Intel Arc B370 vs NVIDIA Jetson Orin Nano Super Comparison

Head-to-Head Benchmarks

The recorded database contains one direct benchmark result for the Intel Arc B370: a 3DMark Steel Nomad DX12 score of 1184. This places the part at the 5th percentile among all GPUs in the database, meaning the vast majority of recorded graphics processors outperform it in this specific test. The NVIDIA Jetson Orin Nano Super has no corresponding benchmark entries in the database, so a direct score comparison is not possible from recorded data.

The Arc B370's nearest rivals in the database provide useful context. The ATI Mobility Radeon HD 5570 scores 1186, a delta of -0.2 percent relative to the Arc B370, indicating near parity. The ATI Radeon HD 5770 scores 1190, a delta of -0.5 percent. The AMD Radeon HD 7650M scores 1192, a delta of -0.7 percent. The AMD FirePro M2000 scores 1168, a delta of 1.4 percent, meaning the Arc B370 is 1.4 percent ahead of that part. These are extremely tight margins, with the Arc B370 effectively trading places within a 24-point band across four rival products. The data shows the Arc B370 sits in a performance cluster where a few points separate adjacent entries.

The Jetson Orin Nano Super's computed figures tell a different story. Its FP32 throughput is 2.089 TFLOPS, while the Arc B370 delivers 6.144 TFLOPS. That places the Intel part at roughly 2.94 times the single-precision compute rate. In FP16, the Arc B370 reaches 12.29 TFLOPS compared to the Jetson's 4.178 TFLOPS, a factor of roughly 2.94 as well, since both parts use a 2:1 FP16 ratio. Pixel throughput favors the Arc B370 at 48.00 GPixel/s versus 16.32 GPixel/s, a 2.94x advantage. Texture rate follows the same pattern: 96.00 GTexel/s versus 32.64 GTexel/s, again a 2.94x margin. These ratios are consistent because the shader, TMU, and ROP counts scale in the same proportion between the two chips.

The Jetson Orin Nano Super shows no wins in recorded head-to-head benchmarks because none exist in the database. Its percentile rank of 50 reflects its position among all GPUs, but with an average benchmark score of zero, that percentile is not tied to any measured performance entry. The Arc B370's percentile of 5 is tied to its single Steel Nomad result. The absence of a shared benchmark means the only numerical comparisons available are the theoretical throughput figures listed in each part's specification block.

FAQ

Q: Which part has the higher FP32 throughput?

A: The Intel Arc B370 delivers 6.144 TFLOPS FP32, while the NVIDIA Jetson Orin Nano Super delivers 2.089 TFLOPS FP32. The Arc B370 is roughly 2.94 times faster in this metric.

Q: What is the Jetson Orin Nano Super's launch MSRP?

A: The launch MSRP for the NVIDIA Jetson Orin Nano Super is 249 USD, as recorded in the database.

Q: Does the Jetson Orin Nano Super have any recorded 3DMark results?

A: No. The database lists no benchmark entries for the Jetson Orin Nano Super, and its average benchmark score is zero. The Arc B370 has one recorded result, a 3DMark Steel Nomad DX12 score of 1184.

Q: How does the Arc B370 compare to its nearest database rivals?

A: The Arc B370's Steel Nomad score of 1184 sits within 0.7 percent of the ATI Mobility Radeon HD 5570 (1186), ATI Radeon HD 5770 (1190), and AMD Radeon HD 7650M (1192). It is 1.4 percent ahead of the AMD FirePro M2000 (1168).

Q: Do both parts support the same graphics APIs?

A: Yes. Both the Intel Arc B370 and NVIDIA Jetson Orin Nano Super support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: What are the memory specifications for each part?

A: The Arc B370 uses system shared memory with system dependent bandwidth. The Jetson Orin Nano Super has 8 GB of LPDDR5 on a 128-bit bus with 102.4 GB/s of bandwidth and memory clocked at 800 MHz (6.4 Gbps effective).

Architecture Differences

The two accelerators come from different design lineages. The Intel Arc B370 is built on the Panther Lake chip using the Xe3-LPG architecture, placed in the Arc Graphics-M (Panther Lake) generation. The NVIDIA Jetson Orin Nano Super uses the GA10B chip with the Ampere architecture, belonging to the Tegra (Ampere) generation. These are fundamentally different product categories: the Intel part is an integrated graphics processor for portable devices, while the NVIDIA part is a Tegra-class system-on-module with a PCIe 4.0 x4 bus interface.

Process technology diverges sharply. The Arc B370 is fabricated on a 3 nm node at Intel, while the Jetson Orin Nano Super uses an 8 nm node at Samsung. The die size for the Intel chip is not recorded in the database, but the NVIDIA chip has a 200 mm² die. Transistor counts are unknown for both parts. The process node difference implies a significant density gap, but the database provides no transistor density figures to quantify it.

Compute resource allocation differs in structure. The Arc B370 carries 1280 shading units, 40 texture mapping units, 20 ROPs, and 10 ray tracing cores. The Jetson Orin Nano Super carries 1024 shading units, 32 TMUs, 16 ROPs, and no recorded ray tracing cores. Instead, the NVIDIA part includes 32 tensor cores, a feature absent from the Arc B370's recorded specifications. The Intel part has no tensor core field populated, so the database does not confirm or deny their presence, but the record shows none.

