Intel Arc Pro B390 vs NVIDIA Jetson Orin Nano Super Comparison

FAQ

Q: What are the core architectural identities of the Intel Arc Pro B390 and the NVIDIA Jetson Orin Nano Super?

A: The Intel Arc Pro B390 is built on Intel’s 3 nm process using the Xe3-LPG architecture, part of the Arc Graphics-WM (Panther Lake) generation. The NVIDIA Jetson Orin Nano Super uses Samsung’s 8 nm process with the Ampere architecture, specifically the GA10B chip in the Tegra (Ampere) generation.

Q: How do their memory configurations differ?

A: The Intel Arc Pro B390 uses system-shared memory, with bandwidth described as system-dependent and a memory clock of system shared. The NVIDIA Jetson Orin Nano Super has 8 GB of LPDDR5 memory on a 128-bit bus, delivering 102.4 GB/s of bandwidth.

Q: Which device has higher raw shading and texture throughput?

A: The Intel Arc Pro B390 has 1536 shading units, 48 texture mapping units, and 24 raster operation units. The NVIDIA Jetson Orin Nano Super has 1024 shading units, 32 TMUs, and 16 ROPs. The Arc Pro B390 also outputs 60.00 GPixel/s and 120.0 GTexel/s, versus 16.32 GPixel/s and 32.64 GTexel/s for the Jetson.

Q: What is the difference in FP32 compute performance?

A: The Intel Arc Pro B390 delivers 7.680 TFLOPS of FP32 performance. The NVIDIA Jetson Orin Nano Super delivers 2.089 TFLOPS. This makes the Intel part roughly 3.7 times higher in FP32 throughput, based on the recorded data.

Q: Do both devices support the same graphics APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. However, the Intel Arc Pro B390 also includes 12 dedicated ray tracing cores, while the Jetson Orin Nano Super does not list any RT cores but includes 32 tensor cores.

Q: What are the power and physical form factor differences?

A: The Intel Arc Pro B390 has a TDP of 80 W and is an integrated graphics processor (IGP) with no power connectors. The NVIDIA Jetson Orin Nano Super has a TDP of 25 W, is also an IGP, but uses a PCIe 4.0 x4 bus interface. The Jetson has physical dimensions of 70 mm by 45 mm.

Where Each One Wins

The Intel Arc Pro B390 is the clear winner in raw graphics throughput. Its pixel rate of 60.00 GPixel/s is 3.7 times that of the Jetson Orin Nano Super’s 16.32 GPixel/s. Texture rate follows the same pattern: 120.0 GTexel/s versus 32.64 GTexel/s, a 3.7 times advantage. FP32 compute is 7.680 TFLOPS versus 2.089 TFLOPS, and FP16 throughput is 15.36 TFLOPS versus 4.178 TFLOPS. The Arc Pro B390 also brings 12 ray tracing cores, which the Jetson lacks entirely. For any workload that stresses traditional graphics rendering, rasterization, or ray tracing, the Intel part dominates.

The NVIDIA Jetson Orin Nano Super wins in memory capacity and integration. It has 8 GB of dedicated LPDDR5 memory with a fixed 102.4 GB/s bandwidth, while the Arc Pro B390 depends entirely on system-shared memory with system-dependent bandwidth. The Jetson’s 32 tensor cores provide a hardware path for AI and inference workloads that the Intel part does not list. The Jetson also operates at 25 W, one-third the TDP of the Arc Pro B390, and has a compact 70 mm by 45 mm footprint, making it suitable for embedded or edge deployments where power and space are constrained. Its release date of December 2024 also predates the Arc Pro B390’s January 2026 launch, giving it a longer deployment history in the field.

Architecture Differences

The two devices come from different architectural lineages. The Intel Arc Pro B390 uses Xe3-LPG, the third-generation Xe architecture, implemented on Intel’s 3 nm process. The chip is part of the Arc Graphics-WM (Panther Lake) generation, and it inherits the integrated graphics design philosophy of its predecessor, HD Graphics-WM. The NVIDIA Jetson Orin Nano Super uses Ampere, NVIDIA’s GPU architecture from the Tegra (Ampere) generation, fabricated by Samsung on an 8 nm process with a die size of 200 mm².

The compute layouts diverge significantly. The Arc Pro B390 has 1536 shading units, 48 TMUs, and 24 ROPs, plus 12 dedicated ray tracing cores. The Jetson Orin Nano Super has 1024 shading units, 32 TMUs, and 16 ROPs, with no ray tracing cores but 32 tensor cores. This reflects different priorities: Intel’s part is built for graphics-heavy workloads, while NVIDIA’s part is oriented toward parallel compute and AI acceleration.

