AMD Radeon HD 8730M vs NVIDIA GeForce 930A Comparison

Where Each One Wins

The benchmark data splits cleanly along a single axis: the AMD Radeon HD 8730M wins the only recorded head-to-head test, the Geekbench OpenCL workload, by a 12% margin. That is the entirety of the win column for AMD, while the NVIDIA GeForce 930A has no recorded wins in the database. This does not mean the 930A is without merit; rather, the recorded measurements simply favor the older AMD part in compute throughput.

Looking at the broader context, the AMD Radeon HD 8730M sits at the 34th percentile among all GPUs in the database, while the GeForce 930A rests at the 31st percentile. The three-point percentile gap is small, suggesting both parts occupy a similar performance tier. The AMD card lands within a tight cluster of rivals: it is statistically even with the NVIDIA Quadro K620M (0% delta), marginally behind the AMD Radeon HD 8750M by 0.3%, and slightly ahead of the Intel UHD Graphics 730 by 0.4%. The GeForce 930A, meanwhile, matches the NVIDIA GeForce 840M within 0.1%, trails the NVIDIA GeForce GTX 980M by 0.2% (an odd pairing given that 980M is a much larger chip in the wider database), and sits 0.6% ahead of the GeForce 940M.

The use-case split is therefore not about one card dominating all workloads. The recorded data shows the AMD Radeon HD 8730M as the stronger OpenCL compute device, which matters for GPU-accelerated tasks such as video encoding, image processing, or any OpenCL-driven application. The GeForce 930A, despite losing the head-to-head, still posts a respectable score that places it within striking distance of the AMD part. For users who rely on NVIDIA-specific software ecosystems or driver optimizations, the 930A may still be the practical choice even if the raw compute benchmark favors AMD.

Architecture Differences

The two chips come from different architectural generations and different vendors, and the specifications reflect that divergence. The AMD Radeon HD 8730M uses the Mars chip built on GCN 1.0 architecture, manufactured by TSMC on a 28 nm process. The NVIDIA GeForce 930A uses the GM108 chip built on Maxwell architecture, also manufactured by TSMC on the same 28 nm node. Both are therefore products of the same foundry and process generation, which makes their performance differences more interesting.

Transistor counts differ slightly: the Mars chip packs 950 million transistors into a 77 mm² die, while the GM108 packs 1,020 million transistors into the same 77 mm² area. The resulting transistor density is 12.3 million per mm² for AMD and 13.2 million per mm² for NVIDIA, a 7% density advantage for the Maxwell part. That density advantage does not translate into a compute win in the recorded benchmark, which suggests clock speeds and memory configuration matter more.

Clock speeds are substantially different. The AMD part runs a 650 MHz base and 700 MHz boost, while the NVIDIA part runs a 928 MHz base and 941 MHz boost. The NVIDIA chip is clocked roughly 43% higher at base and 34% higher at boost, which should give it a raw execution advantage. Yet the AMD part still wins the OpenCL test, pointing to architectural efficiency differences between GCN 1.0 and Maxwell in this specific workload.

Memory configurations also diverge sharply. Both cards use 2 GB of DDR3, but the AMD Radeon HD 8730M uses a 128-bit bus with 28.80 GB/s of bandwidth, while the GeForce 930A uses a 64-bit bus with only 16.02 GB/s of bandwidth. That is a 44% bandwidth advantage for AMD, which is likely the decisive factor in the OpenCL result. The NVIDIA card's memory clock is 1001 MHz with 2 Gbps effective, while the AMD card runs 900 MHz with 1800 Mbps effective, but the wider bus more than compensates for the lower clock.

Both cards share identical shading unit counts (384), texture mapping units (24), and render output units (8). This is unusual for cross-vendor comparisons and suggests the two parts were designed for the same performance tier. The feature set differs in API support: AMD lists DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170, while NVIDIA lists DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. The Vulkan version difference is notable, with NVIDIA supporting a much newer revision.

Head-to-Head Benchmarks

The database records a single head-to-head benchmark: Geekbench OpenCL. The AMD Radeon HD 8730M scores 5955 against the GeForce 930A's 5317, a 12% delta in favor of AMD. This is the only direct comparison available, so the analysis must lean heavily on this result.

The 12% margin is meaningful but not overwhelming. For context, the AMD card's nearest rival, the NVIDIA Quadro K620M, scores 5957, which is 0% different from the 8730M. The GeForce 930A's nearest rival, the GeForce 840M, scores 5322, which is 0.1% below the 930A. This means the AMD part is effectively tied with a workstation-class NVIDIA GPU, while the 930A is effectively tied with a mainstream NVIDIA mobile GPU. The performance gap between these two clusters is consistent with the 12% head-to-head delta.

The AMD card also shows a 0.4% edge over the Intel UHD Graphics 730 and a 0.3% deficit against the AMD Radeon HD 8750M, placing it in a tightly packed group of four GPUs spanning just 53 points (from 5929 to 5982). The GeForce 930A's cluster is similarly tight: five GPUs spanning 74 points (from 5284 to 5358). The two clusters do not overlap, reinforcing the head-to-head result.

