AMD Radeon R5 M230 vs NVIDIA GeForce GTX 650 Comparison

The Verdict

The recorded data presents a fascinating near-tie. In the single shared benchmark, Geekbench OpenCL, the AMD Radeon R5 M230 scores 4577, while the NVIDIA GeForce GTX 650 scores 4545. The AMD part wins by a razor-thin margin of 0.7%. This result is almost within the margin of error, suggesting that for raw compute workloads measured by OpenCL, the two are effectively peers.

However, the broader database context reveals a more nuanced picture. The AMD Radeon R5 M230 sits at the 27th percentile of all GPUs, while the GeForce GTX 650 sits lower at the 22nd percentile. This discrepancy is explained by the GTX 650's average benchmark score of 3823, which is dragged down by its additional Metal and Vulkan scores of 2400 and 4524, respectively. The M230 only has an OpenCL score recorded, so its average of 4577 is not directly comparable to the GTX 650's multi-API average.

For a user choosing between these two, the decision hinges on the target application. If the workload is strictly OpenCL compute, the AMD Radeon R5 M230 holds a very slight edge. If the workload involves Vulkan, the GeForce GTX 650 is the only option with a recorded score in that API, posting 4524. The database indicates the R5 M230 is an integrated graphics processor (IGP), making it the default choice for a portable, low-power system. The GTX 650, with its single-slot design, dedicated power connector, and 65 W TDP, is a discrete desktop card that requires a power supply. Ultimately, the GTX 650 is the more versatile and feature-complete package for a desktop system, despite losing the single head-to-head compute test.

Architecture Differences

The two GPUs represent fundamentally different design philosophies from their respective manufacturers. The AMD Radeon R5 M230 is built on the Graphics Core Next 1.0 architecture, using the "Jet" chip. The NVIDIA GeForce GTX 650 uses the Kepler architecture, based on the "GK106" chip. Both are fabricated on a 28 nm process at TSMC, but the similarities end there.

The transistor counts and die sizes are starkly different. The AMD chip packs 690 million transistors into a 56 mm² die, resulting in a transistor density of 12.3 million per mm². The NVIDIA chip is significantly larger and more complex, containing 2,540 million transistors on a 221 mm² die, with a slightly lower density of 11.5 million per mm². This makes the GTX 650's chip roughly four times larger in terms of transistor count and die area.

Memory architecture is another major divergence. The R5 M230 uses 2 GB of DDR3 memory on a 64-bit bus, yielding a bandwidth of 16.00 GB/s. The GTX 650, in contrast, uses 1024 MB of GDDR5 memory on a 128-bit bus, delivering 80.00 GB/s of bandwidth, which is exactly five times higher. The memory clocks also reflect this: the AMD runs at 1000 MHz (2 Gbps effective), while the NVIDIA runs at 1250 MHz (5 Gbps effective).

The feature sets differ as well. The AMD part supports DirectX 12 (11_1), while the NVIDIA part supports DirectX 12 (11_0). Both support OpenGL 4.6, but the Vulkan versions differ, with AMD at 1.2.170 and NVIDIA at 1.2.175. The bus interface is also different: the R5 M230 uses PCIe 3.0 x8, while the GTX 650 uses the full PCIe 3.0 x16.

FAQ

Q: Which GPU has a higher transistor count?

A: The NVIDIA GeForce GTX 650 has a significantly higher transistor count at 2,540 million, compared to the AMD Radeon R5 M230's 690 million.

Q: What is the difference in memory bus width?

A: The AMD Radeon R5 M230 uses a 64-bit bus, while the NVIDIA GeForce GTX 650 uses a 128-bit bus, which is twice as wide.

Q: Which GPU supports the Vulkan API?

A: Both support Vulkan. The AMD Radeon R5 M230 supports version 1.2.170, and the NVIDIA GeForce GTX 650 supports version 1.2.175.

Q: What is the form factor of the AMD part?

A: The AMD Radeon R5 M230 is listed as an integrated graphics processor (IGP), meaning it is built into a system, not a standalone card.

Q: Does the NVIDIA card require external power?

A: Yes, the GeForce GTX 650 has a 1x 6-pin power connector and a suggested power supply of 250 W.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce GTX 650 has 384 shading units, which is more than the AMD Radeon R5 M230's 320 shading units.

