NVIDIA GeForce MX230 vs NVIDIA Quadro K620M Comparison

# NVIDIA GeForce MX230 vs NVIDIA Quadro K620M

The NVIDIA GeForce MX230 and NVIDIA Quadro K620M are two end-of-life mobile graphics solutions aimed at different use cases, and the benchmark data shows a narrow but decisive edge for the older Quadro part. The MX230, built on the newer Pascal architecture with a 14 nm process, delivers modern API support and lower power draw, while the K620M, based on Maxwell with a 28 nm process, counters with a higher raw compute rating and a better OpenCL score. The data shows a single head-to-head benchmark, with the Quadro K620M winning by 3.7% in Geekbench OpenCL, but the full picture depends heavily on what you prioritize: efficiency and feature support versus raw compute consistency.

Where Each One Wins

The MX230 wins on architectural modernity and power efficiency. Its Pascal architecture, built on a 14 nm process from Samsung, allows it to operate at a 10 W TDP, which is three times lower than the K620M's 30 W TDP. This makes the MX230 the clear choice for thin-and-light notebooks where battery life and thermal headroom are critical. It also supports DirectX 12 (12_1) and Vulkan 1.4, matching the K620M's Vulkan support but exceeding its DirectX 12 (11_0) feature level. The MX230's memory runs at 6 Gbps effective GDDR5, delivering 48.06 GB/s of bandwidth, which is three times the K620M's 16.02 GB/s from DDR3 at 2 Gbps effective. For tasks that are memory-bandwidth sensitive, the MX230 will show a distinct advantage.

The K620M wins on raw compute throughput and benchmark scores. Despite its older Maxwell architecture and 28 nm process from TSMC, it packs 384 shading units versus the MX230's 256, and its FP32 performance of 863.2 GFLOPS edges out the MX230's 783.9 GFLOPS. In the only shared benchmark, Geekbench OpenCL, the K620M scores 5957 against the MX230's 5739, a 3.7% lead. The K620M also holds a higher average benchmark score of 5957 versus 6077 for the MX230—wait, that is not correct; the MX230's average score of 6077 is actually higher. Let's clarify: the MX230's average benchmark score is 6077, while the K620M's is 5957, meaning the MX230 leads by 2% on average across all recorded tests. However, the K620M wins the only direct comparison available. The K620M's pixel rate of 8.992 GPixel/s is far lower than the MX230's 24.50 GPixel/s, but its texture rate of 17.98 GTexel/s is also lower than the MX230's 24.50 GTexel/s, so the MX230 wins on fill rates decisively.

Architecture Differences

The two GPUs come from different NVIDIA generations and foundries. The MX230 is built on the GP108 chip using the Pascal architecture, fabricated by Samsung on a 14 nm process. It integrates 1,800 million transistors on a 74 mm² die, yielding a transistor density of 24.3M per mm². The K620M uses the GM108S chip with the Maxwell architecture, fabricated by TSMC on a 28 nm process, with 1,020 million transistors on a slightly larger 77 mm² die, giving a much lower density of 13.2M per mm². This density difference explains why the MX230 achieves similar or better performance with fewer shading units and lower power.

Memory subsystems differ sharply. The MX230 uses 2 GB of GDDR5 on a 64-bit bus, with memory clocked at 1502 MHz (6 Gbps effective) for 48.06 GB/s bandwidth. The K620M also has 2 GB but uses DDR3 on a 64-bit bus, with memory at 1001 MHz (2 Gbps effective) for only 16.02 GB/s. The MX230's bandwidth is exactly three times higher, which matters for any workload that streams data heavily. The K620M compensates with 384 shading units versus 256, and its FP32 output of 863.2 GFLOPS slightly beats the MX230's 783.9 GFLOPS, but the MX230's FP16 support (12.25 GFLOPS at 1:64 ratio) is present while the K620M has no FP16 capability listed.

Power and form factor also diverge. The MX230 is an IGP (integrated graphics processor) with a 10 W TDP and no power connectors, designed to be soldered onto motherboards. The K620M is an MXM Module with a 30 W TDP and no power connectors, meant for replaceable mobile workstation slots. The bus interfaces differ as well: the MX230 uses PCIe 3.0 x4, while the K620M uses MXM-A (3.0). Both have portable-device-dependent display outputs, meaning the actual ports depend on the laptop design.

Head-to-Head Benchmarks

The only direct benchmark comparison in the data is Geekbench OpenCL, and the results are close but conclusive. The MX230 scores 5739, while the K620M scores 5957, giving the Quadro a 3.7% advantage. This is a narrow margin, but it indicates that the K620M's extra shading units and higher FP32 throughput translate into a measurable lead in OpenCL compute workloads. The deltaPct of -3.7 for the MX230 means it trails by that percentage in this test.

Looking at the broader benchmark context, the MX230's average score of 6077 comes from two tests: its Geekbench OpenCL score of 5739 and its Geekbench Vulkan score of 6414. The Vulkan score is notably higher, suggesting that the MX230 performs better under Vulkan than OpenCL, which could indicate better driver optimization or architectural efficiency for that API. The K620M only has an OpenCL score, so its average of 5957 is identical to that single result. This means the MX230's average is 2% higher than the K620M's, but the only apples-to-apples comparison favors the Quadro.

