NVIDIA Jetson Orin Nano Super vs NVIDIA Switch 2 GPU Comparison

Head-to-Head Benchmarks

The recorded database contains no direct head-to-head benchmark entries for the NVIDIA Jetson Orin Nano Super and the NVIDIA Switch 2 GPU. The benchmark arrays are empty, and the win counters for both parts show zero. This means there are no measured frame rates, synthetic scores, or application-specific performance numbers to compare directly. What the data does provide is a full set of compute specifications, memory characteristics, and architectural details that allow for a derived comparison based on theoretical throughput and hardware configuration.

The most significant difference in raw compute appears in the FP32 (single precision) throughput figures. The Jetson Orin Nano Super delivers 2.089 TFLOPS, while the Switch 2 GPU delivers 4.301 TFLOPS. That is more than double the FP32 output for the Switch 2 part. In FP16 (half precision) with the 2:1 ratio, the Jetson reaches 4.178 TFLOPS, and the Switch 2 reaches 8.602 TFLOPS, again a direct doubling. The Switch 2 GPU also has 1536 shading units against 1024 on the Jetson, and 48 texture mapping units versus 32. The texture rate reflects this: the Jetson produces 32.64 GTexel/s, while the Switch 2 produces 67.20 GTexel/s, which is just over double. Pixel rate is also higher on the Switch 2, at 22.40 GPixel/s versus 16.32 GPixel/s on the Jetson, though the difference here is smaller, about 37% ahead.

Clock speeds explain part of this gap. The Jetson Orin Nano Super has no listed base or boost clock in the database, but its memory clock is 800 MHz with 6.4 Gbps effective. The Switch 2 GPU lists a base clock of 561 MHz and a boost clock of 1400 MHz. That boost clock is substantially higher than the base, and when combined with the larger shading unit count, it drives the FP32 and texture rate advantages. The memory configuration differs as well: the Jetson uses 8 GB of LPDDR5, while the Switch 2 uses 12 GB of LPDDR5X, both on a 128-bit bus. Despite the different memory types, the recorded bandwidth is identical at 102.4 GB/s for both parts. The memory clock is the same 800 MHz, so the bandwidth parity is consistent with the bus width being equal.

Both GPUs share the same chip name, GA10B, the same architecture (Ampere), the same process node (8 nm), the same foundry (Samsung), and the same die size (200 mm²). The transistor count is listed as unknown for both. The API support is identical: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Both parts also share the same percentile ranking at 50% of all GPUs in the database, and both have an average benchmark score of zero, which reflects the absence of recorded benchmark data rather than a performance assessment.

The Jetson Orin Nano Super has 32 tensor cores, while the Switch 2 GPU has 48 tensor cores. The Switch 2 also lists 12 ray tracing cores, while the Jetson lists none. This is a notable structural difference in hardware features. The Jetson has 16 ROPs, same as the Switch 2, so the pixel rate difference comes only from the clock speed advantage of the Switch 2. The Jetson has a TDP of 25 W, while the Switch 2 has a TDP of 40 W, a 15 W difference that reflects the higher clocked and more heavily equipped Switch 2 silicon.

The Verdict

From the recorded data, the NVIDIA Switch 2 GPU is the stronger compute part in every measured specification category where a difference exists. It has more shading units, more texture mapping units, more tensor cores, ray tracing cores, higher clock speeds, more memory capacity, and higher FP32, FP16, pixel rate, and texture rate figures. The only areas where the two parts are equal are the memory bus width, memory bandwidth, ROP count, die size, process node, architecture, and API support. The Jetson Orin Nano Super is a lower power, lower clocked, and less equipped version of the same underlying chip.

Given that there are no benchmark scores, the verdict must rest on the specification data. The Switch 2 GPU is positioned for console-class rendering workloads, with a 40 W TDP and 12 GB of memory, while the Jetson Orin Nano Super is positioned for embedded or edge applications, with a 25 W TDP and 8 GB of memory. For any task that relies on raw shader throughput or texture fill, the Switch 2 GPU should be selected based on the numbers. For tasks where lower power consumption is a priority and the smaller memory footprint is acceptable, the Jetson Orin Nano Super is the more conservative choice.

