Intel Arc B570 vs NVIDIA GeForce RTX 2060 SUPER Comparison

Intel Arc B570 and NVIDIA GeForce RTX 2060 SUPER are two graphics cards from different eras, and the benchmark data reflects a clear split in their strengths. The Arc B570, built on Intel’s Xe2-HPG architecture, is a newer design with a smaller process node and higher clock speeds, while the RTX 2060 SUPER is a Turing-generation card with a wider memory bus and more RT cores. In the recorded head-to-head tests, each card wins exactly five tests out of ten, making this a genuinely balanced comparison that comes down to workload type. The Arc B570 takes the lead in modern DirectX 12 and compute-heavy tasks, while the RTX 2060 SUPER dominates older DirectX versions and 2D performance. Below is a breakdown of where each card wins, followed by architectural context, specifications, and benchmark-by-benchmark analysis.

Where Each One Wins

The Intel Arc B570 is the stronger choice for modern graphics APIs and compute workloads. In 3DMark Steel Nomad, a DirectX 12 test, the Arc B570 scores 2649 against 2011 for the RTX 2060 SUPER, a 31.7% advantage. It also wins Geekbench Vulkan with 96844 points versus 77402, a 25.1% lead, and Geekbench OpenCL with 83514 versus 76957, an 8.5% margin. In PassMark DirectX 12, the Arc B570 wins 72 to 61, an 18% advantage. For GPU compute, the Arc B570 scores 7281 against 6721, a smaller 8.3% win. These results show that the Arc B570 is built for modern DX12 titles and general compute tasks, where its newer architecture and higher shader throughput pay off.

The NVIDIA GeForce RTX 2060 SUPER, by contrast, wins in older DirectX versions and 2D tests. In PassMark DirectX 10, it scores 111 versus 65, a 41.4% lead, which is the largest margin in either direction across all tests. It also wins PassMark DirectX 9 with 218 against 164, a 24.8% advantage, and PassMark DirectX 11 with 130 versus 118, a 9.2% margin. In PassMark G2D, the RTX 2060 SUPER scores 854 against 661, a 22.6% win. Its PassMark G3D score is also higher at 16462 versus 14195, a 13.8% lead. This pattern suggests the RTX 2060 SUPER retains an edge in legacy DirectX titles and 2D rendering, while the Arc B570 pulls ahead in modern APIs.

Architecture Differences

The two cards are built on fundamentally different process nodes and architectures. The Intel Arc B570 uses the BMG-G21 chip with Xe2-HPG architecture, fabricated on a 5 nm process at TSMC. It packs 19,600 million transistors into a 272 mm² die, giving a transistor density of 72.1 million per mm². The NVIDIA GeForce RTX 2060 SUPER uses the TU106 chip with Turing architecture, on a 12 nm process also at TSMC. It has 10,800 million transistors on a much larger 445 mm² die, with a transistor density of only 24.3 million per mm². This difference in process node and die size explains why the Arc B570 achieves higher clock speeds and throughput despite having a smaller physical footprint.

Core configuration also differs. The Arc B570 has 2304 shading units, 144 texture mapping units, and 80 render output units, with 18 ray tracing cores. The RTX 2060 SUPER has 2176 shading units, 136 TMUs, and 64 ROPs, with 34 ray tracing cores and 272 tensor cores. The Arc B570 has more shading units and ROPs, which contributes to its higher pixel rate of 200.0 GPixel/s versus 105.6 GPixel/s, and texture rate of 360.0 GTexel/s versus 224.4 GTexel/s. The RTX 2060 SUPER has more ray tracing cores and tensor cores, which are relevant for ray tracing and AI-accelerated features, though the raw FP32 throughput of the Arc B570 is higher at 11.52 TFLOPS versus 7.181 TFLOPS. FP16 performance is similarly higher on the Arc B570 at 23.04 TFLOPS versus 14.36 TFLOPS, both at a 2:1 ratio.

Memory subsystems also differ. The Arc B570 has 10 GB of GDDR6 on a 160-bit bus, with 380.0 GB/s bandwidth. The RTX 2060 SUPER has 8 GB of GDDR6 on a 256-bit bus, with 448.0 GB/s bandwidth. The RTX 2060 SUPER has a wider bus and higher bandwidth, but the Arc B570 offers more total memory capacity. The Arc B570 has a base and boost clock of 2500 MHz, while the RTX 2060 SUPER runs at 1470 MHz base and 1650 MHz boost. Memory speed is also higher on the Arc B570 at 2375 MHz (19 Gbps effective) versus 1750 MHz (14 Gbps effective) for the RTX 2060 SUPER.

FAQ

Q: Which card is faster in DirectX 12?

A: The Intel Arc B570 wins in PassMark DirectX 12 with a score of 72 against 61 for the RTX 2060 SUPER, an 18% advantage. It also wins 3DMark Steel Nomad, a DX12 test, with 2649 versus 2011, a 31.7% margin.

Q: Which card has better compute performance?

A: The Intel Arc B570 leads in PassMark GPU Compute with 7281 points against 6721, an 8.3% win. It also wins Geekbench OpenCL with 83514 versus 76957, an 8.5% margin.

Q: Does the RTX 2060 SUPER win in any tests?

