Intel Arc A770M vs NVIDIA GeForce GTX 1060 3 GB Comparison
Intel Arc A770M
GeForce GTX 1060 3 GB
PERFORMANCE BENCHMARKS
Analysis: Intel Arc A770M vs NVIDIA GeForce GTX 1060 3 GB
# Where Each One Wins
The Intel Arc A770M is the clear winner in every head-to-head benchmark recorded, taking both available tests outright. The NVIDIA GeForce GTX 1060 3 GB fails to secure a single win in direct comparisons, with the Intel part dominating in both compute and graphics API workloads. In the Geekbench OpenCL test, the Arc A770M scores 89,494 against the GTX 1060's 27,001, a staggering 69.8% advantage. The Vulkan result tells a similar story: 74,422 for Intel versus 17,475 for NVIDIA, a 76.5% lead. These are not marginal victories — they represent generational gaps in raw throughput and modern API efficiency.
Looking at the broader benchmark landscape, the GTX 1060 3 GB posts an average benchmark score of 19,334, placing it in the 64th percentile of all GPUs. The Arc A770M averages 18,383, sitting at the 62nd percentile. Despite the Intel part winning every head-to-head test, the aggregate scores are remarkably close — the GTX 1060 actually edges ahead by roughly 5% in average score. This paradox stems from the test sets: the GTX 1060's three Geekbench scores (Metal, OpenCL, Vulkan) are all competitive with its peer group, while the Arc A770M's ten tests include several legacy DirectX and 2D workloads where it underperforms. The Intel GPU's lead in modern compute is real, but its overall profile is uneven across older APIs.
For use-case selection, the Arc A770M wins decisively for OpenCL compute and Vulkan gaming workloads. The GTX 1060 3 GB, however, demonstrates stronger consistency across its tested applications, with no catastrophic failures in any benchmark. The data suggests the Arc A770M is the choice for users prioritizing modern API performance and raw compute throughput, while the GTX 1060 offers more balanced results across a wider variety of legacy and contemporary test scenarios.
# Architecture Differences
The two GPUs represent fundamentally different design philosophies separated by six years of silicon evolution. The NVIDIA GeForce GTX 1060 3 GB uses the GP106 chip built on TSMC's 16 nm process, packing 4,400 million transistors into a 200 mm² die. This yields a transistor density of 22.0 million per square millimeter. The Intel Arc A770M employs the DG2-512 chip on a much more advanced 6 nm TSMC node, integrating 21,700 million transistors across a 406 mm² die — a density of 53.4 million per square millimeter, more than double that of the NVIDIA part.
The architectural lineage diverges sharply. The GTX 1060 is based on Pascal, the GeForce 10-series architecture from 2016, supporting DirectX 12 (12_1) and OpenGL 4.6 with Vulkan 1.4. The Arc A770M uses Intel's Xe-HPG architecture from the Alchemist generation (Arc 7 Mobile), featuring DirectX 12 Ultimate (12_2) support alongside the same OpenGL 4.6 and Vulkan 1.4 APIs. The Intel part also includes 32 dedicated ray tracing cores, a feature entirely absent from the Pascal-based NVIDIA GPU.
Memory configurations highlight the generational divide. The GTX 1060 3 GB ships with 3 GB of GDDR5 on a 192-bit bus, delivering 192.2 GB/s of bandwidth. The Arc A770M offers 16 GB of GDDR6 across a 256-bit interface, quadrupling capacity while more than doubling bandwidth to 512.0 GB/s. This 2.66x bandwidth advantage directly explains the Intel part's dominance in compute-heavy workloads.
The compute resources are similarly lopsided. The GTX 1060 fields 1,152 shading units, 72 texture mapping units, and 48 raster output units, achieving 3.935 TFLOPS of FP32 performance. The Arc A770M counters with 4,096 shading units, 256 TMUs, and 128 ROPs, delivering 16.79 TFLOPS FP32 — a 4.27x raw compute advantage. Pixel rate jumps from 81.98 GPixel/s to 262.4 GPixel/s, and texture rate from 123.0 GTexel/s to 524.8 GTexel/s. The Intel part's FP16 performance of 33.59 TFLOPS (2:1 ratio) dwarfs the GTX 1060's 61.49 GFLOPS (1:64 ratio), reflecting a fundamental shift toward mixed-precision workloads.
Clock speeds show the Intel part running higher out of the box: 1650 MHz base and 2050 MHz boost versus 1506 MHz base and 1708 MHz boost for NVIDIA. Both GPUs share a 120 W TDP, but the Arc A770M achieves this in an IGP (integrated graphics package) form factor with no dedicated power connectors, while the GTX 1060 is a dual-slot discrete card requiring a single 6-pin connector and a 300 W suggested PSU. The GTX 1060 uses PCIe 3.0 x16; the Arc A770M uses PCIe 4.0 x16. Display outputs also differ: the GTX 1060 offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.4a, while the Arc A770M's outputs are portable device dependent.
# Head-to-Head Benchmarks
The direct comparison data contains only two tests, but both are decisive Intel victories. In Geekbench OpenCL, the Arc A770M scores 89,494 against the GTX 1060's 27,001, a delta of -69.8% from the Intel part's perspective (meaning the NVIDIA GPU trails by that margin). This 3.31x performance ratio reflects the massive compute resource disparity — the Intel part's 16.79 TFLOPS FP32 and 512.0 GB/s bandwidth simply overwhelm the older Pascal card.
