AMD Radeon RX 6650 XT vs Intel Arc A770M Comparison

The AMD Radeon RX 6650 XT and Intel Arc A770M represent two different philosophies in GPU design: a desktop-focused RDNA 2 part versus a mobile-oriented Xe-HPG solution. Benchmark data shows a clear split, with the AMD card winning 7 of 10 head-to-head tests, yet the Intel part dominates in raw compute and modern DX12 workloads. The average benchmark scores are close (19,765 vs 18,383), but the distribution of wins reveals fundamental architectural priorities.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon RX 6650 XT scores 19,765 on average, while the Intel Arc A770M scores 18,383. This places the AMD card at the 65th percentile of all GPUs, versus the Intel card at the 62nd percentile.

Q: How large is the performance gap in the most modern DX12 test?

A: In 3DMark Steel Nomad DX12, the Intel Arc A770M scores 2,278 versus the AMD’s 1,859. This represents an 18.4% advantage for Intel, the largest single-test margin in its favor.

Q: What is the biggest win for the AMD card?

A: The RX 6650 XT leads by 144.9% in Passmark DirectX 11, scoring 169 versus Intel’s 69. This is the most lopsided result in the entire comparison.

Q: Does the Intel card have any compute advantage?

A: Yes, decisively. In Geekbench OpenCL, the Intel Arc A770M scores 89,494 versus AMD’s 10,111 — an 88.7% difference. This suggests vastly different compute throughput characteristics.

Q: Which card has more memory and bandwidth?

A: The Intel Arc A770M features 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth. The AMD RX 6650 XT has 8 GB on a 128-bit bus with 280.3 GB/s.

Q: Are both cards still in production?

A: No. Both the AMD Radeon RX 6650 XT and Intel Arc A770M are marked as end-of-life in the production status field.

Architecture Differences

The AMD Radeon RX 6650 XT uses the Navi 23 chip built on RDNA 2.0 architecture, manufactured on TSMC’s 7 nm process. The die measures 237 mm² and packs 11,060 million transistors, yielding a density of 46.7 million transistors per square millimeter. In contrast, the Intel Arc A770M employs the DG2-512 chip with Xe-HPG architecture on a 6 nm TSMC node. Its die is substantially larger at 406 mm², accommodating 21,700 million transistors at a density of 53.4M per mm².

The compute configuration diverges sharply. The AMD card has 2,048 shading units, 128 texture mapping units, and 64 raster output units. The Intel card doubles these counts: 4,096 shading units, 256 TMUs, and 128 ROPs. Both feature 32 ray tracing cores, but the Intel part’s raw throughput is higher — 16.79 TFLOPS FP32 versus AMD’s 10.79 TFLOPS. The AMD card compensates with higher clocks: 2,055 MHz base and 2,635 MHz boost, compared to Intel’s 1,650 MHz base and 2,050 MHz boost.

Memory architecture also differs fundamentally. The Intel Arc A770M offers 16 GB of GDDR6 on a 256-bit bus, delivering 512.0 GB/s. The AMD RX 6650 XT provides 8 GB on a 128-bit bus at 280.3 GB/s. The Intel card’s PCIe interface is wider (x16 versus x8), though both use PCIe 4.0. Power characteristics are opposite: the AMD card draws 176 W and requires a 1x 8-pin connector with a 450 W suggested PSU, while the Intel card is rated at 120 W with an IGP slot width and no external power connector listed.

Head-to-Head Benchmarks

The benchmark results paint a nuanced picture. In 3DMark Steel Nomad DX12, the Intel Arc A770M wins with 2,278 points against AMD’s 1,859, a margin of 18.4%. This is a modern, demanding workload where Intel’s architectural strengths emerge clearly.

Geekbench OpenCL shows the most extreme divergence. Intel scores 89,494 versus AMD’s 10,111 — an 88.7% advantage. This massive gap likely stems from the Intel card’s doubled shader count and higher FP32 throughput (16.79 vs 10.79 TFLOPS). However, in Geekbench Vulkan, the tables turn: AMD wins 81,306 to 74,422, a 9.2% lead. This suggests AMD’s driver or architecture handles Vulkan’s explicit control model more efficiently.

The Passmark suite reveals a pattern of AMD dominance in legacy and general 3D workloads. In DirectX 10, AMD leads 109 to 56 (94.6%). DirectX 11 shows AMD ahead 169 to 69 (144.9%). DirectX 9 gives AMD 215 versus 178 (20.8%). The 2D test also favors AMD at 918 to 711 (29.1%). Most tellingly, Passmark G3D shows AMD at 17,166 versus Intel’s 11,774 — a 45.8% advantage. The GPU compute test continues this trend: AMD scores 7,532 against Intel’s 4,778, a 57.6% lead.

