AMD Radeon 680M vs Intel Arc B370 Comparison

Head-to-Head Benchmarks

The recorded data contains a single direct comparison between the AMD Radeon 680M and the Intel Arc B370: the 3DMark Steel Nomad DX12 test. The results are decisive. The Intel Arc B370 scores 1184 points, while the AMD Radeon 680M scores 378 points. The delta is -68.1% from the perspective of the AMD part, meaning the Intel solution delivers a score that is more than three times higher. This is a substantial margin in a modern DirectX 12 workload, and it establishes the Intel Arc B370 as the clear leader in raw graphics performance based on this measurement.

The average benchmark score for each part reinforces the gap. The AMD Radeon 680M carries an average benchmark score of 15270 across its submitted tests, which includes Geekbench OpenCL and Vulkan results in addition to the Steel Nomad run. The Intel Arc B370 has an average benchmark score of 1184, but that figure is based solely on the single Steel Nomad result available in the database. The comparison is therefore somewhat lopsided: the AMD part benefits from a broader set of measurements, while the Intel part has only one recorded test. Still, for the one workload where both are measured under identical conditions, the Intel Arc B370 wins outright.

The AMD Radeon 680M does post strong numbers in other tests. Its Geekbench OpenCL score is 23468, and its Geekbench Vulkan score is 21965. These are not directly comparable to the Intel Arc B370, since the Intel part has no recorded Geekbench results in the database. The difference in test coverage means the 680M's overall average is heavily influenced by those two strong compute results. The Intel Arc B370, by contrast, has no compute-oriented benchmark entries, so its average score reflects only the 3DMark test.

Percentile placement tells a similar story. The AMD Radeon 680M sits at the 57th percentile among all GPUs in the database. The Intel Arc B370 sits at the 5th percentile. This is a striking divergence, and it suggests that while the Intel part wins the one shared test, its overall standing in the database is far lower. The explanation lies in the rival sets. The AMD Radeon 680M's nearest rivals include the NVIDIA GeForce GTX 580 with an average score of 15283 and a delta of -0.1%, the NVIDIA GeForce RTX 2060 with an average score of 15290 and a delta of -0.1%, the NVIDIA GeForce RTX 3050 OEM with an average score of 15199 and a delta of 0.5%, and the AMD Radeon RX 7600 with an average score of 15171 and a delta of 0.7%. These are all desktop-class discrete GPUs, and the 680M lands within a fraction of a percent of each of them. Its integrated design is effectively competitive with those dedicated cards in the database's aggregate scoring.

The Intel Arc B370's nearest rivals are much older parts. The ATI Mobility Radeon HD 5570 scores 1186 with a delta of -0.2%, the ATI Radeon HD 5770 scores 1190 with a delta of -0.5%, the AMD Radeon HD 7650M scores 1192 with a delta of -0.7%, and the AMD FirePro M2000 scores 1168 with a delta of 1.4%. The Intel part's single Steel Nomad score of 1184 places it in the company of legacy hardware from over a decade ago. The 5th percentile rank reflects this positioning. The 680M, by contrast, is surrounded by much more recent and much faster discrete GPUs in the database's aggregate ranking.

Architecture Differences

The two integrated GPUs come from different manufacturing and design schools. The AMD Radeon 680M uses the Rembrandt+ chip built on RDNA 2.0 architecture, produced on a 6 nm process at TSMC. It belongs to the Navi II IGP generation for Rembrandt Mobile. The Intel Arc B370 uses the Panther Lake chip built on Xe3-LPG architecture, produced on a 3 nm process at Intel's own foundry. It belongs to the Arc Graphics-M generation for Panther Lake. The process node gap is significant: 6 nm versus 3 nm. The AMD part is fabricated on an older node, while the Intel part uses a newer one, which typically enables higher transistor density and better power efficiency per unit of area.

The AMD Radeon 680M contains 13,100 million transistors on a die size of 208 mm², yielding a transistor density of 63.0 million transistors per square millimeter. The Intel Arc B370's transistor count and die size are recorded as unknown in the database, so no direct density comparison is possible. The clock behavior also differs. The AMD part has a base clock of 2000 MHz and a boost clock of 2200 MHz. The Intel part has a base clock of 300 MHz and a boost clock of 2400 MHz. The Intel part's base clock is far lower, but its boost clock is higher by 200 MHz. This suggests a wider dynamic range in the Intel design, likely a consequence of aggressive power management on the 3 nm node.