Clock behavior also separates the two. The Arc B370 lists a base clock of 300 MHz and a boost clock of 2400 MHz. The Jetson Orin Nano Super lists no base or boost clock in the database, only a memory clock of 800 MHz with 6.4 Gbps effective data rate. The Intel part's memory clock is listed as system shared, meaning it depends on the host platform. The NVIDIA part's memory operates at a fixed recorded speed.

The Arc B370's display outputs are portable device dependent, and its bus interface is IGP. The Jetson Orin Nano Super also has portable device dependent display outputs but uses PCIe 4.0 x4 as its bus interface. The Arc B370 has no power connectors recorded, with a slot width of IGP. The Jetson Orin Nano Super has no power connector entry at all. Physical dimensions are recorded only for the NVIDIA part: 70 mm by 45 mm. The Intel part has no dimensions in the database.

Specification Differences

The two parts differ across nearly every recorded specification field. The Arc B370 uses a 3 nm Intel process; the Jetson Orin Nano Super uses an 8 nm Samsung process. Foundries are Intel and Samsung respectively. The Jetson's die size is 200 mm²; the Arc B370's die size is not recorded.

Shading units: 1280 on the Arc B370 versus 1024 on the Jetson. TMUs: 40 versus 32. ROPs: 20 versus 16. Ray tracing cores: 10 on the Arc B370, none recorded on the Jetson. Tensor cores: none recorded on the Arc B370, 32 on the Jetson. Pixel rate: 48.00 GPixel/s versus 16.32 GPixel/s. Texture rate: 96.00 GTexel/s versus 32.64 GTexel/s. FP32: 6.144 TFLOPS versus 2.089 TFLOPS. FP16: 12.29 TFLOPS versus 4.178 TFLOPS, both using a 2:1 ratio.

Memory configurations are fundamentally different. The Arc B370 uses system shared memory with system shared type, bus width, and system dependent bandwidth. The Jetson has 8 GB of LPDDR5 with a 128-bit bus and 102.4 GB/s bandwidth. The Jetson's memory clocks at 800 MHz with 6.4 Gbps effective; the Arc B370's memory clock is system shared. Both parts have a TDP of 25 W. Both use IGP slot width. The Arc B370 lists no power connectors; the Jetson has no power connector field populated. The Arc B370's bus interface is IGP; the Jetson's is PCIe 4.0 x4. Both list portable device dependent display outputs.

API support matches exactly: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 on both parts. Release dates differ: the Arc B370 is dated 2026-01-26, while the Jetson Orin Nano Super is dated 2024-12-16. Both have active production status. The Jetson has a launch MSRP of 249 USD; the Arc B370 has no recorded launch MSRP. The Arc B370's base clock is 300 MHz and boost is 2400 MHz; the Jetson has no base or boost clock recorded. The Jetson's physical footprint is 70 mm by 45 mm; the Arc B370 has no recorded dimensions.

Where Each One Wins

The Intel Arc B370 wins on raw compute metrics across the board. Its FP32 throughput of 6.144 TFLOPS is nearly three times the Jetson's 2.089 TFLOPS. Its 48.00 GPixel/s pixel rate and 96.00 GTexel/s texture rate dwarf the Jetson's 16.32 GPixel/s and 32.64 GTexel/s. For workloads that depend on shader throughput, rasterization, or texture filtering, the recorded specifications place the Arc B370 clearly ahead. The presence of 10 ray tracing cores on the Arc B370 gives it a feature the Jetson does not list at all, which matters for any DirectX 12 Ultimate ray-traced content. The single recorded 3DMark Steel Nomad score of 1184 confirms the Arc B370 can complete a modern DX12 workload, even if its 5th percentile standing shows it is not a high-end part.

The NVIDIA Jetson Orin Nano Super wins on memory determinism and embedded integration. Its 8 GB of LPDDR5 with 102.4 GB/s of bandwidth is a fixed, known quantity. The Arc B370's system shared memory is entirely dependent on the host platform, so its bandwidth cannot be stated as a single number. The Jetson also provides 32 tensor cores, which the Arc B370 does not list. Any workload that relies on tensor core acceleration, such as AI inference or matrix operations, has a hardware path on the Jetson that the Intel part lacks in the database record. The Jetson's PCIe 4.0 x4 bus interface allows it to operate as a discrete module, whereas the Arc B370 is an IGP tied to its host processor.

The Jetson's 50th percentile standing among all GPUs, combined with a lack of recorded benchmark scores, suggests the database treats it as a mid-pack embedded part. The Arc B370's 5th percentile standing, tied to its Steel Nomad result, indicates it sits near the bottom of the desktop GPU spectrum in that test. The nearest rival data for the Arc B370 shows a cluster of older ATI and AMD parts within a 1.4 percent range, meaning the Intel part is competitive with legacy discrete GPUs from roughly a decade earlier in this workload.

For sustained compute tasks where the host platform provides ample system memory bandwidth, the Arc B370's higher shader counts and clock speeds should deliver better frame rates or throughput. For power-constrained embedded deployments where memory bandwidth is fixed and tensor operations are required, the Jetson Orin Nano Super has the recorded advantages. Both parts share a 25 W TDP and identical API support, so power draw and API compatibility do not differentiate them. The release dates show the Jetson arrived in December 2024, while the Arc B370 is dated January 2026, making the Intel part the newer design in the database.