Specification Differences

The following table outlines the key differences between the two GPUs.

| Specification | AMD Radeon R5 M230 | NVIDIA GeForce GTX 650 |

| :--- | :--- | :--- |

| Manufacturer | AMD | NVIDIA |

| Chip | Jet | GK106 |

| Architecture | GCN 1.0 | Kepler |

| Transistors | 690 million | 2,540 million |

| Die Size | 56 mm² | 221 mm² |

| Transistor Density | 12.3M / mm² | 11.5M / mm² |

| Memory Size | 2 GB | 1024 MB |

| Memory Type | DDR3 | GDDR5 |

| Memory Bus Width | 64 bit | 128 bit |

| Memory Bandwidth | 16.00 GB/s | 80.00 GB/s |

| Memory Clock | 1000 MHz (2 Gbps effective) | 1250 MHz (5 Gbps effective) |

| Shading Units | 320 | 384 |

| TMUs | 20 | 32 |

| ROPs | 8 | 16 |

| Pixel Rate | 4.880 GPixel/s | 8.464 GPixel/s |

| Texture Rate | 12.20 GTexel/s | 33.86 GTexel/s |

| FP32 | 390.4 GFLOPS | 812.5 GFLOPS |

| TDP | Not listed | 65 W |

| Slot Width | IGP | Single-slot |

| Power Connectors | Not listed | 1x 6-pin |

| Suggested PSU | Not listed | 250 W |

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

| Display Outputs | Portable Device Dependent | 1x DVI, 2x DisplayPort 1.2 |

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

| Vulkan Version | 1.2.170 | 1.2.175 |

| Dimensions (Length) | Not listed | 147 mm (5.8 inches) |

| Release Date | 2014-01-06 | 2013-11-26 |

Head-to-Head Benchmarks

The only recorded head-to-head benchmark is Geekbench OpenCL. In this test, the AMD Radeon R5 M230 scores 4577, and the NVIDIA GeForce GTX 650 scores 4545. The AMD part wins this test with a delta of 0.7%. This is a very narrow margin, indicating that the two GPUs perform almost identically in this specific compute workload.

Looking at the broader benchmark data, the GTX 650 has additional results that the AMD part lacks. In Geekbench Metal, the GTX 650 scores 2400. In Geekbench Vulkan, it scores 4524. The M230 has no recorded scores for these APIs. This means the GTX 650 is the only one of the two with a measurable performance footprint in Apple's Metal API and the cross-platform Vulkan API.

The theoretical peak figures also favor the NVIDIA card substantially. The GTX 650's FP32 performance is 812.5 GFLOPS, which is more than double the R5 M230's 390.4 GFLOPS. Similarly, the texture rate is 33.86 GTexel/s versus 12.20 GTexel/s, and the pixel rate is 8.464 GPixel/s versus 4.880 GPixel/s. These figures suggest that in raw throughput, the GTX 650 should be significantly faster, yet the OpenCL score does not reflect this advantage, pointing to potential driver overhead or architectural differences in how compute is scheduled.

Where Each One Wins

The AMD Radeon R5 M230 wins in the only direct comparison available. Its OpenCL score of 4577 is 0.7% higher than the GTX 650's 4545. This makes it the winner for purely OpenCL-based compute tasks, as measured by Geekbench. The M230 also has a slight advantage in API support for DirectX, supporting version 12 (11_1) versus the NVIDIA's 12 (11_0).

The NVIDIA GeForce GTX 650 wins in almost every other measurable category. It has a much larger transistor budget, a wider memory bus, and over five times the memory bandwidth. Its shading unit count is higher (384 vs 320), and it has more texture mapping units (32 vs 20) and more render output units (16 vs 8). The GTX 650 is the only one with a Vulkan score (4524), making it the choice for Vulkan-based applications. It also has a dedicated display output configuration (1x DVI, 2x DisplayPort 1.2), whereas the AMD part's outputs are described as "Portable Device Dependent," meaning it relies on the host laptop's ports.

The GTX 650 is also the only one with a recorded TDP of 65 W, a power connector, and a suggested PSU of 250 W, making it a self-contained desktop solution. The AMD part, being an IGP, is inherently tied to a mobile platform. In terms of raw theoretical throughput, the GTX 650 is the clear victor in pixel rate, texture rate, and FP32 compute. The data shows a classic case of a discrete, high-bandwidth desktop GPU versus an integrated, power-efficient mobile GPU, where the desktop part dominates in specifications, but the mobile part manages to hold its own in one specific compute benchmark.