A: Yes. The RTX 2060 SUPER wins PassMark DirectX 10, DirectX 9, DirectX 11, G2D, and G3D. Its largest win is in DirectX 10 with 111 versus 65, a 41.4% lead, and its G3D score is 16462 versus 14195, a 13.8% margin.

Q: How much memory does each card have?

A: The Intel Arc B570 has 10 GB of GDDR6 on a 160-bit bus with 380.0 GB/s bandwidth. The NVIDIA GeForce RTX 2060 SUPER has 8 GB of GDDR6 on a 256-bit bus with 448.0 GB/s bandwidth.

Q: What is the average benchmark score for each card?

A: The Intel Arc B570 has an average benchmark score of 20556, while the RTX 2060 SUPER has an average of 18093. This places the Arc B570 in the 65th percentile of all GPUs, and the RTX 2060 SUPER in the 62nd percentile.

Q: Which card has more ray tracing cores?

A: The NVIDIA GeForce RTX 2060 SUPER has 34 ray tracing cores, while the Intel Arc B570 has 18. The RTX 2060 SUPER also has 272 tensor cores, which the Arc B570 does not list.

Specification Differences

The two cards differ in nearly every major specification. The Intel Arc B570 uses a 5 nm process with 19,600 million transistors on a 272 mm² die, while the RTX 2060 SUPER uses a 12 nm process with 10,800 million transistors on a 445 mm² die. Clock speeds are 2500 MHz base and boost for the Arc B570, versus 1470 MHz base and 1650 MHz boost for the RTX 2060 SUPER. Memory capacity is 10 GB versus 8 GB, bus width is 160-bit versus 256-bit, and bandwidth is 380.0 GB/s versus 448.0 GB/s. The Arc B570 has 2304 shading units, 144 TMUs, 80 ROPs, and 18 RT cores, while the RTX 2060 SUPER has 2176 shading units, 136 TMUs, 64 ROPs, 34 RT cores, and 272 tensor cores. Pixel rate is 200.0 GPixel/s versus 105.6 GPixel/s, texture rate is 360.0 GTexel/s versus 224.4 GTexel/s, and FP32 is 11.52 TFLOPS versus 7.181 TFLOPS. FP16 is 23.04 TFLOPS versus 14.36 TFLOPS, both at 2:1. TDP is 150 W for the Arc B570 and 175 W for the RTX 2060 SUPER. The Arc B570 uses PCIe 4.0 x8, while the RTX 2060 SUPER uses PCIe 3.0 x16. Display outputs differ: the Arc B570 has 1x HDMI 2.1a and 3x DisplayPort 2.1, while the RTX 2060 SUPER has 1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, and 1x USB Type-C. Dimensions also differ, with the Arc B570 measuring 272 mm in length and 115 mm in height, versus 229 mm length, 113 mm height, and 35 mm width for the RTX 2060 SUPER. The Arc B570 launched with a 219 USD launch MSRP, while the RTX 2060 SUPER launched with a 399 USD launch MSRP.

Head-to-Head Benchmarks

The largest win for the Intel Arc B570 is in 3DMark Steel Nomad, where it scores 2649 against 2011, a 31.7% margin. This is a decisive victory in a modern DX12 benchmark and highlights the architectural advantage of the newer Xe2-HPG design. The second-largest win is in Geekbench Vulkan, where the Arc B570 scores 96844 versus 77402, a 25.1% lead. In Geekbench OpenCL, the Arc B570 wins 83514 to 76957, an 8.5% margin. PassMark DirectX 12 also goes to the Arc B570 with 72 versus 61, an 18% win, and PassMark GPU Compute goes to the Arc B570 with 7281 versus 6721, an 8.3% margin.

The NVIDIA GeForce RTX 2060 SUPER has its largest win in PassMark DirectX 10, scoring 111 against 65, a 41.4% lead. This is the single biggest delta in either direction across all tests. It also wins PassMark DirectX 9 with 218 versus 164, a 24.8% margin, and PassMark G2D with 854 versus 661, a 22.6% lead. In PassMark G3D, the RTX 2060 SUPER scores 16462 versus 14195, a 13.8% win, and in PassMark DirectX 11 it wins 130 versus 118, a 9.2% margin. These results show that the RTX 2060 SUPER is still competitive in legacy DirectX workloads and 2D performance, but it lags in modern APIs and compute.

The Verdict

Based on the recorded data, the Intel Arc B570 is the better choice for modern DirectX 12 gaming and compute-heavy workloads. Its 31.7% lead in 3DMark Steel Nomad and 25.1% lead in Geekbench Vulkan demonstrate clear superiority in current-generation graphics APIs. It also offers more memory at 10 GB versus 8 GB, higher clock speeds, and a much smaller process node. The average benchmark score of 20556 versus 18093, along with its 65th percentile placement versus 62nd for the RTX 2060 SUPER, confirms that it is the faster card overall in the database's aggregated measurements.

The NVIDIA GeForce RTX 2060 SUPER is still the pick for users who rely on older DirectX versions or 2D rendering tasks. Its wins in DirectX 10, DirectX 9, and DirectX 11, plus a higher G3D score, make it a viable option for legacy software. However, its 12 nm process, older Turing architecture, and lower average benchmark score put it at a disadvantage in modern workloads. For a builder choosing between these two today, the data points squarely to the Arc B570 for forward-looking systems, while the RTX 2060 SUPER remains relevant only for specific older applications.