The data indicates that the Switch 2 GPU is roughly twice as fast in FP32 and FP16 throughput, and just over twice as fast in texture rate. The pixel rate advantage is smaller but still present. The Jetson Orin Nano Super does not win any category in the specification comparison. The recorded wins counters are zero for both parts, so there is no benchmark-based victory to assign. Based on the hardware configuration alone, the Switch 2 GPU is the performance leader.

FAQ

Q: Which GPU has higher FP32 performance?

A: The NVIDIA Switch 2 GPU delivers 4.301 TFLOPS in FP32, while the NVIDIA Jetson Orin Nano Super delivers 2.089 TFLOPS. The Switch 2 is more than double the Jetson in this metric.

Q: Do the two GPUs use the same chip?

A: Yes, both the Jetson Orin Nano Super and the Switch 2 GPU are listed with the chip name GA10B, built on the Ampere architecture, using the 8 nm process from Samsung, with a die size of 200 mm².

Q: What is the memory bandwidth of each GPU?

A: Both GPUs have identical memory bandwidth at 102.4 GB/s, achieved over a 128-bit bus. The Jetson uses 8 GB of LPDDR5, while the Switch 2 uses 12 GB of LPDDR5X.

Q: Does the Switch 2 GPU have ray tracing cores?

A: Yes, the Switch 2 GPU lists 12 ray tracing cores. The Jetson Orin Nano Super does not list any ray tracing cores.

Q: What is the difference in texture fill rate?

A: The Jetson Orin Nano Super produces 32.64 GTexel/s, while the Switch 2 GPU produces 67.20 GTexel/s, which is just over double the Jetson's rate.

Q: Which GPU has a higher power limit?

A: The Switch 2 GPU has a TDP of 40 W, while the Jetson Orin Nano Super has a TDP of 25 W.

Q: Are the API features the same?

A: Yes, both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 exactly as listed in the database.

Specification Differences

The two GPUs differ in several recorded specification fields. The Jetson Orin Nano Super has 1024 shading units, while the Switch 2 GPU has 1536 shading units. The texture mapping unit count is 32 on the Jetson versus 48 on the Switch 2. Both have 16 ROPs. The tensor core count is 32 on the Jetson and 48 on the Switch 2. The Switch 2 adds 12 ray tracing cores, while the Jetson lists none.

The clock configuration shows the Jetson has no base or boost clock recorded, while the Switch 2 has a base clock of 561 MHz and a boost clock of 1400 MHz. The memory clock is identical at 800 MHz with 6.4 Gbps effective for both. Memory size differs: 8 GB on the Jetson versus 12 GB on the Switch 2. Memory type differs as well, with LPDDR5 on the Jetson and LPDDR5X on the Switch 2, though the bus width remains 128-bit for both and bandwidth remains 102.4 GB/s for both.

The pixel rate is 16.32 GPixel/s on the Jetson and 22.40 GPixel/s on the Switch 2. The texture rate is 32.64 GTexel/s on the Jetson and 67.20 GTexel/s on the Switch 2. FP32 throughput is 2.089 TFLOPS versus 4.301 TFLOPS, and FP16 throughput is 4.178 TFLOPS versus 8.602 TFLOPS, both with the 2:1 ratio. The TDP is 25 W for the Jetson and 40 W for the Switch 2.

The physical dimensions differ. The Jetson is listed with a length of 70 mm (2.8 inches) and a height of 45 mm (1.8 inches), with no width recorded. The Switch 2 GPU is listed with a length of 272 mm (10.7 inches), a height of 116 mm (4.6 inches), and a width of 14 mm (0.6 inches). The Jetson uses a PCIe 4.0 x4 bus interface, while the Switch 2 has no bus interface recorded. The Jetson has display outputs listed as portable device dependent, while the Switch 2 has no outputs listed. The Jetson has a slot width of IGP, while the Switch 2 has no slot width recorded. Release dates differ: the Jetson was released on 2024-12-16, and the Switch 2 GPU on 2025-06-04.

