AMD Radeon HD 8730M vs NVIDIA GeForce GTX 760M Comparison

The Verdict

The benchmark database places the AMD Radeon HD 8730M and NVIDIA GeForce GTX 760M in adjacent performance tiers, but the data favors AMD in the only direct head-to-head comparison recorded. The HD 8730M scores 5955 in Geekbench OpenCL, while the GTX 760M scores 5604, giving AMD a 6.3% lead. That edge is modest, yet it flips the expected hierarchy given the transistor counts and memory bandwidth on paper.

For users choosing between these two end-of-life mobile GPUs, the AMD Radeon HD 8730M is the safer pick for compute workloads that rely on OpenCL acceleration. Its 34th percentile ranking across all GPUs slightly edges out the GTX 760M's 31st percentile, and its average benchmark score of 5955 versus 5236 for NVIDIA reflects a broader consistency across tests. The GTX 760M does have a Vulkan result in the database (4868), which the HD 8730M lacks entirely, so anyone targeting Vulkan-based applications should consider the NVIDIA part despite its lower OpenCL showing.

The GTX 760M is not without merits. Its 768 shading units, 64 texture mapping units, and 16 ROPs dwarf the HD 8730M's 384 shaders, 24 TMUs, and 8 ROPs. The NVIDIA chip also carries twice the transistor budget at 2,540 million versus 950 million, and its 64.13 GB/s memory bandwidth is more than double AMD's 28.80 GB/s. Yet in the recorded OpenCL test, raw resources did not translate into a win. The database shows AMD's GCN 1.0 architecture delivering higher compute efficiency in this specific workload.

The verdict is straightforward: pick the AMD Radeon HD 8730M if OpenCL throughput is the priority, as the data shows a 6.3% victory. Pick the NVIDIA GeForce GTX 760M if you need Vulkan API support or expect driver-level optimization for NVIDIA-specific features, since the HD 8730M has no recorded Vulkan score. Neither card is competitive by modern standards, but between these two, the AMD part wins the only head-to-head benchmark available.

Where Each One Wins

The AMD Radeon HD 8730M wins the sole head-to-head benchmark in the database. In Geekbench OpenCL, it scores 5955 against the GTX 760M's 5604, a delta of 6.3%. This is the only direct comparison recorded, so AMD holds a clean 1-0 record in head-to-head tests. The win is notable because the GTX 760M has superior theoretical specifications: 768 shading units versus 384, 64 TMUs versus 24, 16 ROPs versus 8, and 64.13 GB/s bandwidth versus 28.80 GB/s. The HD 8730M compensates with a higher base clock of 650 MHz versus 628 MHz and a boost clock of 700 MHz versus 719 MHz, though NVIDIA's boost is actually higher. The AMD chip also has a smaller die (77 mm² versus 221 mm²) and fewer transistors (950 million versus 2,540 million), suggesting its GCN 1.0 architecture extracts more performance per transistor in OpenCL.

The NVIDIA GeForce GTX 760M wins in areas not covered by the head-to-head test. It has a recorded Vulkan score of 4868, which the HD 8730M cannot match because the database lists no Vulkan result for the AMD part. The GTX 760M also supports a newer Vulkan version (1.2.175 versus 1.2.170) and DirectX 12 (11_0) versus AMD's DirectX 12 (11_1), though the latter is a higher feature level. For memory-bound tasks, the GTX 760M's GDDR5 memory at 4 Gbps effective and 64.13 GB/s bandwidth is substantially faster than the HD 8730M's DDR3 at 1800 Mbps effective and 28.80 GB/s. The NVIDIA chip also has a higher pixel rate (11.50 GPixel/s versus 5.600 GPixel/s) and texture rate (46.02 GTexel/s versus 16.80 GTexel/s), which could benefit fill-rate-limited workloads even if the OpenCL test does not reflect that advantage.

In terms of PCIe connectivity, the GTX 760M uses PCIe 3.0 x16, while the HD 8730M uses PCIe 3.0 x8. That halves the available bus bandwidth for AMD, yet it still won the OpenCL test. The GTX 760M is also rated at 55 W TDP, while the HD 8730M has no TDP listed in the database, so power efficiency comparisons cannot be made from the recorded data.

Architecture Differences

The AMD Radeon HD 8730M is built on GCN 1.0 architecture, using the Mars chip, and belongs to the Solar System generation (HD 8700M). It is manufactured on a 28 nm process at TSMC, packing 950 million transistors into a 77 mm² die, yielding a transistor density of 12.3 million per square millimeter. The NVIDIA GeForce GTX 760M uses the Kepler architecture with the GK106S chip, part of the GeForce 700M generation. It is also on TSMC's 28 nm node but uses 2,540 million transistors across a 221 mm² die, with a lower density of 11.5 million per square millimeter.

The compute resources differ sharply. AMD provides 384 shading units, 24 texture mapping units, and 8 ROPs. NVIDIA doubles the shader count to 768, nearly triples TMUs to 64, and doubles ROPs to 16. The FP32 throughput reflects this: the GTX 760M delivers 1,104.4 GFLOPS against the HD 8730M's 537.6 GFLOPS. Pixel rate is 11.50 GPixel/s for NVIDIA versus 5.600 GPixel/s for AMD, and texture rate is 46.02 GTexel/s versus 16.80 GTexel/s. These are large gaps on paper, yet the OpenCL benchmark did not follow that pattern.