The single exception in Passmark is DirectX 12, where Intel wins 70 to 62 (11.4%). This small margin, combined with the larger Steel Nomad win, suggests Intel’s DX12 implementation is solid while its older API paths lag significantly.

Specification Differences

| Specification | AMD Radeon RX 6650 XT | Intel Arc A770M |

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

| Process Node | 7 nm | 6 nm |

| Transistors | 11,060 million | 21,700 million |

| Die Size | 237 mm² | 406 mm² |

| Transistor Density | 46.7M / mm² | 53.4M / mm² |

| Base Clock | 2055 MHz | 1650 MHz |

| Boost Clock | 2635 MHz | 2050 MHz |

| Memory Size | 8 GB | 16 GB |

| Memory Bus | 128 bit | 256 bit |

| Memory Bandwidth | 280.3 GB/s | 512.0 GB/s |

| Shading Units | 2048 | 4096 |

| TMUs | 128 | 256 |

| ROPs | 64 | 128 |

| Pixel Rate | 168.6 GPixel/s | 262.4 GPixel/s |

| Texture Rate | 337.3 GTexel/s | 524.8 GTexel/s |

| FP32 Performance | 10.79 TFLOPS | 16.79 TFLOPS |

| FP16 Performance | 21.59 TFLOPS (2:1) | 33.59 TFLOPS (2:1) |

| TDP | 176 W | 120 W |

| Slot Width | Dual-slot | IGP |

| Power Connectors | 1x 8-pin | None |

| Suggested PSU | 450 W | None |

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

| Display Outputs | 1x HDMI 2.1, 3x DisplayPort 1.4a | Portable Device Dependent |

The AMD card’s launch MSRP was 399 USD, while the Intel card has no listed launch MSRP.

The Verdict

The data indicates that the AMD Radeon RX 6650 XT is the stronger all-around performer for most gaming and general graphics workloads. It wins 7 of 10 head-to-head benchmarks, including all legacy DirectX tests, both Passmark 3D tests, and Vulkan. Its average benchmark score is 7.5% higher than the Intel card’s. The nearest rival data confirms this: AMD’s closest competitors (NVIDIA Tesla K40m at -0.6%, AMD FirePro W7000 at -0.7%) are within 1% of its score, while Intel’s rivals (AMD Radeon RX 460 at 0.1%, NVIDIA GeForce RTX 3060 Mobile at 1.2%) are similarly close but at a lower absolute level.

However, the Intel Arc A770M is not without merit. Its 18.4% win in 3DMark Steel Nomad DX12 and 88.7% margin in Geekbench OpenCL indicate superior raw compute and modern API performance. The 16 GB memory capacity and 512.0 GB/s bandwidth are double the AMD card’s specifications, though these do not translate into consistent benchmark wins. The Intel card also operates at a lower TDP (120 W versus 176 W), making it a more power-efficient choice for its intended mobile form factor.

Where Each One Wins

The AMD Radeon RX 6650 XT is the clear choice for users prioritizing legacy API compatibility and traditional 3D rendering. Its wins in DirectX 9 (20.8%), DirectX 10 (94.6%), DirectX 11 (144.9%), and Passmark G3D (45.8%) demonstrate strong driver optimization for established game engines. The Vulkan advantage (9.2%) further cements its position for cross-platform titles. The 2D performance lead (29.1%) suggests better desktop and UI responsiveness. This card suits gamers running a broad mix of older and current titles, especially those using Vulkan or DirectX 11.

The Intel Arc A770M excels in compute-heavy and forward-looking scenarios. Its Geekbench OpenCL score (89,494) is nearly nine times higher than AMD’s, indicating massive parallel compute headroom for non-graphics workloads like machine learning inference or scientific calculations. The 3DMark Steel Nomad DX12 win (18.4%) shows modern API efficiency, and the DirectX 12 Passmark advantage (11.4%) reinforces this. With double the memory (16 GB versus 8 GB) and bandwidth (512.0 GB/s versus 280.3 GB/s), it is better equipped for large textures, high-resolution assets, or future game titles that demand more VRAM. The lower TDP (120 W versus 176 W) also makes it suitable for thinner, more power-constrained laptops.

In summary, the AMD card dominates the benchmark suite overall, but the Intel card’s strengths in compute and modern APIs — plus its memory and power advantages — make it a compelling option for specific workloads. The choice depends on whether the user prioritizes broad game compatibility and established performance (AMD) or raw compute density and future-proofing (Intel).