Intel Arc Graphics 64EU vs NVIDIA Jetson Orin Nano Super Comparison

Head-to-Head Benchmarks

The recorded data shows no direct head-to-head benchmark comparison between the Intel Arc Graphics 64EU and the NVIDIA Jetson Orin Nano Super. The database contains a single 3DMark Steel Nomad DX12 result for the Intel part, while the NVIDIA part has no benchmark entries at all. This makes a direct score-for-score comparison impossible from the available measurements.

What can be established is the Intel Arc Graphics 64EU's standing in the broader database. Its 3DMark Steel Nomad DX12 score of 733 places it in the 3rd percentile of all GPUs, a very low position. The nearest recorded rivals confirm this: the Intel Arc Graphics 32EU and Intel Arc Graphics 24EU both average exactly 733, a 0% delta. The AMD Radeon HD 6470M is 1.4% behind at 723, while the NVIDIA GeForce GT 415M is 2.4% ahead at 751. These margins are small, indicating that the Arc 64EU operates in a performance class that is essentially interchangeable with those older or lower-end parts.

The NVIDIA Jetson Orin Nano Super has an average benchmark score of 0 in the database, with an empty benchmark list. Its percentile rank is 50, but this figure cannot be tied to any measured test result in the available data. The absence of recorded scores means any performance comparison would rely on theoretical specifications rather than measured outcomes.

The FP32 compute figures offer a partial proxy. The Intel Arc Graphics 64EU delivers 1.946 TFLOPS, while the NVIDIA Jetson Orin Nano Super delivers 2.089 TFLOPS. The NVIDIA part holds a 7.3% advantage in raw FP32 throughput. In FP16, the Intel part reaches 3.891 TFLOPS (2:1) and the NVIDIA part reaches 4.178 TFLOPS (2:1), a similar proportional gap. Pixel rate and texture rate move in the opposite direction: the Intel part records 30.40 GPixel/s and 60.80 GTexel/s, while the NVIDIA part records 16.32 GPixel/s and 32.64 GTexel/s. The Intel part is 86.3% ahead in pixel rate and 86.3% ahead in texture rate. These are theoretical rates, however, not benchmark scores.

Given the lack of a head-to-head test, the data cannot confirm which part wins any specific benchmark. The Intel part has a measured result, and the NVIDIA part does not. That is the extent of what the database supports.

FAQ

Q: What is the Intel Arc Graphics 64EU's measured benchmark score?

A: The database records a 3DMark Steel Nomad DX12 score of 733, which places it in the 3rd percentile of all GPUs.

Q: How does the Intel Arc Graphics 64EU compare to its nearest rivals?

A: It is exactly level with the Intel Arc Graphics 32EU and Intel Arc Graphics 24EU (both 733, 0% delta). It is 1.4% ahead of the AMD Radeon HD 6470M (723) and 2.4% behind the NVIDIA GeForce GT 415M (751).

Q: Does the NVIDIA Jetson Orin Nano Super have any benchmark results in the database?

A: No. Its benchmark list is empty, and its average benchmark score is recorded as 0, though its percentile rank is listed as 50.

Q: Which part has higher FP32 compute?

A: The NVIDIA Jetson Orin Nano Super records 2.089 TFLOPS, which is 7.3% higher than the Intel Arc Graphics 64EU's 1.946 TFLOPS.

Q: Which part has higher pixel and texture rates?

A: The Intel Arc Graphics 64EU records 30.40 GPixel/s and 60.80 GTexel/s, versus 16.32 GPixel/s and 32.64 GTexel/s for the NVIDIA Jetson Orin Nano Super, an 86.3% advantage in each.

Q: What memory does the NVIDIA Jetson Orin Nano Super use?

A: It has 8 GB of LPDDR5 memory on a 128-bit bus, with 102.4 GB/s of bandwidth and a memory clock of 800 MHz (6.4 Gbps effective). The Intel Arc Graphics 64EU uses system-shared memory with bandwidth described as system dependent.

Where Each One Wins

The Intel Arc Graphics 64EU wins in rasterization throughput metrics. Its pixel rate of 30.40 GPixel/s and texture rate of 60.80 GTexel/s are both 86.3% higher than the NVIDIA part's 16.32 GPixel/s and 32.64 GTexel/s. This suggests an advantage in fill-rate-bound workloads, where the output of shaded pixels and filtered textures is the limiting factor. The Intel part also has a measured benchmark result, 733 in 3DMark Steel Nomad DX12, which at least places it in a known performance tier.

The NVIDIA Jetson Orin Nano Super wins in compute throughput. Its FP32 figure of 2.089 TFLOPS is 7.3% above the Intel part, and its FP16 figure of 4.178 TFLOPS is 7.4% above the Intel part's 3.891 TFLOPS. It also brings 32 tensor cores, which the Intel part lacks entirely, indicating a capability for tensor-accelerated workloads that the Intel part cannot match on paper. The NVIDIA part's 8 GB of dedicated LPDDR5 memory with 102.4 GB/s of bandwidth is a structural advantage over system-shared memory, as the Intel part's bandwidth is system dependent and cannot be quantified from the data.

