GPU Comparison

The Intel Arc A770M and NVIDIA GeForce RTX 2060 SUPER represent two very different approaches to high-performance graphics, yet their aggregate benchmark scores place them in the same percentile tier. The data shows the Arc A770M edges out the RTX 2060 SUPER with an average benchmark score of 18383 against 18093, a gap of roughly 1.6%. Both cards sit at the 62nd percentile among all GPUs, making them direct competitors in overall capability. However, the similarity ends at the aggregate level; the individual benchmark results reveal starkly divergent strengths that depend heavily on API generation and workload type.

The Arc A770M’s most decisive victories come in modern, compute-heavy workloads. In 3DMark Steel Nomad DX12, the Intel part scores 2278 versus the NVIDIA card’s 2011, a 13.3% advantage. The gap widens further in Geekbench OpenCL, where the Arc A770M posts 89494 against 76957, a 16.3% lead. These results indicate that Intel’s architecture scales aggressively with raw throughput and modern API features. The Arc also wins in Passmark DirectX 12, scoring 70 versus 61, a 14.8% margin. That is a notable result because it shows the Intel GPU holding its advantage even in a legacy rasterization test, provided it uses the DX12 path.

The RTX 2060 SUPER counters with overwhelming wins in older DirectX APIs. In Passmark DirectX 11, the NVIDIA card scores 130 versus 69, a staggering 46.9% lead. The DirectX 10 result is similarly lopsided: 111 versus 56, a 49.5% advantage. Even in DirectX 9, the RTX 2060 SUPER leads 218 to 178, an 18.3% margin. These are not small differences; they suggest that the older the API, the more dominant the Turing architecture becomes. The RTX 2060 SUPER also wins in Passmark G3D (16462 vs 11774, 28.5% ahead) and Passmark GPU Compute (6721 vs 4778, 28.9% ahead), which are aggregate tests that may weight legacy paths more heavily. The Vulkan result is close but favors NVIDIA: 77402 versus 74422, a 3.9% edge.

Where Each One Wins

The benchmark split is clean: the Intel Arc A770M wins in modern, DX12-centric and compute-heavy scenarios, while the NVIDIA GeForce RTX 2060 SUPER dominates legacy DirectX titles and general-purpose rasterization tests. The Arc A770M’s wins in 3DMark Steel Nomad and Geekbench OpenCL point to a GPU that is built for the future, it rewards applications that leverage explicit multi-threading, async compute, and the full feature set of DX12 Ultimate. The Passmark DX12 win reinforces this: when both cards are constrained to the same modern API, Intel’s hardware pulls ahead by 14.8%.

The RTX 2060 SUPER is the clear pick for backwards compatibility. Its 46.9% lead in DX11 and 49.5% lead in DX10 are the largest margins in the entire comparison. For users running older game engines, emulators, or productivity software that still relies on legacy DirectX paths, the NVIDIA card will deliver significantly better frame rates and stability. The Passmark G3D and GPU Compute results, where NVIDIA leads by 28.5% and 28.9% respectively, suggest that the Turing card also holds an edge in general-purpose compute workloads that are not optimized for Intel’s architecture. The Vulkan result, while narrow, still favors NVIDIA, indicating that Intel’s driver stack has not yet fully closed the gap in that API.

The practical takeaway is that the Arc A770M is a specialist for modern titles, while the RTX 2060 SUPER is a generalist with a strong bias toward established software. If the workload is a new AAA game built on DX12, the Intel card is the performance leader. If the workload includes older titles or mixed API usage, the NVIDIA card is the safer, faster option.

Architecture Differences

The two GPUs are built on fundamentally different silicon. The Intel Arc A770M uses the DG2-512 chip with the Xe-HPG architecture, fabricated on a 6 nm process at TSMC. It packs 21,700 million transistors into a 406 mm² die, yielding a transistor density of 53.4M per mm². The NVIDIA GeForce RTX 2060 SUPER uses the TU106 chip with the Turing architecture, also built at TSMC but on a 12 nm process. It contains 10,800 million transistors on a slightly larger 445 mm² die, resulting in a much lower density of 24.3M per mm².

The core configurations are just as divergent. The Arc A770M has 4096 shading units, 256 texture mapping units, and 128 ROPs. The RTX 2060 SUPER has 2176 shading units, 136 TMUs, and 64 ROPs. Intel’s GPU has nearly twice the shading units and double the TMUs and ROPs. This explains the Arc’s huge theoretical throughput numbers: 16.79 TFLOPS FP32 versus 7.181 TFLOPS for NVIDIA, and a pixel rate of 262.4 GPixel/s versus 105.6 GPixel/s. The texture rate is similarly lopsided at 524.8 GTexel/s for Intel versus 224.4 GTexel/s for NVIDIA. Clock speeds differ as well, with the Arc boosting to 2050 MHz versus 1650 MHz for the RTX 2060 SUPER, though base clocks are closer at 1650 MHz and 1470 MHz respectively.

