Intel Arc A380 vs NVIDIA GeForce GTX 650 Ti Boost Comparison

The Intel Arc A380 and NVIDIA GeForce GTX 650 Ti Boost represent two very different eras of GPU design. The data places them in adjacent performance percentiles — the Arc A380 at the 44th percentile and the GTX 650 Ti Boost at the 42nd — but their benchmark results reveal a generational gulf. This analysis compares the two strictly on the provided benchmark scores and architectural specifications.

Head-to-Head Benchmarks

The head-to-head data includes only two common tests, and the Intel Arc A380 wins both decisively. In Geekbench OpenCL, the Arc A380 scores 38,224 against the GTX 650 Ti Boost’s 9,302, a 310.9% advantage. This is not a marginal win; it is a three-fold-plus increase in raw compute throughput as measured by that workload. The Vulkan result is similarly lopsided: the Arc A380 posts 36,736 versus 10,041, a 265.9% delta. For any modern API-based workload, the Arc A380 is in a different performance class.

Looking beyond the head-to-head table, the Arc A380’s broader benchmark suite shows consistent strength. Its Passmark DirectX 11 score is 38, while its DirectX 9 score is 73; the GTX 650 Ti Boost has no Passmark entries in the provided data, so direct comparison there is impossible. However, the Arc A380’s average benchmark score of 8,558 versus the GTX 650 Ti Boost’s 8,067 tells the story — a 5.7% overall average deficit for the older card. The Arc A380 also has a higher percentile ranking across all GPUs (44th versus 42nd), confirming that its wins are not isolated to one test.

The GTX 650 Ti Boost does have one unique benchmark entry: Geekbench Metal, where it scores 4,858. The Arc A380 has no Metal score in the data, so this is a niche advantage only on Apple’s API, which is irrelevant for a Windows-focused discrete GPU. In every overlapping test, the Arc A380 wins by a wide margin. The data shows zero wins for the GTX 650 Ti Boost in the head-to-head table.

Where Each One Wins

The Intel Arc A380 wins in every scenario where both cards have data. Its OpenCL and Vulkan scores indicate strong compute and modern API performance, making it the clear choice for any workload that uses those interfaces. The Passmark suite for the Arc A380 shows balanced results across DirectX 10 (37), DirectX 11 (38), and DirectX 12 (35), with a much higher DirectX 9 score of 73. This suggests the card handles older legacy APIs surprisingly well, though the modern API scores are more relevant for current games.

The GTX 650 Ti Boost’s only claim to a win is the Geekbench Metal score of 4,858, which exists solely because the Arc A380 lacks a Metal result. In practical terms, this is not a use case for either card on typical desktop systems. The GTX 650 Ti Boost also has a lower average benchmark score and a lower percentile, so it cannot be recommended for any modern workload where the Arc A380 is available.

Where the GTX 650 Ti Boost might still be relevant is in legacy system compatibility. It uses a PCIe 3.0 x16 interface, while the Arc A380 uses PCIe 4.0 x8. For very old motherboards without PCIe 4.0 support, the GTX 650 Ti Boost could be a drop-in replacement. The data does not quantify this, but the bus interface difference is a real specification distinction. Additionally, the GTX 650 Ti Boost has a lower transistor density (11.5M / mm² versus 45.9M / mm²), which historically correlates with simpler driver requirements, though no driver data is provided here.

Architecture Differences

The architectural gap is enormous. The Intel Arc A380 uses the DG2-128 chip on the Xe-HPG architecture, built on a 6 nm TSMC process. It contains 7,200 million transistors on a 157 mm² die, yielding a transistor density of 45.9M / mm². The NVIDIA card uses the GK106 chip on the much older Kepler architecture, fabricated on a 28 nm TSMC process. It has 2,540 million transistors on a 221 mm² die, for a density of only 11.5M / mm². This means the Arc A380 packs nearly three times the transistors into a smaller physical area, explaining its massive performance lead.

Core counts differ significantly. The Arc A380 has 1,024 shading units, 64 TMUs, and 32 ROPs, plus 8 dedicated ray tracing cores. The GTX 650 Ti Boost has 768 shading units, 64 TMUs, and 24 ROPs, with no ray tracing hardware. The Arc A380 also supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the GTX 650 Ti Boost is limited to DirectX 12 (11_0) and Vulkan 1.2.175. This means the Intel card can handle hardware ray tracing and the latest graphics features, while the NVIDIA card cannot.

Memory is another major divide. The Arc A380 has 6 GB of GDDR6 on a 96-bit bus, providing 186.0 GB/s of bandwidth. The GTX 650 Ti Boost has 2 GB of GDDR5 on a 192-bit bus, with 144.2 GB/s. The Arc A380’s higher capacity is critical for modern textures, while its narrower bus is compensated by faster memory clocks — 1937 MHz (15.5 Gbps effective) versus 1502 MHz (6 Gbps effective). The Arc A380’s pixel rate is 65.60 GPixel/s versus 16.51 GPixel/s for the GTX 650 Ti Boost, and its texture rate is 131.2 GTexel/s versus 66.05 GTexel/s.

