AMD Radeon R5 M435 vs NVIDIA GeForce GTX 670MX Comparison

# AMD Radeon R5 M435 vs NVIDIA GeForce GTX 670MX

The AMD Radeon R5 M435 and NVIDIA GeForce GTX 670MX represent two very different approaches to mobile graphics, separated by nearly four years of development. The data shows a single head-to-head benchmark where the NVIDIA part wins, yet the two GPUs land in the same overall percentile bracket, making the choice less obvious than raw scores suggest.

Where Each One Wins

The NVIDIA GeForce GTX 670MX claims the only direct head-to-head victory, winning the Geekbench OpenCL test with a score of 6,125 against the R5 M435’s 5,859 — a 4.3% margin. This makes the GTX 670MX the clear pick for compute-heavy workloads that stress OpenCL performance, particularly those that benefit from higher raw throughput and wider memory interfaces.

The AMD Radeon R5 M435, despite losing the head-to-head, does not trail by a wide margin. Its OpenCL score of 5,859 falls within 4.3% of the GTX 670MX, which means for many real-world tasks the difference may be imperceptible. The R5 M435 also holds an advantage in portability, being an integrated graphics processor (IGP) with a slot width of "IGP," whereas the GTX 670MX is a discrete solution with a 75 W TDP. For thin-and-light systems where power draw and thermals matter more than peak compute, the R5 M435’s integrated nature is a structural win.

Interestingly, the GTX 670MX also posts a Geekbench Vulkan score of 5,316, while the R5 M435 has no Vulkan benchmark recorded. This suggests the NVIDIA GPU has been validated for Vulkan workloads, whereas AMD’s part lacks comparable data in this area — a point in NVIDIA’s favor for modern API support.

Architecture Differences

The two GPUs could not be more different under the hood. The AMD Radeon R5 M435 uses the Jet chip built on GCN 1.0 architecture, fabricated on a 28 nm process at TSMC. The GTX 670MX uses the GK104 chip with Kepler architecture, also on a 28 nm TSMC process. Both are end-of-life products, but their design philosophies diverge sharply.

The transistor counts tell a dramatic story. The GTX 670MX packs 3,540 million transistors on a 294 mm² die, while the R5 M435 has just 690 million transistors on a 56 mm² die. That is over five times more transistors on five times the silicon area for the NVIDIA part. Transistor density is nearly identical — 12.3M / mm² for AMD versus 12.0M / mm² for NVIDIA — confirming that both use similar manufacturing maturity, but NVIDIA chose to scale up the chip massively while AMD kept things small and efficient.

The memory subsystems differ fundamentally. The R5 M435 has 2 GB of GDDR5 on a 64-bit bus, yielding 36.00 GB/s of bandwidth. The GTX 670MX has 3 GB of GDDR5 on a 192-bit bus, delivering 67.20 GB/s — nearly double the bandwidth. Clock speeds also favor AMD: the R5 M435 runs at 780 MHz base and 1030 MHz boost, while the GTX 670MX is locked at 601 MHz for both base and boost. This lower clock is offset by the GTX 670MX’s far wider execution resources.

Compute unit counts reinforce the disparity. The GTX 670MX has 960 shading units, 80 TMUs, and 24 ROPs, versus the R5 M435’s 320 shading units, 20 TMUs, and 8 ROPs. The NVIDIA part also achieves higher pixel rate (12.02 GPixel/s versus 8.240 GPixel/s) and texture rate (48.08 GTexel/s versus 20.60 GTexel/s). FP32 throughput is 1,153.9 GFLOPS for NVIDIA versus 659.2 GFLOPS for AMD — a 75% advantage. Neither GPU has RT or tensor cores, and both support DirectX 12 with feature-level limitations (AMD at 11_1, NVIDIA at 11_0). Both support OpenGL 4.6, with Vulkan versions close (AMD 1.2.170, NVIDIA 1.2.175).

The bus interface also differs: AMD uses PCIe 3.0 x8, while NVIDIA uses PCIe 3.0 x16. This could impact data transfer in bandwidth-sensitive scenarios, though for most mobile workloads the difference is likely minor.

Head-to-Head Benchmarks

The only direct comparison available is Geekbench OpenCL, where the GTX 670MX scores 6,125 against the R5 M435’s 5,859. The delta of -4.3% from AMD’s perspective means the R5 M435 trails by a modest margin. This is a smaller gap than the hardware specifications would suggest — the GTX 670MX has 3x the shading units and nearly 2x the bandwidth, yet only manages a 4.3% win in this particular workload. This suggests that OpenCL performance here is not purely scaling with compute resources, possibly due to driver optimizations, memory latency, or the specific workload characteristics.

Looking at the nearest rivals provides context. The R5 M435’s average benchmark score is 5,859, placing it between the AMD Radeon R7 M465 (5,841, +0.3%) and the Intel UHD Graphics 730 (5,929, -1.2%). It also sits just below the AMD Radeon HD 8730M (5,955, -1.6%) and NVIDIA Quadro K620M (5,957, -1.6%). This clustering shows the R5 M435 is competitive with a broad range of mid-range mobile GPUs from multiple generations.

The GTX 670MX’s average benchmark score is 5,721, which is actually lower than its OpenCL score of 6,125 because the Vulkan score of 5,316 drags the average down. Its nearest rivals include the Intel Iris Pro Graphics P6300 (5,712, +0.2%), NVIDIA GeForce GTX 550 Ti (5,731, -0.2%), AMD Radeon HD 8790M (5,691, +0.5%), and NVIDIA Quadro M500M (5,604, +2.1%). This places the GTX 670MX in a similar performance tier to the R5 M435, despite its much larger physical hardware.