The Geekbench Vulkan test shows an even wider gap. The Arc A770M posts 74,422, while the GTX 1060 manages 17,475, resulting in a -76.5% delta. This 4.26x advantage aligns closely with the FP32 compute ratio, suggesting Vulkan scales nearly linearly with shading resources in these workloads. The Intel part's support for DirectX 12 Ultimate and modern API features likely contributes to its efficiency, though Vulkan 1.4 is present on both cards.
The GTX 1060's best showing comes in the aggregate. Its average benchmark score of 19,334 places it just 0.4% behind the AMD Radeon RX 7900 XTX (19,410) and 0.8% ahead of the NVIDIA TITAN Xp (19,177). This is remarkable for a card that launched at 199 USD in 2016 with only 3 GB of memory. By contrast, the Arc A770M's average of 18,383 sits within 0.1% of the AMD Radeon RX 460 (18,373), 0.8% below the AMD FirePro D500 (18,533), and 1.1% above the AMD Radeon Pro 5700 (18,189). The Intel part actually trails the NVIDIA GeForce RTX 3060 Mobile (18,159) by 1.2%.
The passmark results for the Arc A770M reveal why its average is dragged down. DirectX 9 scores 178, DirectX 10 scores 56, DirectX 11 scores 69, and DirectX 12 scores 70. The G2D test returns 711, while G3D hits 11,774 and GPU compute reaches 4,778. These legacy DirectX scores are anomalously low, suggesting driver maturity issues or architectural inefficiencies in older API paths. The GTX 1060's three Geekbench scores — 13,527 Metal, 27,001 OpenCL, 17,475 Vulkan — show no such weakness, maintaining consistent performance across different API layers.
# FAQ
Q: Which GPU has higher raw compute performance?
A: The Intel Arc A770M delivers 16.79 TFLOPS FP32 and 33.59 TFLOPS FP16, versus 3.935 TFLOPS FP32 and 61.49 GFLOPS FP16 for the GTX 1060 3 GB. The Intel part leads by over 4x in FP32 and over 500x in FP16.
Q: How do memory capacities compare?
A: The Arc A770M offers 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth. The GTX 1060 3 GB has 3 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth. The Intel part provides more than 5x capacity and 2.66x bandwidth.
Q: What explains the GTX 1060's higher average benchmark score despite losing head-to-head?
A: The GTX 1060's average of 19,334 across three tests exceeds the Arc A770M's 18,383 across ten tests. The Intel part's low legacy DirectX scores (56-178 in DX10-DX12) and 2D test (711) drag down its aggregate, while the NVIDIA card performs consistently across Metal, OpenCL, and Vulkan.
Q: Do both GPUs support ray tracing?
A: No. The Arc A770M includes 32 dedicated ray tracing cores, while the GTX 1060 3 GB has no ray tracing hardware at all. The Intel part also supports DirectX 12 Ultimate, whereas the NVIDIA card is limited to DirectX 12 (12_1).
Q: How do the power requirements differ?
A: Both GPUs share a 120 W TDP. The GTX 1060 3 GB requires a single 6-pin power connector and a 300 W suggested PSU, while the Arc A770M is an IGP with no dedicated power connectors and no suggested PSU listed.
Q: Where does each GPU rank among all GPUs?
A: The GTX 1060 3 GB sits in the 64th percentile, with its nearest rival being the AMD Radeon RX 7900 XTX at -0.4% delta. The Arc A770M ranks in the 62nd percentile, with the AMD Radeon RX 460 as its closest competitor at +0.1% delta.
# The Verdict
The data presents a clear split: the Intel Arc A770M is the superior GPU for modern compute and Vulkan workloads, while the NVIDIA GeForce GTX 1060 3 GB holds its own in aggregate benchmark consistency. The Arc A770M wins both head-to-head tests decisively — 69.8% in OpenCL and 76.5% in Vulkan — and offers 4.27x the FP32 throughput, 2.66x the memory bandwidth, and 32 ray tracing cores the NVIDIA part lacks entirely.
However, the GTX 1060 3 GB's 64th percentile ranking versus the Arc A770M's 62nd percentile tells a nuanced story. The NVIDIA card's average score of 19,334 places it within 0.4% of the AMD Radeon RX 7900 XTX, a flagship-class GPU, while the Intel part's 18,383 average sits near the mid-range AMD Radeon RX 460. The Intel GPU's poor legacy DirectX performance (scores of 56-178 across DX10-DX12) suggests driver maturity issues that undermine its modern architectural advantages.
For users prioritizing OpenCL compute, Vulkan gaming, or ray tracing, the Arc A770M is the clear choice based on benchmark results. Its 16 GB memory capacity and 512.0 GB/s bandwidth provide substantial headroom for large datasets and high-resolution textures. For those needing consistent performance across a broad range of APIs — including legacy DirectX titles — the GTX 1060 3 GB offers more balanced results, despite being an end-of-life product from 2016.
The GTX 1060's 199 USD launch MSRP (stated once here) reflects its mainstream positioning, while the Arc A770M's lack of a launch MSRP suggests a different market segment entirely. Ultimately, the Intel part wins on raw capability and modern features, but the NVIDIA card's reliability across diverse workloads keeps it competitive in aggregate scoring. Pick the Arc A770M for future-proof compute and API support; pick the GTX 1060 3 GB for dependable legacy performance.