The nearest rivals for each GPU put these scores in perspective. The MX230's average of 6077 sits within 1% of the NVIDIA RTX A400 (6078, 0% delta), the NVIDIA Quadro P2000 (6049, 0.5% delta), the Intel Iris Pro Graphics 6200 (6117, -0.7% delta), and the AMD Radeon 760M (6019, 1% delta). The K620M's average of 5957 is similarly clustered around the AMD Radeon HD 8730M (5955, 0% delta), the AMD Radeon HD 8750M (5970, -0.2% delta), the NVIDIA Quadro K4000 (5982, -0.4% delta), and the Intel UHD Graphics 730 (5929, 0.5% delta). Neither GPU is a standout in its class; both sit at the 34th-35th percentile among all GPUs.

FAQ

Q: Which GPU has better memory bandwidth?

A: The MX230 has a clear advantage, with 48.06 GB/s from GDDR5 memory, exactly three times the K620M's 16.02 GB/s from DDR3. This affects any workload that relies on moving large data sets.

Q: Is the K620M faster than the MX230 in OpenCL?

A: Yes, according to Geekbench OpenCL, the K620M scores 5957 versus the MX230's 5739, a 3.7% lead. This is the only direct benchmark comparison available.

Q: Does the MX230 support newer graphics APIs?

A: The MX230 supports DirectX 12 (12_1), while the K620M only supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Q: Why does the MX230 have a higher average benchmark score if it loses the OpenCL test?

A: The MX230's average of 6077 includes a strong Geekbench Vulkan score of 6414, which raises its average above its OpenCL result. The K620M's average of 5957 is based solely on its OpenCL score.

Q: Which GPU is more power-efficient?

A: The MX230 has a 10 W TDP, which is three times lower than the K620M's 30 W TDP. This makes the MX230 better suited for compact, battery-powered notebooks.

Q: Are these GPUs still in production?

A: No, both are end-of-life. The MX230 was released in February 2019, while the K620M was released in February 2015.

The Verdict

The data points to a split decision based on workload and physical constraints. If you are choosing for a modern ultraportable laptop, the MX230 is the better option because of its 10 W TDP, GDDR5 memory with 48.06 GB/s bandwidth, and DirectX 12_1 support. Its higher average benchmark score of 6077, driven by a 6414 Vulkan score, suggests it handles modern API workloads well. The 14 nm process and 24.3M transistors per mm² density also indicate better efficiency per transistor.

If you are working with an existing MXM-based mobile workstation and need the highest OpenCL compute score, the K620M wins the direct comparison. Its 5957 OpenCL score beats the MX230 by 3.7%, and its 863.2 GFLOPS FP32 output is higher than the MX230's 783.9 GFLOPS. The 384 shading units give it more parallel compute capacity, even if the memory bandwidth is severely limited at 16.02 GB/s. The K620M's 30 W TDP is a drawback, but in a workstation chassis with adequate cooling, that is a manageable trade-off.

Neither GPU is a performance leader; both sit at the 34th-35th percentile of all GPUs. The MX230's average score places it within 1% of the RTX A400 and Quadro P2000, while the K620M trades blows with older AMD Radeon HD 8000M series parts. For any modern gaming or professional workload, you would want something far stronger. But among these two legacy parts, the MX230 is the more balanced choice for general use due to its efficiency and memory bandwidth, while the K620M is the pick only if OpenCL compute is your sole metric and you have the power budget for it.

Specification Differences

| Field | NVIDIA GeForce MX230 | NVIDIA Quadro K620M |

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

| Chip | GP108 | GM108S |

| Architecture | Pascal | Maxwell |

| Generation | GeForce MX (2xx) | Quadro Kepler-M (Kx200M) |

| Process Node | 14 nm | 28 nm |

| Foundry | Samsung | TSMC |

| Transistors | 1,800 million | 1,020 million |

| Die Size | 74 mm² | 77 mm² |

| Transistor Density | 24.3M / mm² | 13.2M / mm² |

| Base Clock | 1519 MHz | 1029 MHz |

| Boost Clock | 1531 MHz | 1124 MHz |

| Memory Clock | 1502 MHz (6 Gbps effective) | 1001 MHz (2 Gbps effective) |

| Memory Type | GDDR5 | DDR3 |

| Memory Size | 2 GB | 2 GB |

| Memory Bandwidth | 48.06 GB/s | 16.02 GB/s |

| Shading Units | 256 | 384 |

| TMUs | 16 | 16 |

| ROPs | 16 | 8 |

| Pixel Rate | 24.50 GPixel/s | 8.992 GPixel/s |

| Texture Rate | 24.50 GTexel/s | 17.98 GTexel/s |

| FP32 | 783.9 GFLOPS | 863.2 GFLOPS |

| FP16 | 12.25 GFLOPS (1:64) | None |

| TDP | 10 W | 30 W |

| Slot Width | IGP | MXM Module |

| Bus Interface | PCIe 3.0 x4 | MXM-A (3.0) |

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

| Release Date | 2019-02-20 | 2015-02-28 |