Shader resources favor the Intel part. The Intel Arc B370 has 1280 shading units, 40 texture mapping units, and 20 render output units. The AMD Radeon 680M has 768 shading units, 48 TMUs, and 32 ROPs. Intel leads in shading units by a wide margin, while AMD leads in texture units and ROPs. The AMD part also has 12 ray tracing cores, while the Intel part has 10. The throughput figures reflect these configurations. The AMD Radeon 680M delivers 70.40 GPixel/s of pixel rate and 105.6 GTexel/s of texture rate, with 3.379 TFLOPS of FP32 compute and 6.758 TFLOPS of FP16 compute at a 2:1 ratio. The Intel Arc B370 delivers 48.00 GPixel/s of pixel rate and 96.00 GTexel/s of texture rate, with 6.144 TFLOPS of FP32 compute and 12.29 TFLOPS of FP16 compute at a 2:1 ratio. The Intel part produces roughly 82% more FP32 throughput and roughly 82% more FP16 throughput, despite having fewer TMUs and ROPs. The higher shader count drives that advantage.

Power draw is another major differentiator. The AMD Radeon 680M has a TDP of 50 W. The Intel Arc B370 has a TDP of 25 W. The Intel part delivers a higher Steel Nomad score while drawing half the thermal budget. This is a notable efficiency result. Both parts are IGPs with no power connectors, and both have portable-device-dependent display outputs. The AMD part uses a PCIe 4.0 x8 bus interface, while the Intel part uses an IGP bus interface. Memory configurations are identical in kind: both use system shared memory with system-dependent bandwidth, so neither has dedicated VRAM.

API support is identical. Both parts support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The production status for both is Active. The AMD Radeon 680M was released on 2023-01-02, while the Intel Arc B370 has a release date of 2026-01-26, indicating a substantially newer product. The AMD part lists a predecessor (Vega II IGP) and a successor (Navi III IGP), while the Intel part has neither recorded.

The Verdict

The benchmark data points in one direction for the shared workload. The Intel Arc B370 achieves 1184 points in 3DMark Steel Nomad DX12, compared to 378 points for the AMD Radeon 680M. That is a 68.1% deficit for the AMD part, and it is the only direct head-to-head measurement available. For any user prioritizing that specific 3DMark test, the Intel part is the only choice.

The broader database picture complicates the decision. The AMD Radeon 680M has a 57th percentile rank among all GPUs, with an average benchmark score of 15270. It sits within 1% of the NVIDIA GeForce GTX 580, the NVIDIA GeForce RTX 2060, the NVIDIA GeForce RTX 3050 OEM, and the AMD Radeon RX 7600. Those are formidable neighbors for an integrated GPU. The Intel Arc B370 has a 5th percentile rank and an average score of 1184, placing it alongside the ATI Mobility Radeon HD 5570, the ATI Radeon HD 5770, the AMD Radeon HD 7650M, and the AMD FirePro M2000. Those are all legacy parts.

The discrepancy between the Steel Nomad result and the aggregate percentile likely stems from the test coverage. The AMD part has three recorded benchmarks, including two Geekbench compute tests. The Intel part has only one. The 680M's strong compute scores lift its average and its percentile, while the Intel part's single result does not benefit from any additional data. The Intel Arc B370 wins the one test both share, but the AMD Radeon 680M has a more complete profile in the database.

The efficiency data favors Intel. At 25 W versus 50 W TDP, the Intel part uses half the power budget of the AMD part while scoring materially higher in the shared DX12 test. The newer 3 nm process and the higher shader count support this outcome. The AMD part counters with a higher pixel rate, a higher texture rate, more ROPs, more TMUs, and more ray tracing cores. The AMD part also has a much higher base clock (2000 MHz versus 300 MHz), which may benefit sustained workloads that do not rely on boost behavior.

The release dates indicate the Intel part is three years newer. The AMD part launched on 2023-01-02, and the Intel part on 2026-01-26. For a database that tracks hardware across generations, the Intel Arc B370 represents the more recent design point, and its single benchmark win aligns with that timeline. The AMD Radeon 680M, despite being older and less efficient in the one shared test, maintains a far higher aggregate standing due to its compute results and its proximity to modern discrete GPUs.

The verdict from the recorded data is conditional. The Intel Arc B370 is the stronger part in the only directly comparable workload, and it does so at half the TDP. The AMD Radeon 680M is the stronger part in overall database ranking, driven by its Geekbench OpenCL and Vulkan scores. Neither part covers the other's strengths completely.

FAQ

Q: Which GPU wins the only direct benchmark comparison?