Process node differences also matter. The 3 nm Intel process versus the 8 nm Samsung process gives the Arc Pro B390 a transistor density advantage, although the exact transistor counts are unknown for both. The Arc Pro B390’s higher clock range, from a base of 300 MHz to a boost of 2500 MHz, contrasts with the Jetson, which lists no base or boost clocks but runs its memory at 800 MHz with 6.4 Gbps effective speed. The memory architecture is fundamentally different: system-shared for Intel, dedicated LPDDR5 for NVIDIA.

Specification Differences

| Specification | Intel Arc Pro B390 | NVIDIA Jetson Orin Nano Super |

|---------------|-------------------|-------------------------------|

| Process Node | 3 nm | 8 nm |

| Foundry | Intel | Samsung |

| Die Size | Unknown | 200 mm² |

| Base Clock | 300 MHz | Not listed |

| Boost Clock | 2500 MHz | Not listed |

| Memory Size | System Shared | 8 GB |

| Memory Type | System Shared | LPDDR5 |

| Memory Bus Width | System Shared | 128 bit |

| Memory Bandwidth | System Dependent | 102.4 GB/s |

| Shading Units | 1536 | 1024 |

| TMUs | 48 | 32 |

| ROPs | 24 | 16 |

| RT Cores | 12 | None listed |

| Tensor Cores | None listed | 32 |

| Pixel Rate | 60.00 GPixel/s | 16.32 GPixel/s |

| Texture Rate | 120.0 GTexel/s | 32.64 GTexel/s |

| FP32 | 7.680 TFLOPS | 2.089 TFLOPS |

| FP16 | 15.36 TFLOPS | 4.178 TFLOPS |

| TDP | 80 W | 25 W |

| Bus Interface | IGP | PCIe 4.0 x4 |

| Release Date | 2026-01-26 | 2024-12-16 |

| Launch MSRP | Not listed | 249 USD |

Head-to-Head Benchmarks

The recorded data shows no direct head-to-head benchmark scores between the two devices, but the specification-derived metrics provide a clear comparison. The most decisive difference is in pixel throughput. The Arc Pro B390’s 60.00 GPixel/s is 3.7 times the Jetson’s 16.32 GPixel/s. This translates directly to fill-rate-bound scenarios: higher resolution rendering, more complex fragment shaders, and multi-sample anti-aliasing will all favor the Intel part.

Texture throughput shows the same ratio. The Arc Pro B390 delivers 120.0 GTexel/s versus 32.64 GTexel/s for the Jetson. Games and applications that rely heavily on texture sampling, such as detailed terrain rendering or high-resolution material mapping, will see a substantial advantage on the Intel side. The 48 TMUs versus 32 TMUs, combined with the higher clock speeds, explain this gap.

Compute performance is where the separation becomes stark. The Arc Pro B390’s 7.680 TFLOPS of FP32 is 3.7 times the Jetson’s 2.089 TFLOPS. In FP16, the Intel part reaches 15.36 TFLOPS versus 4.178 TFLOPS. This means general-purpose GPU compute, physics simulations, and shader-heavy effects will run significantly faster on the Arc Pro B390. The 1536 shading units versus 1024 give the Intel part a 1.5 times raw unit count advantage, and the higher boost clock of 2500 MHz amplifies that.

The Jetson Orin Nano Super does not win in any graphics or compute metric listed here. Its advantages are structural. The 8 GB of dedicated LPDDR5 memory with 102.4 GB/s bandwidth is a fixed, predictable resource. The Arc Pro B390’s system-shared memory means its effective bandwidth is system-dependent, so performance will vary based on the host platform’s memory configuration. For embedded applications where memory access patterns are critical and latency must be controlled, the Jetson’s dedicated memory is the more reliable choice.

The 32 tensor cores on the Jetson provide a hardware acceleration path for AI inference, a feature the Arc Pro B390 does not list. Tensor core performance is not quantified in the data, but the presence of dedicated hardware for matrix operations gives the Jetson a functional edge in neural network workloads, even if its raw FP32 throughput is lower.

Power consumption is another clear split. The Jetson operates at 25 W, versus 80 W for the Arc Pro B390. For fanless or battery-powered designs, the Jetson’s lower TDP is a decisive factor. The Arc Pro B390’s higher performance comes at 3.2 times the power draw, based on the TDP figures. The Jetson’s compact dimensions of 70 mm by 45 mm also suit it for space-constrained enclosures.

The release timeline favors the Jetson for current deployments. It launched on 2024-12-16, while the Arc Pro B390 is dated 2026-01-26. Both are listed as Active in production status. The Jetson’s launch MSRP is 249 USD, while the Arc Pro B390 has no listed launch MSRP.

In summary, the data indicates the Intel Arc Pro B390 is the faster graphics processor by a wide margin in every measured throughput category, while the NVIDIA Jetson Orin Nano Super offers lower power, dedicated memory, tensor cores, and a smaller physical footprint. The choice depends on whether the workload prioritizes raw rendering performance or embedded efficiency and AI acceleration.