Ray tracing and tensor hardware tell a different story. The RTX 2060 SUPER has 34 RT cores and 272 tensor cores. The Arc A770M has 32 RT cores and no listed tensor cores. Despite having fewer RT cores, the Intel part’s higher clock and wider shading array may compensate in some ray tracing scenarios, but the absence of dedicated tensor cores means it lacks NVIDIA’s AI-accelerated features, such as DLSS. Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Memory configurations also differ significantly. The Arc A770M ships with 16 GB of GDDR6 on a 256-bit bus, delivering 512.0 GB/s of bandwidth. The RTX 2060 SUPER has 8 GB of GDDR6 on the same 256-bit bus, but tops out at 448.0 GB/s. The Intel card’s memory runs at 2000 MHz (16 Gbps effective) versus 1750 MHz (14 Gbps effective) for NVIDIA. Power consumption is also a major separator: the Arc A770M is rated at 120 W TDP and is designed as an IGP (integrated graphics package) for mobile devices, while the RTX 2060 SUPER is a 175 W dual-slot desktop card requiring a single 8-pin power connector and a 450 W suggested PSU. The NVIDIA card also has a defined physical footprint at 229 mm long, 113 mm tall, and 35 mm wide, whereas the Intel part’s dimensions are listed as portable device dependent.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The Intel Arc A770M has a slightly higher average benchmark score of 18383, compared to 18093 for the NVIDIA GeForce RTX 2060 SUPER. This puts the Intel card 1.6% ahead overall.

Q: Why does the RTX 2060 SUPER win in Passmark DirectX 11 by such a large margin?

A: The RTX 2060 SUPER scores 130 in Passmark DirectX 11 versus 69 for the Arc A770M, a 46.9% lead. This suggests the Turing architecture has far more mature drivers and hardware optimization for legacy DirectX 11 workloads.

Q: Does the Intel Arc A770M have any advantage in modern APIs?

A: Yes. The Arc A770M leads by 13.3% in 3DMark Steel Nomad DX12 and by 14.8% in Passmark DirectX 12. It also wins Geekbench OpenCL by 16.3%, indicating a strong compute advantage in modern workloads.

Q: Which card has more memory bandwidth?

A: The Intel Arc A770M has 512.0 GB/s of bandwidth, while the RTX 2060 SUPER has 448.0 GB/s. The Intel card also has double the memory capacity at 16 GB versus 8 GB.

Q: How do the transistor densities compare?

A: The Intel Arc A770M has a transistor density of 53.4M per mm² on a 6 nm process, while the RTX 2060 SUPER has 24.3M per mm² on a 12 nm process. Intel packs more than twice the density into a slightly smaller die.

Q: What is the power consumption difference?

A: The Arc A770M is rated at 120 W TDP, while the RTX 2060 SUPER is rated at 175 W TDP. The Intel part is designed for mobile integration (IGP), whereas the NVIDIA card is a dual-slot desktop component.

The Verdict

The data presents a clear trade-off. The Intel Arc A770M is the better choice for users running modern DX12 titles, compute-heavy applications, or workloads that benefit from high memory bandwidth and capacity. Its 16 GB frame buffer and 512.0 GB/s bandwidth give it a substantial edge for high-resolution textures and large datasets. Its wins in 3DMark Steel Nomad and Geekbench OpenCL confirm that it is a forward-looking GPU that rewards developers who target the latest APIs. The 120 W TDP also makes it a more power-efficient option, though that is partly a function of its mobile-oriented design.

The NVIDIA GeForce RTX 2060 SUPER is the safer pick for compatibility and legacy software. Its 46.9% lead in DirectX 11 and 49.5% lead in DirectX 10 are the most dramatic margins in the comparison, making it the undisputed king of older game engines and enterprise applications that have not migrated to DX12. The 272 tensor cores also provide dedicated AI acceleration hardware that the Arc A770M lacks entirely. For users who play a mix of old and new games, or who rely on productivity software with legacy rendering paths, the RTX 2060 SUPER will deliver a more consistent experience.

There is no universal winner. The Arc A770M is the performance leader for the next generation of software, while the RTX 2060 SUPER is the performance leader for the current and past generation. The aggregate scores are nearly identical, 18383 versus 18093, but the distribution of those scores could not be more different. Users should choose based on the software they actually run: modern DX12 workloads favor Intel, while anything else favors NVIDIA.