A: The Intel Arc B370 wins 3DMark Steel Nomad DX12 with a score of 1184, while the AMD Radeon 680M scores 378. The delta is -68.1% for the AMD part.

Q: How do the two GPUs rank among all GPUs in the database?

A: The AMD Radeon 680M is at the 57th percentile, while the Intel Arc B370 is at the 5th percentile.

Q: What are the closest rivals for each GPU?

A: The AMD Radeon 680M's nearest rivals are the NVIDIA GeForce GTX 580 (15283), the NVIDIA GeForce RTX 2060 (15290), the NVIDIA GeForce RTX 3050 OEM (15199), and the AMD Radeon RX 7600 (15171). The Intel Arc B370's nearest rivals are the ATI Mobility Radeon HD 5570 (1186), the ATI Radeon HD 5770 (1190), the AMD Radeon HD 7650M (1192), and the AMD FirePro M2000 (1168).

Q: What is the TDP of each GPU?

A: The AMD Radeon 680M has a TDP of 50 W, and the Intel Arc B370 has a TDP of 25 W.

Q: What are the FP32 compute figures for each GPU?

A: The AMD Radeon 680M delivers 3.379 TFLOPS of FP32, and the Intel Arc B370 delivers 6.144 TFLOPS of FP32.

Q: What process nodes and foundries are used?

A: The AMD Radeon 680M uses a 6 nm process at TSMC, and the Intel Arc B370 uses a 3 nm process at Intel.

Where Each One Wins

The Intel Arc B370 wins in the single shared benchmark. The 3DMark Steel Nomad DX12 result of 1184 versus 378 gives Intel a clear victory in that specific DirectX 12 workload. The Intel part also wins on raw compute throughput, with 6.144 TFLOPS FP32 and 12.29 TFLOPS FP16, compared to 3.379 TFLOPS FP32 and 6.758 TFLOPS FP16 for the AMD part. The Intel part wins on efficiency as well, delivering a higher Steel Nomad score at a TDP of 25 W, half the 50 W TDP of the AMD part. The Intel part also has a higher boost clock at 2400 MHz versus 2200 MHz, and it uses a newer 3 nm process at Intel's foundry. The Intel part has more shading units (1280 versus 768), which explains its compute lead.

The AMD Radeon 680M wins in the aggregate. Its average benchmark score of 15270 is based on three tests, and its 57th percentile rank places it among modern discrete GPUs. Its nearest rivals include the NVIDIA GeForce RTX 2060 and the AMD Radeon RX 7600, both of which are far more powerful than the legacy parts surrounding the Intel Arc B370. The AMD part wins on pixel rate (70.40 GPixel/s versus 48.00 GPixel/s) and texture rate (105.6 GTexel/s versus 96.00 GTexel/s). It has more TMUs (48 versus 40), more ROPs (32 versus 20), and more ray tracing cores (12 versus 10). The AMD part has a much higher base clock (2000 MHz versus 300 MHz), which may help in sustained workloads. The AMD part also has a known transistor count of 13,100 million on a 208 mm² die, while the Intel part's transistor count and die size are unknown in the database.

The use-case split follows these strengths. For a modern DirectX 12 benchmark like Steel Nomad, the Intel Arc B370 is the stronger performer, and it achieves that at lower power. For general compute workloads, the AMD Radeon 680M's Geekbench OpenCL score of 23468 and Vulkan score of 21965 indicate substantial compute capability, though the Intel part has no recorded Geekbench results for comparison. For integrated graphics in a power-constrained portable device, the Intel part's 25 W TDP is a distinct advantage. For a system where the GPU must match the aggregate performance profile of discrete cards like the GTX 580 or RTX 2060, the AMD part's database standing is far stronger.

The database also records production status as Active for both, and both support the same API set: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Neither has a launch MSRP recorded. The AMD Radeon 680M has a PCIe 4.0 x8 bus interface, while the Intel Arc B370 uses an IGP bus interface. Memory is system shared for both, with bandwidth dependent on the host system. Display outputs are portable-device dependent for both. The AMD part has a predecessor (Vega II IGP) and a successor (Navi III IGP), while the Intel part has neither recorded. The release dates differ by roughly three years, with the AMD part at 2023-01-02 and the Intel part at 2026-01-26.

The final split is straightforward. The Intel Arc B370 wins the one measured graphics test, the compute throughput comparison, and the power efficiency comparison. The AMD Radeon 680M wins the aggregate database ranking, the pixel and texture throughput comparison, the ray tracing core count, and the overall benchmark profile. The choice depends on whether the single Steel Nomad result or the broader database standing carries more weight.