AMD Radeon HD 7730M vs NVIDIA GeForce GTX 670MX Comparison

FAQ

Q: How does the AMD Radeon HD 7730M compare to the NVIDIA GeForce GTX 670MX in OpenCL performance?

A: In Geekbench OpenCL, the AMD Radeon HD 7730M scores 6581, while the NVIDIA GeForce GTX 670MX scores 6125. That gives the AMD part a 7.4% lead in this specific test.

Q: Which GPU has a higher average benchmark score?

A: The AMD Radeon HD 7730M has an average benchmark score of 6581, whereas the NVIDIA GeForce GTX 670MX averages 5721. The AMD part sits at the 38th percentile of all GPUs, while the NVIDIA part sits at the 33rd percentile.

Q: What are the closest rivals to each GPU?

A: For the AMD Radeon HD 7730M, the nearest rival is the Intel UHD Graphics P750, which scores 6554, a 0.4% difference. The AMD Radeon R7 M460 scores 6612, which is 0.5% higher. The NVIDIA GeForce GTX 670MX is closest to the Intel Iris Pro Graphics P6300, which scores 5712, a 0.2% difference, and the NVIDIA GeForce GTX 550 Ti, which scores 5731, 0.2% higher.

Q: What memory configurations do these two GPUs use?

A: The AMD Radeon HD 7730M uses 2 GB of DDR3 memory on a 128-bit bus, yielding 28.80 GB/s of bandwidth. The NVIDIA GeForce GTX 670MX uses 3 GB of GDDR5 memory on a 192-bit bus, yielding 67.20 GB/s of bandwidth.

Q: Which GPU has a higher transistor count?

A: The NVIDIA GeForce GTX 670MX has 3,540 million transistors on a 294 mm² die, while the AMD Radeon HD 7730M has 1,500 million transistors on a 123 mm² die. Both are built on TSMC's 28 nm process.

Q: What is the power draw difference between the two?

A: The AMD Radeon HD 7730M has a TDP of 25 W, while the NVIDIA GeForce GTX 670MX has a TDP of 75 W. The NVIDIA part consumes three times the power budget of the AMD part.

Where Each One Wins

The benchmark data points to a clear split: the AMD Radeon HD 7730M wins in raw compute throughput in the recorded OpenCL test. Its score of 6581 versus 6125 for the NVIDIA GeForce GTX 670MX means that for workloads that stress general-purpose compute, the AMD part is the stronger choice. The 7.4% delta is not a marginal difference; it is a measurable gap that shows up in the database's single head-to-head test.

The NVIDIA GeForce GTX 670MX, however, has its own advantages that do not appear in the win column but matter for real-world usage. It has a higher memory bandwidth at 67.20 GB/s versus 28.80 GB/s, more shading units at 960 versus 512, and more texture mapping units at 80 versus 32. These specifications suggest that in texture-heavy or memory-bandwidth-bound scenarios, the NVIDIA part could pull ahead, even though the recorded OpenCL score does not capture that. The GTX 670MX also supports Vulkan 1.2.175 versus 1.2.170 for the AMD part, and it uses a PCIe 3.0 x16 interface versus PCIe 2.0 x16.

For users who prioritize compute workloads, the AMD Radeon HD 7730M is the winner in the recorded data. For users who need more memory, more bandwidth, or more shading units, the NVIDIA GeForce GTX 670MX has the specification-level edge. The two GPUs are not interchangeable; each has a distinct profile that suits different tasks.

Architecture Differences

The AMD Radeon HD 7730M is built on GCN 1.0 architecture with the Chelsea chip. The NVIDIA GeForce GTX 670MX uses the Kepler architecture with the GK104 chip. Both are manufactured on a 28 nm process at TSMC, but the similarities end there.

GCN 1.0 is designed around a unified shader architecture that emphasizes compute performance. The Chelsea chip packs 512 shading units, 32 texture mapping units, and 16 raster output units. The transistor density is 12.2 million per square millimeter, which is slightly higher than the NVIDIA part's 12.0 million per square millimeter. This density comes from the smaller die: 123 mm² versus 294 mm².

Kepler, on the other hand, is NVIDIA's architecture that focuses on efficiency and scalability. The GK104 chip is a larger design with 960 shading units, 80 texture mapping units, and 24 raster output units. The transistor count is more than double that of the AMD part: 3,540 million versus 1,500 million. The die size is also more than double at 294 mm².

The memory subsystems reflect the architectural priorities. The AMD part uses DDR3 memory with a 128-bit bus, while the NVIDIA part uses GDDR5 memory with a 192-bit bus. GDDR5 provides significantly higher bandwidth, which is critical for texture-heavy workloads. The AMD part's smaller memory bus and slower memory type limit its bandwidth to 28.80 GB/s, while the NVIDIA part reaches 67.20 GB/s.

