filtering textures dominates the rendering pipeline. The NVIDIA also has a higher FP32 compute rating at 722.7 GFLOPS versus 624.0 GFLOPS, a 15.8% lead in raw single-precision floating-point math. This suggests the Maxwell architecture can execute more math operations per second, even if the OpenCL benchmark tells a different story due to real-world memory or driver overhead.

In terms of memory, the AMD has a clear bandwidth advantage. It uses GDDR5 memory with an effective data rate of 4 Gbps, yielding 32.00 GB/s across its 64-bit bus. The NVIDIA uses DDR3 with an effective rate of 2 Gbps, yielding only 16.02 GB/s. That is exactly half the bandwidth, a severe handicap for memory-intensive workloads. The AMD's higher bandwidth is likely a key reason for its benchmark victory. The NVIDIA compensates with more shading units (384 versus 320) and more texture mapping units (24 versus 20), but it cannot overcome the memory deficit in the recorded test.

For users who prioritize compute or memory-heavy tasks, the AMD is the stronger pick. For users who favor texture throughput or raw FP32 math, the NVIDIA has a theoretical edge, though its lower memory bandwidth may throttle those capabilities in practice. The data shows no other benchmark where the NVIDIA wins outright, so its strengths are purely architectural rather than validated by a database score.

Architecture Differences

The two GPUs come from different architectural eras and design philosophies. The AMD Radeon HD 8690M is built on GCN 1.0, using the Sun chip, fabricated by TSMC on a 28 nm process. The NVIDIA GeForce 930A uses the Maxwell architecture, with the GM108 chip, also on TSMC's 28 nm process. Both are 28 nm parts, so process node is identical, but the transistor counts differ significantly.

The AMD chip packs 690 million transistors into a die size of 56 mm², resulting in a transistor density of 12.3 million per square millimeter. The NVIDIA chip has 1,020 million transistors across a 77 mm² die, giving it a density of 13.2 million per square millimeter. The NVIDIA chip is physically larger and contains nearly 48% more transistors, which explains its higher shading unit count and texture unit count.

The memory subsystems diverge sharply. The AMD uses 2 GB of GDDR5 with a 64-bit bus, an effective speed of 4 Gbps, and a bandwidth of 32.00 GB/s. The NVIDIA also uses 2 GB but of DDR3, with the same 64-bit bus, an effective speed of 2 Gbps, and a bandwidth of 16.02 GB/s. This is a critical difference: the AMD has exactly double the memory bandwidth, which is a major factor in compute tasks that stream data. The NVIDIA's DDR3 is lower-power but slower.

Clock speeds are near-identical, despite the architectural differences. The AMD has a base clock of 925 MHz and a boost of 975 MHz. The NVIDIA has a base of 928 MHz and a boost of 941 MHz. The AMD boots 34 MHz higher, which helps its compute performance. Memory clocks differ: the AMD runs at 1000 MHz (4 Gbps effective), while the NVIDIA runs at 1001 MHz (2 Gbps effective). The effective rate is what matters, and the AMD is twice as fast.

The shading resources favor the NVIDIA. The 930A has 384 shading units, 24 texture mapping units, and 8 ROPs. The AMD has 320 shading units, 20 texture mapping units, and 8 ROPs. The NVIDIA thus has 20% more shading units and 20% more TMUs, but the same ROP count. The pixel rate is close: 7.800 GPixel/s for the AMD versus 7.528 GPixel/s for the NVIDIA, because the AMD's higher boost clock offsets its fewer ROPs. The texture rate favors the NVIDIA: 22.58 GTexel/s versus 19.50 GTexel/s.

The FP32 compute is higher on the NVIDIA at 722.7 GFLOPS versus 624.0 GFLOPS. This is a pure arithmetic throughput number that does not account for memory bottlenecks. The AMD has no FP16 rating, and the NVIDIA also lists none, so that is a non-issue.

The NVIDIA has a TDP of 33 W, while the AMD has no listed TDP in the database, so a direct power comparison is not possible. The NVIDIA is listed as an integrated GPU (IGP) with no power connectors, while the AMD has no slot width or power connector data. The NVIDIA's display outputs are "portable device dependent," meaning they vary by laptop, and the AMD has no display output information listed.