Pixel and texture rates tell a complementary story. The NVIDIA part achieves 7.528 GPixel/s and 22.58 GTexel/s, while the AMD part achieves 5.600 GPixel/s and 16.80 GTexel/s. NVIDIA leads by 34% in pixel throughput and 34% in texture throughput, yet loses the OpenCL compute test. This suggests the AMD architecture converts its raw resources into OpenCL performance more efficiently, likely due to the memory bandwidth advantage. FP32 compute follows a similar pattern: NVIDIA posts 722.7 GFLOPS against AMD's 537.6 GFLOPS, a 34% lead, yet still loses the compute benchmark. The pattern is consistent: NVIDIA has higher throughput on paper, but AMD wins the actual measurement.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The AMD Radeon HD 8730M scores 5955 in Geekbench OpenCL, which is 12% higher than the NVIDIA GeForce 930A's 5317.

Q: How do the memory buses compare between the two cards?

A: The AMD Radeon HD 8730M uses a 128-bit bus with 28.80 GB/s bandwidth, while the NVIDIA GeForce 930A uses a 64-bit bus with 16.02 GB/s bandwidth. The AMD card has a 44% bandwidth advantage.

Q: Are the shading unit counts different?

A: No, both GPUs have exactly 384 shading units, 24 texture mapping units, and 8 render output units.

Q: Which GPU has the higher clock speed?

A: The NVIDIA GeForce 930A runs at a 928 MHz base and 941 MHz boost, compared to the AMD Radeon HD 8730M's 650 MHz base and 700 MHz boost.

Q: How does the AMD Radeon HD 8730M compare to its nearest rivals?

A: It is 0% different from the NVIDIA Quadro K620M, 0.3% behind the AMD Radeon HD 8750M, and 0.4% ahead of the Intel UHD Graphics 730.

Q: What API versions does each card support?

A: The AMD Radeon HD 8730M supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The NVIDIA GeForce 930A supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4.

The Verdict

The data points to a clear but narrow conclusion. The AMD Radeon HD 8730M is the better choice for OpenCL compute workloads, based on its 12% head-to-head win and its higher percentile ranking (34th versus 31st). The likely driver is memory bandwidth: the AMD card's 128-bit bus delivers 28.80 GB/s, nearly double the 16.02 GB/s of the NVIDIA part. For users running OpenCL-accelerated applications, the AMD card should deliver noticeably faster results.

The NVIDIA GeForce 930A is not without reasons for selection. It offers higher clock speeds (928 MHz base versus 650 MHz), higher pixel rate (7.528 GPixel/s versus 5.600 GPixel/s), higher texture rate (22.58 GTexel/s versus 16.80 GTexel/s), and higher FP32 throughput (722.7 GFLOPS versus 537.6 GFLOPS). It also supports Vulkan 1.4, a much newer API revision than the AMD card's 1.2.170. For graphics-oriented tasks or software that leverages newer Vulkan features, the 930A may be the more sensible pick despite losing the compute benchmark.

The transistor density advantage also favors NVIDIA: 13.2 million transistors per mm² versus 12.3 million, with 1,020 million total transistors against 950 million. This suggests the Maxwell architecture is more efficient in its use of silicon, even if that efficiency does not translate into an OpenCL win. The GeForce 930A also has a defined 33 W TDP, while the AMD card's power consumption is not recorded in the database, so power-constrained systems may prefer the NVIDIA part.

Final recommendation from the recorded data: choose the AMD Radeon HD 8730M for compute-accelerated workloads where OpenCL performance matters most. Choose the NVIDIA GeForce 930A for graphics performance, newer API support, and a defined power envelope. Neither card dominates the other across all metrics; the decision depends on the specific workload.

Specification Differences

The following table covers only the fields where the two GPUs differ, according to the database:

| Field | AMD Radeon HD 8730M | NVIDIA GeForce 930A |

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

| Chip | Mars | GM108 |

| Architecture | GCN 1.0 | Maxwell |

| Transistors | 950 million | 1,020 million |

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

| Base Clock | 650 MHz | 928 MHz |

| Boost Clock | 700 MHz | 941 MHz |

| Memory Clock | 900 MHz, 1800 Mbps effective | 1001 MHz, 2 Gbps effective |

| Memory Bus Width | 128 bit | 64 bit |

| Memory Bandwidth | 28.80 GB/s | 16.02 GB/s |

| Pixel Rate | 5.600 GPixel/s | 7.528 GPixel/s |

| Texture Rate | 16.80 GTexel/s | 22.58 GTexel/s |

| FP32 | 537.6 GFLOPS | 722.7 GFLOPS |

| TDP | Not recorded | 33 W |

| Slot Width | Not recorded | IGP |

| Power Connectors | Not recorded | None |

| Display Outputs | Not recorded | Portable Device Dependent |

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

| Vulkan Support | 1.2.170 | 1.4 |

| Release Date | 2013-03-31 | 2015-03-12 |

| Predecessor | London | GeForce 800A |

| Successor | Gem System | Not recorded |

| Geekbench OpenCL Score | 5955 | 5317 |

| Percentile | 34th | 31st |

Fields not listed, such as memory size (both 2 GB), memory type (both DDR3), shading units (both 384), TMUs (both 24), ROPs (both 8), process node (both 28 nm), foundry (both TSMC), die size (both 77 mm²), bus interface (both PCIe 3.0 x8), OpenGL support (both 4.6), and production status (both end-of-life), are identical between the two parts.