Both GPUs sit at the 33rd percentile among all GPUs, meaning they outperform roughly one-third of the database. This parity in percentile, despite wildly different architectures, is a key finding — the R5 M435 achieves similar overall standing with far less silicon, while the GTX 670MX uses its massive transistor budget to achieve a similar result.

FAQ

Q: Which GPU has more raw compute power?

A: The GTX 670MX has significantly more raw compute resources, with 960 shading units versus 320 and FP32 throughput of 1,153.9 GFLOPS versus 659.2 GFLOPS. However, in the OpenCL benchmark, it only leads by 4.3%.

Q: Is the R5 M435 more power-efficient?

A: The data shows the R5 M435 is an integrated GPU with an "IGP" slot width and no listed TDP, while the GTX 670MX has a 75 W TDP and no power connectors. This suggests the AMD part is designed for lower-power systems, though exact efficiency figures are not provided.

Q: Which GPU has better memory bandwidth?

A: The GTX 670MX delivers 67.20 GB/s over a 192-bit bus, while the R5 M435 provides 36.00 GB/s over a 64-bit bus. The NVIDIA part has nearly double the bandwidth.

Q: Do both GPUs support modern APIs?

A: Both support DirectX 12, OpenGL 4.6, and Vulkan. The GTX 670MX has a recorded Vulkan benchmark score of 5,316, while the R5 M435 has no Vulkan benchmark data; its Vulkan version is 1.2.170 versus 1.2.175 for NVIDIA.

Q: How do their average benchmark scores compare?

A: The R5 M435 has an average score of 5,859, while the GTX 670MX averages 5,721. Despite the GTX 670MX winning the single head-to-head OpenCL test, its average is dragged down by its Vulkan score.

Q: Which GPU has more VRAM?

A: The GTX 670MX has 3 GB of GDDR5, while the R5 M435 has 2 GB. Both use GDDR5 memory.

The Verdict

The data presents a nuanced picture. If the sole criterion is the recorded OpenCL head-to-head, the GTX 670MX wins with a 4.3% margin. Its higher shading unit count, wider memory bus, and greater bandwidth suggest it should dominate more decisively, yet the benchmark gap is narrow. This implies the R5 M435 is remarkably efficient for its size, achieving 89% of the GTX 670MX’s OpenCL score with only 20% of the transistor count.

For users prioritizing raw compute in OpenCL workloads, the GTX 670MX is the safer choice, especially given its Vulkan validation and higher pixel/texture rates. However, the R5 M435’s integrated nature means it can fit into systems where a discrete GPU with a 75 W TDP would be impractical. The R5 M435 also has a higher base clock (780 MHz versus 601 MHz) and boost clock (1030 MHz versus 601 MHz), which may help in latency-sensitive tasks.

The percentile standings are identical at 33, indicating that in the broader GPU landscape, neither card holds a meaningful advantage over the other. The GTX 670MX’s average score of 5,721 is actually lower than the R5 M435’s 5,859, suggesting that across multiple workloads, the AMD part may be the more consistent performer. The absence of a Vulkan score for the R5 M435 is a data gap, not necessarily a capability gap.

For gaming or general use where OpenCL is not the primary workload, the R5 M435’s higher clocks and integrated form factor may serve better. For compute-heavy tasks where OpenCL and Vulkan are used, the GTX 670MX’s wider hardware and proven Vulkan score give it an edge, albeit a smaller one than specifications suggest.

Specification Differences

| Field | AMD Radeon R5 M435 | NVIDIA GeForce GTX 670MX |

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

| Chip | Jet | GK104 |

| Architecture | GCN 1.0 | Kepler |

| Process Node | 28 nm | 28 nm |

| Transistors | 690 million | 3,540 million |

| Die Size | 56 mm² | 294 mm² |

| Transistor Density | 12.3M / mm² | 12.0M / mm² |

| Base Clock | 780 MHz | 601 MHz |

| Boost Clock | 1030 MHz | 601 MHz |

| Memory Clock | 1125 MHz / 4.5 Gbps effective | 700 MHz / 2.8 Gbps effective |

| Memory Size | 2 GB | 3 GB |

| Memory Bus Width | 64 bit | 192 bit |

| Memory Bandwidth | 36.00 GB/s | 67.20 GB/s |

| Shading Units | 320 | 960 |

| TMUs | 20 | 80 |

| ROPs | 8 | 24 |

| Pixel Rate | 8.240 GPixel/s | 12.02 GPixel/s |

| Texture Rate | 20.60 GTexel/s | 48.08 GTexel/s |

| FP32 | 659.2 GFLOPS | 1,153.9 GFLOPS |

| TDP | Not specified | 75 W |

| Slot Width | IGP | Not specified |

| Power Connectors | Not specified | None |

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

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

| Vulkan | 1.2.170 | 1.2.175 |

| Release Date | 2016-05-14 | 2012-09-30 |

| Predecessor | Solar System | GeForce 500M |

| Successor | Polaris Mobile | GeForce 700M |

| Geekbench OpenCL | 5,859 | 6,125 |

| Geekbench Vulkan | Not recorded | 5,316 |

| Average Benchmark Score | 5,859 | 5,721 |

| Percentile vs All GPUs | 33 | 33 |