Memory architecture also diverges. The HD 8730M uses 2 GB of DDR3 on a 128-bit bus, running at 900 MHz (1800 Mbps effective), for 28.80 GB/s bandwidth. The GTX 760M uses 2 GB of GDDR5 on the same 128-bit bus, at 1002 MHz (4 Gbps effective), for 64.13 GB/s bandwidth. That is a 2.2x bandwidth advantage for NVIDIA, but it did not secure a win in the recorded test.

API support reveals another split. AMD lists DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. NVIDIA lists DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The Vulkan version is marginally newer on the NVIDIA side, and the GTX 760M has a Vulkan benchmark score while AMD does not. The HD 8730M's DirectX feature level is one notch higher, which could matter for certain legacy titles.

Physical and power characteristics differ as well. The GTX 760M is a 55 W MXM module with no power connectors, and its display outputs are marked "Portable Device Dependent." The HD 8730M has no TDP, slot width, or display output data in the database. The GTX 760M uses PCIe 3.0 x16, while the HD 8730M is limited to PCIe 3.0 x8. Both are end-of-life products, with the HD 8730M released on 2013-03-31 and the GTX 760M on 2013-05-29. The HD 8730M's predecessor is London and its successor is Gem System; the GTX 760M's predecessor is GeForce 600M and its successor is GeForce 800M.

FAQ

Q: Which GPU has a higher OpenCL benchmark score?

A: The AMD Radeon HD 8730M scores 5955 in Geekbench OpenCL, while the NVIDIA GeForce GTX 760M scores 5604. AMD leads by 6.3% in the direct head-to-head test.

Q: Does the NVIDIA GeForce GTX 760M support Vulkan?

A: Yes, it has a recorded Geekbench Vulkan score of 4868 and supports Vulkan version 1.2.175. The AMD Radeon HD 8730M lists Vulkan 1.2.170 but has no Vulkan benchmark score in the database.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce GTX 760M has 768 shading units, exactly double the 384 found on the AMD Radeon HD 8730M.

Q: What is the memory bandwidth difference?

A: The GTX 760M offers 64.13 GB/s with GDDR5 memory, while the HD 8730M provides 28.80 GB/s with DDR3 memory. Both use a 128-bit bus and 2 GB capacity.

Q: Are these GPUs still in production?

A: No, both are marked as end-of-life in the database. The HD 8730M was released on 2013-03-31, and the GTX 760M followed on 2013-05-29.

Q: Which GPU has a higher transistor count?

A: The NVIDIA GeForce GTX 760M uses 2,540 million transistors on a 221 mm² die, compared to the AMD Radeon HD 8730M's 950 million transistors on a 77 mm² die.

Head-to-Head Benchmarks

The only recorded head-to-head benchmark is Geekbench OpenCL, and the AMD Radeon HD 8730M takes it decisively. AMD scores 5955, NVIDIA scores 5604, and the delta is 6.3% in AMD's favor. This result is striking because the GTX 760M has a massive theoretical resource advantage. Its FP32 throughput of 1,104.4 GFLOPS is more than double the HD 8730M's 537.6 GFLOPS. Its texture rate of 46.02 GTexel/s nearly triples AMD's 16.80 GTexel/s, and its pixel rate of 11.50 GPixel/s is roughly double AMD's 5.600 GPixel/s. The memory bandwidth gap is even more pronounced: 64.13 GB/s versus 28.80 GB/s, a 2.2x difference.

Yet the OpenCL score tells a different story. The HD 8730M's GCN 1.0 architecture appears to execute OpenCL workloads with higher efficiency per compute unit than Kepler, at least in this specific test. The AMD chip's smaller footprint (77 mm² versus 221 mm²) and lower transistor count (950 million versus 2,540 million) suggest it does more with less, at least for this workload. The GTX 760M's higher boost clock of 719 MHz versus 700 MHz on AMD did not help close the gap.

The nearest rival data reinforces the positioning. The HD 8730M sits near the NVIDIA Quadro K620M, which scores 5957 (a 0% delta), and the AMD Radeon HD 8750M at 5970 (-0.3%). It also edges out the Intel UHD Graphics 730 at 5929 (0.4% ahead) and trails the NVIDIA Quadro K4000 at 5982 (-0.5%). The GTX 760M, meanwhile, clusters with the NVIDIA Quadro 4000M at 5211 (0.5% ahead), the NVIDIA GeForce 940M at 5284 (-0.9%), and the AMD Radeon R7 M260X at 5161 (1.5% ahead). Notably, the GTX 760M's average benchmark score of 5236 is dragged down by its Vulkan result of 4868, while its OpenCL score of 5604 is closer to the HD 8730M's 5955.

Looking at the broader picture, the HD 8730M's percentile ranking of 34 across all GPUs beats the GTX 760M's 31, and its average score of 5955 outstrips NVIDIA's 5236 by roughly 13.7%. If the database included more head-to-head tests, the GTX 760M's superior fill rates and bandwidth might produce wins in gaming or texture-heavy scenarios, but no such data exists. The recorded evidence gives AMD one win and NVIDIA zero. For OpenCL compute, the Radeon HD 8730M is the clear choice based on the numbers. For Vulkan or bandwidth-sensitive tasks, the GTX 760M's spec sheet argues in its favor, but the benchmark record does not support a victory.