Power and physical specs also differ. The Arc A380 has a 75 W TDP with a 1x 8-pin power connector and a suggested 250 W PSU. The GTX 650 Ti Boost has a 134 W TDP with a 1x 6-pin connector and a suggested 300 W PSU. Despite the GTX 650 Ti Boost drawing more power, it delivers far less performance. The Arc A380 is also shorter at 222 mm versus 241 mm for the NVIDIA card, and it supports modern display outputs: 1x HDMI 2.1 and 3x DisplayPort 2.0, whereas the GTX 650 Ti Boost offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2.

The Verdict

The data is unambiguous: the Intel Arc A380 is the superior GPU in every measurable way. Its 310.9% lead in OpenCL and 265.9% lead in Vulkan are not close calls. Its average benchmark score is 491 points higher, and its percentile rank is two points higher. The GTX 650 Ti Boost is an end-of-life product from 2013, while the Arc A380 is also end-of-life but from 2022, and that nine-year gap is visible in every specification.

Who should pick the GTX 650 Ti Boost? Only someone with a legacy system that cannot support PCIe 4.0 and requires a dual-DVI output setup, as the NVIDIA card offers 2x DVI ports. Even then, the GTX 650 Ti Boost’s 2 GB memory and 144.2 GB/s bandwidth will bottleneck any modern workload. The Arc A380’s 6 GB memory and 186.0 GB/s bandwidth are simply more future-proof.

Who should pick the Arc A380? Anyone building a low-power, modern system. Its 75 W TDP is nearly half the GTX 650 Ti Boost’s 134 W, yet it delivers over three times the compute performance. Its support for DirectX 12 Ultimate and ray tracing cores makes it viable for contemporary games at low settings. The Arc A380’s launch MSRP is 149 USD, while the GTX 650 Ti Boost launched at 169 USD. Given the performance delta, the Arc A380 is the only rational choice from the data.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The Intel Arc A380 has an average benchmark score of 8,558, which is 491 points higher than the NVIDIA GeForce GTX 650 Ti Boost’s 8,067.

Q: How much faster is the Arc A380 in Vulkan performance?

A: The Arc A380 scores 36,736 in Geekbench Vulkan, which is 265.9% higher than the GTX 650 Ti Boost’s 10,041.

Q: Does the GTX 650 Ti Boost support ray tracing?

A: No, the GTX 650 Ti Boost has no ray tracing cores, while the Arc A380 has 8 RT cores and supports DirectX 12 Ultimate.

Q: What is the memory capacity difference?

A: The Arc A380 has 6 GB of GDDR6 memory, while the GTX 650 Ti Boost has 2 GB of GDDR5 memory. The Arc A380 also has higher bandwidth at 186.0 GB/s versus 144.2 GB/s.

Q: Which card has a lower power draw?

A: The Arc A380 has a 75 W TDP, which is 59 W lower than the GTX 650 Ti Boost’s 134 W TDP.

Q: Are there any benchmarks where the GTX 650 Ti Boost wins?

A: In the head-to-head benchmark table, the GTX 650 Ti Boost wins 0 tests. It has a Geekbench Metal score of 4,858, but the Arc A380 has no Metal score in the data, so no comparison is possible.

Specification Differences

| Specification | Intel Arc A380 | NVIDIA GeForce GTX 650 Ti Boost |

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

| Chip | DG2-128 | GK106 |

| Architecture | Xe-HPG | Kepler |

| Process Node | 6 nm | 28 nm |

| Transistors | 7,200 million | 2,540 million |

| Die Size | 157 mm² | 221 mm² |

| Base Clock | 2000 MHz | 980 MHz |

| Boost Clock | 2050 MHz | 1032 MHz |

| Memory Clock | 1937 MHz (15.5 Gbps effective) | 1502 MHz (6 Gbps effective) |

| Memory Size | 6 GB | 2 GB |

| Memory Type | GDDR6 | GDDR5 |

| Memory Bus | 96 bit | 192 bit |

| Bandwidth | 186.0 GB/s | 144.2 GB/s |

| Shading Units | 1024 | 768 |

| TMUs | 64 | 64 |

| ROPs | 32 | 24 |

| RT Cores | 8 | 0 |

| Pixel Rate | 65.60 GPixel/s | 16.51 GPixel/s |

| Texture Rate | 131.2 GTexel/s | 66.05 GTexel/s |

| FP32 | 4.198 TFLOPS | 1.585 TFLOPS |

| FP16 | 8.397 TFLOPS (2:1) | N/A |

| TDP | 75 W | 134 W |

| Power Connectors | 1x 8-pin | 1x 6-pin |

| Suggested PSU | 250 W | 300 W |

| Bus Interface | PCIe 4.0 x8 | PCIe 3.0 x16 |

| Display Outputs | 1x HDMI 2.1, 3x DisplayPort 2.0 | 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| DirectX | 12 Ultimate (12_2) | 12 (11_0) |

| Vulkan | 1.4 | 1.2.175 |

| Length | 222 mm (8.7 inches) | 241 mm (9.5 inches) |