The NVIDIA part also holds advantages in power efficiency and physical form. Its TDP is 25 W versus 65 W for the Intel part, a 61.5% lower power envelope. Its dimensions are recorded at 70 mm by 45 mm, while the Intel part has no recorded dimensions and is an IGP integrated into a processor. The NVIDIA part is a standalone portable device module, whereas the Intel part depends on the motherboard for display output and memory.

Specification Differences

The two parts differ in nearly every recorded specification category. The Intel Arc Graphics 64EU uses a 3 nm process from TSMC, while the NVIDIA Jetson Orin Nano Super uses an 8 nm process from Samsung. The Intel part's chip is Arrow Lake-S with 17,800 million transistors on a 243 mm² die, giving a transistor density of 73.3M per mm². The NVIDIA part's chip is GA10B on a 200 mm² die, with transistor count recorded as unknown and density not recorded.

Clock behavior differs markedly. The Intel part has a base clock of 300 MHz and a boost clock of 1900 MHz. The NVIDIA part has no base or boost clock recorded; only its memory clock of 800 MHz (6.4 Gbps effective) is listed.

Memory setup is a major divergence. The Intel part uses system-shared memory with system-shared type, bus width, and system-dependent bandwidth. The NVIDIA part has 8 GB of dedicated LPDDR5 on a 128-bit bus with 102.4 GB/s bandwidth.

Compute unit counts differ. The Intel part has 512 shading units, 32 TMUs, and 16 ROPs. The NVIDIA part has 1024 shading units, 32 TMUs, and 16 ROPs, plus 32 tensor cores. Neither part lists RT cores.

Power and interface differ. The Intel part has a 65 W TDP and a Ring Bus interface. The NVIDIA part has a 25 W TDP and a PCIe 4.0 x4 interface. The NVIDIA part has recorded dimensions of 70 mm by 45 mm; the Intel part has none. Display outputs are motherboard dependent for Intel and portable device dependent for NVIDIA. The NVIDIA part has a launch MSRP of 249 USD; the Intel part has no recorded launch MSRP.

Architecture Differences

The Intel Arc Graphics 64EU uses the Xe-LPG architecture from the Arc Graphics (Arrow Lake) generation, built on a 3 nm TSMC process. The NVIDIA Jetson Orin Nano Super uses the Ampere architecture from the Tegra (Ampere) generation, built on an 8 nm Samsung process. These are different architectural families with different design priorities.

The shading unit count is the clearest architectural signal. The NVIDIA part has 1024 shading units, double the Intel part's 512. Yet the NVIDIA part's FP32 throughput is only 7.3% higher, indicating a lower per-shader clock or a different scheduling approach. The Intel part compensates with a higher boost clock of 1900 MHz versus no recorded clock for the NVIDIA part.

The Intel part's Xe-LPG architecture includes no tensor cores, while the NVIDIA Ampere part includes 32. This is a definitive feature difference: the NVIDIA part can execute tensor operations, and the Intel part cannot according to the recorded data.

Both parts support the same API set: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Intel part is an integrated GPU on a Ring Bus, while the NVIDIA part is a portable device module on PCIe 4.0 x4. The Intel part's predecessor is listed as HD Graphics; the NVIDIA part has no predecessor recorded.

The production status of both is Active. The Intel part was released on 2024-10-23 and the NVIDIA part on 2024-12-16.

The Verdict

The data does not support a definitive performance winner because the NVIDIA Jetson Orin Nano Super has no recorded benchmark scores. The Intel Arc Graphics 64EU has a measured 3DMark Steel Nomad DX12 score of 733, which is a very low result at the 3rd percentile. That score ties the Intel Arc Graphics 32EU and 24EU, edges the AMD Radeon HD 6470M by 1.4%, and trails the NVIDIA GeForce GT 415M by 2.4%. This places the Intel part in a performance class far below the mainstream.

For users who prioritize rasterization throughput as expressed in pixel and texture rates, the Intel part is the clear choice from the recorded data. Its 30.40 GPixel/s and 60.80 GTexel/s are 86.3% higher than the NVIDIA part's figures. For users who prioritize compute throughput, the NVIDIA part leads in FP32 and FP16 by about 7.3% and 7.4% respectively, and it is the only one of the two with tensor cores. The NVIDIA part also has dedicated 8 GB LPDDR5 memory with 102.4 GB/s bandwidth, a fixed resource that does not depend on system configuration, whereas the Intel part's memory bandwidth is system dependent.

Power characteristics favor the NVIDIA part decisively: 25 W versus 65 W, a 61.5% reduction. The NVIDIA part's physical footprint of 70 mm by 45 mm makes it a compact, self-contained module, while the Intel part is an IGP with no recorded dimensions. The NVIDIA part also carries a launch MSRP of 249 USD.

The choice depends on the workload. The Intel Arc Graphics 64EU suits fill-rate-bound graphics tasks and has a known, if low, benchmark score. The NVIDIA Jetson Orin Nano Super suits compute-oriented and tensor-accelerated tasks, offers dedicated memory, and consumes far less power, but its performance is unmeasured in the database. Without benchmark results for the NVIDIA part, the only fully verified performance claim belongs to the Intel part, and that claim is a modest one.