Both GPUs support DirectX 12, but with different feature levels. The AMD part supports DirectX 12 (11_1), while the NVIDIA part supports DirectX 12 (11_0). Both support OpenGL 4.6. Vulkan support is present on both, with the NVIDIA part having a slightly newer version.

The power profiles differ dramatically. The AMD Radeon HD 7730M has a TDP of 25 W, which is typical for a low-power mobile part. The NVIDIA GeForce GTX 670MX has a TDP of 75 W, which is typical for a higher-performance mobile part. This difference in power draw is a direct consequence of the larger die, more transistors, and faster memory on the NVIDIA side.

Specification Differences

The two GPUs differ across nearly every major specification category. The AMD Radeon HD 7730M has 512 shading units, 32 TMUs, and 16 ROPs. The NVIDIA GeForce GTX 670MX has 960 shading units, 80 TMUs, and 24 ROPs. The NVIDIA part has nearly double the shading units and more than double the texture mapping units.

Clock speeds differ as well. The AMD part has a base clock of 575 MHz and a boost clock of 675 MHz. The NVIDIA part has a base clock of 601 MHz and a boost clock of 601 MHz, meaning it does not boost beyond its base clock. The AMD part's boost clock is 74 MHz higher than its base, while the NVIDIA part runs at a fixed clock.

Memory specifications show a clear NVIDIA advantage. The AMD part has 2 GB of DDR3 on a 128-bit bus with 28.80 GB/s bandwidth. The NVIDIA part has 3 GB of GDDR5 on a 192-bit bus with 67.20 GB/s bandwidth. The NVIDIA part offers 50% more memory capacity and more than double the bandwidth.

Compute rates reflect the architectural differences. The AMD part delivers 691.2 GFLOPS of FP32 performance, while the NVIDIA part delivers 1,153.9 GFLOPS. Pixel rates are 10.80 GPixel/s for the AMD part and 12.02 GPixel/s for the NVIDIA part. Texture rates are 21.60 GTexel/s for the AMD part and 48.08 GTexel/s for the NVIDIA part.

The bus interface differs: PCIe 2.0 x16 for the AMD part versus PCIe 3.0 x16 for the NVIDIA part. The NVIDIA part also has no power connectors listed, while the AMD part's power connector configuration is not specified. Both are marked as end-of-life products. The AMD part was released on April 23, 2012, while the NVIDIA part was released on September 30, 2012, about five months later.

Head-to-Head Benchmarks

The only recorded head-to-head benchmark is Geekbench OpenCL. In this test, the AMD Radeon HD 7730M scores 6581, while the NVIDIA GeForce GTX 670MX scores 6125. The AMD part wins with a 7.4% delta. This is the single data point where the two GPUs are directly compared.

The 7.4% lead in OpenCL is notable because the NVIDIA part has a substantially larger die, more transistors, more shading units, and more memory bandwidth. The fact that the AMD part still wins in this compute-oriented test suggests that GCN 1.0's architecture is more efficient at translating its resources into OpenCL performance, at least in this workload. The AMD part's higher boost clock of 675 MHz versus 601 MHz likely contributes to this result, as does the architecture's design focus on compute throughput.

The nearest rivals for the AMD part in the database include the Intel UHD Graphics P750 at 6554, which is 0.4% lower, and the AMD Radeon R7 M460 at 6612, which is 0.5% higher. The NVIDIA GeForce GTX 670M scores 6513, which is 1% lower than the AMD part. The NVIDIA GeForce GT 555M scores 6493, which is 1.4% lower. These comparisons show that the AMD Radeon HD 7730M sits in a tight cluster of GPUs with scores between roughly 6500 and 6600.

For the NVIDIA GeForce GTX 670MX, the nearest rivals are the Intel Iris Pro Graphics P6300 at 5712, which is 0.2% lower, and the NVIDIA GeForce GTX 550 Ti at 5731, which is 0.2% higher. The AMD Radeon HD 8790M scores 5691, which is 0.5% lower. The NVIDIA Quadro M500M scores 5604, which is 2.1% lower. The GTX 670MX's average score of 5721 places it in a lower performance tier than the AMD part, which averages 6581.

The database records one win for the AMD part and zero for the NVIDIA part. This is a straightforward result: in the measured OpenCL workload, the AMD Radeon HD 7730M outperforms the NVIDIA GeForce GTX 670MX by a clear margin. The specification-level advantages of the NVIDIA part, such as higher bandwidth and more shading units, do not translate into a higher OpenCL score in this case. The data indicates that for compute-heavy tasks, the AMD part is the better performer, despite its smaller die and lower power draw.