API support is similar but not identical. Both support DirectX 12, but the AMD's version is 12 (11_1) while the NVIDIA is 12 (11_0). The AMD's DirectX feature level is slightly higher. Both support OpenGL 4.6. For Vulkan, the AMD supports 1.2.170, while the NVIDIA supports 1.4, giving the NVIDIA a newer Vulkan API version.

The bus interface is the same for both: PCIe 3.0 x8, which is a narrower link than the typical x16 but sufficient for these low-power parts.

FAQ

Q: Which GPU is faster in compute workloads?

A: The AMD Radeon HD 8690M has a Geekbench OpenCL score of 6,137, which is 15.4% higher than the NVIDIA GeForce 930A's 5,317. This makes the AMD the faster choice for OpenCL-based compute tasks.

Q: Does the NVIDIA GeForce 930A have any performance advantage?

A: Yes, the NVIDIA has a higher texture rate (22.58 GTexel/s versus 19.50 GTexel/s) and higher FP32 compute (722.7 GFLOPS versus 624.0 GFLOPS). However, no benchmark in the database confirms that these translate to a win.

Q: What is the main memory difference between the two?

A: The AMD uses 2 GB of GDDR5 with a 64-bit bus and bandwidth of 32.00 GB/s, while the NVIDIA uses 2 GB of DDR3 with the same 64-bit bus but bandwidth of 16.02 GB/s. The AMD has double the memory bandwidth.

Q: How does each GPU compare to its nearest rivals?

A: The AMD is 0.3% ahead of the Intel Iris Pro Graphics 6200 and 1% ahead of the NVIDIA RTX A400 and GeForce MX230, but 1.2% behind the AMD Radeon HD 6950. The NVIDIA is 0.1% behind the GeForce 840M and 0.2% ahead of the GeForce GTX 980M, but 0.8% behind the AMD Radeon R7 M445.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce 930A has 384 shading units, while the AMD Radeon HD 8690M has 320. The NVIDIA also has 24 texture mapping units versus the AMD's 20.

Q: What are the architectural differences?

A: The AMD uses GCN 1.0 with the Sun chip, while the NVIDIA uses Maxwell with the GM108 chip. Both are 28 nm, but the NVIDIA has more transistors (1,020 million versus 690 million) and a larger die size (77 mm² versus 56 mm²). The NVIDIA supports Vulkan 1.4, while the AMD supports Vulkan 1.2.170.

Specification Differences

The table below lists only the fields where the two GPUs differ, based on the database records.

| Specification | AMD Radeon HD 8690M | NVIDIA GeForce 930A |

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

| Architecture | GCN 1.0 | Maxwell |

| Chip | Sun | GM108 |

| Generation | Solar System (HD 8600M) | GeForce 900A |

| Transistors | 690 million | 1,020 million |

| Die Size | 56 mm² | 77 mm² |

| Transistor Density | 12.3M / mm² | 13.2M / mm² |

| Base Clock | 925 MHz | 928 MHz |

| Boost Clock | 975 MHz | 941 MHz |

| Memory Clock | 1000 MHz, 4 Gbps effective | 1001 MHz, 2 Gbps effective |

| Memory Type | GDDR5 | DDR3 |

| Bandwidth | 32.00 GB/s | 16.02 GB/s |

| Shading Units | 320 | 384 |

| TMUs | 20 | 24 |

| Pixel Rate | 7.800 GPixel/s | 7.528 GPixel/s |

| Texture Rate | 19.50 GTexel/s | 22.58 GTexel/s |

| FP32 | 624.0 GFLOPS | 722.7 GFLOPS |

| TDP | Not listed | 33 W |

| Slot Width | Not listed | IGP |

| Power Connectors | Not listed | None |

| Display Outputs | Not listed | Portable Device Dependent |

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

| Vulkan | 1.2.170 | 1.4 |

| Release Date | 2013-02-28 | 2015-03-12 |

| Predecessor | London | GeForce 800A |

| Successor | Gem System | Not listed |

The two cards share the same process node (28 nm), memory size (2 GB), bus width (64 bit), bus interface (PCIe 3.0 x8), and the same number of ROPs (8). They also both lack ray tracing and tensor cores, and both are end-of-life products. The GeForce 930A launched roughly two years after the Radeon HD 8690M, but the AMD still comes out on top in the only recorded benchmark, demonstrating that architectural efficiency and memory bandwidth can outweigh transistor count and newer design.