Architecture Differences

Both GPUs share the same underlying architecture details. The chip is GA10B for both, built on the Ampere architecture, using an 8 nm process from Samsung, with a die size of 200 mm². The transistor count is unknown for both parts. The generation differs in label: the Jetson is listed as Tegra (Ampere), while the Switch 2 GPU is listed as Console GPU (Nintendo). The architecture itself, however, is identical.

The major architectural differences come from the enabled hardware blocks. The Switch 2 GPU has 1536 shading units, 48 TMUs, 48 tensor cores, and 12 ray tracing cores. The Jetson Orin Nano Super has 1024 shading units, 32 TMUs, 32 tensor cores, and no ray tracing cores. Both have 16 ROPs. This means the Switch 2 GPU has a larger execution resource pool across all compute and texture units, plus the addition of ray tracing hardware that the Jetson does not have.

The memory subsystem differs in type and capacity. The Jetson uses LPDDR5, while the Switch 2 uses LPDDR5X, which is a faster memory standard, though the recorded bandwidth is the same at 102.4 GB/s due to the identical 128-bit bus and 800 MHz memory clock. The Switch 2 has 12 GB of memory versus 8 GB on the Jetson, so the larger capacity is the primary memory advantage, not bandwidth.

The clock strategy differs. The Switch 2 GPU has a base clock of 561 MHz and a boost clock of 1400 MHz, allowing for dynamic scaling. The Jetson has no base or boost clock recorded, suggesting a fixed or unspecified clock behavior in the database. The higher boost clock on the Switch 2, combined with the larger shading unit count, yields the substantial throughput differences in FP32, FP16, pixel rate, and texture rate.

The power envelope differs, with the Jetson at 25 W and the Switch 2 at 40 W. The API support is identical between the two, with DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The production status for both is Active.

Where Each One Wins

Based strictly on the recorded data, the NVIDIA Switch 2 GPU wins in every performance-related category. It has higher FP32 throughput at 4.301 TFLOPS versus 2.089 TFLOPS, higher FP16 throughput at 8.602 TFLOPS versus 4.178 TFLOPS, higher texture rate at 67.20 GTexel/s versus 32.64 GTexel/s, and higher pixel rate at 22.40 GPixel/s versus 16.32 GPixel/s. It also has more shading units, more TMUs, more tensor cores, ray tracing cores, and more memory capacity. The Switch 2 GPU is the clear choice for workloads that demand raw graphics compute, shader-heavy rendering, or ray-traced effects, based on the hardware capabilities listed.

The Jetson Orin Nano Super does not win any performance category in the specification data. Its advantages are limited to lower power consumption at 25 W versus 40 W, a smaller physical footprint at 70 mm by 45 mm versus 272 mm by 116 mm by 14 mm, and a PCIe 4.0 x4 interface that the Switch 2 does not list. The Jetson also has display outputs listed as portable device dependent, while the Switch 2 has no outputs. For embedded systems, portable devices, or edge computing applications where the 25 W TDP and compact size are critical, the Jetson Orin Nano Super is the more appropriate part. It also has the same memory bandwidth of 102.4 GB/s, so memory throughput is not a differentiator.

For gaming or console-class rendering, the Switch 2 GPU is the stronger option, given its higher clock speeds, larger execution units, ray tracing cores, and 12 GB of memory. For low-power embedded inference or lightweight graphics tasks, the Jetson Orin Nano Super fits within a 25 W envelope and a much smaller board footprint. The data does not include any benchmark results, so these conclusions are drawn entirely from the specification differences recorded in the database. Both parts share the same architecture, process node, and die size, so the differences come down to enabled hardware, clocks, memory capacity, and power limits.