Intel Arc Pro B390 vs NVIDIA Switch 2 GPU Comparison

Intel Arc Pro B390 is an integrated graphics processor from Intel, built on the Panther Lake chip and Xe3-LPG architecture at a 3 nm process node. NVIDIA Switch 2 GPU is a console graphics processor from NVIDIA, using the GA10B chip and Ampere architecture at an 8 nm process node fabricated by Samsung. Both devices share the same shading unit count of 1536 and texture mapping unit count of 48, but they differ substantially in clock speeds, memory configuration, and power envelope. The database lists both with a 50th percentile ranking among all GPUs, indicating they occupy a similar tier in overall performance distribution, though their measured benchmark scores are not yet recorded.

FAQ

Q: What are the core clock speeds of the Intel Arc Pro B390 and NVIDIA Switch 2 GPU?

A: The Intel Arc Pro B390 has a base clock of 300 MHz and a boost clock of 2500 MHz. The NVIDIA Switch 2 GPU has a base clock of 561 MHz and a boost clock of 1400 MHz.

Q: How much memory does each GPU have access to?

A: The Intel Arc Pro B390 uses system shared memory, with the size, type, bus width, and bandwidth all listed as system dependent. The NVIDIA Switch 2 GPU has 12 GB of LPDDR5X memory on a 128-bit bus, providing 102.4 GB/s of bandwidth.

Q: What is the difference in pixel fill rate between the two?

A: The Intel Arc Pro B390 achieves a pixel rate of 60.00 GPixel/s, while the NVIDIA Switch 2 GPU achieves 22.40 GPixel/s. This gives the Intel part a 2.68 times higher pixel throughput based on the recorded data.

Q: Which GPU has higher FP32 compute throughput?

A: The Intel Arc Pro B390 delivers 7.680 TFLOPS of FP32 compute, compared to 4.301 TFLOPS for the NVIDIA Switch 2 GPU. The Intel part is approximately 1.79 times higher in this metric.

Q: What is the power consumption of each device?

A: The Intel Arc Pro B390 has a TDP of 80 W, while the NVIDIA Switch 2 GPU has a TDP of 40 W. The NVIDIA part consumes half the power according to the listed specifications.

Q: What API levels do both GPUs support?

A: Both the Intel Arc Pro B390 and NVIDIA Switch 2 GPU support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. There is no difference in API support between the two.

Architecture Differences

The Intel Arc Pro B390 is built on the Panther Lake chip using the Xe3-LPG architecture, manufactured at a 3 nm process node by Intel. The NVIDIA Switch 2 GPU uses the GA10B chip with the Ampere architecture, manufactured at an 8 nm process node by Samsung. The Intel part has a die size listed as unknown, while the NVIDIA part has a die size of 200 mm². Both devices have transistor counts listed as unknown, so no direct comparison of transistor density is possible from the recorded data.

The Intel Arc Pro B390 is an integrated graphics processor with an IGP slot width and no power connectors, meaning it draws power from the host system. The NVIDIA Switch 2 GPU is a console GPU with physical dimensions of 272 mm in length, 116 mm in height, and 14 mm in width. The NVIDIA part has no display outputs listed, while the Intel part has display outputs described as portable device dependent, indicating it relies on the host device for display connectivity.

Both GPUs have 1536 shading units and 48 texture mapping units. The Intel Arc Pro B390 has 24 ROPs, while the NVIDIA Switch 2 GPU has 16 ROPs. Both have 12 ray tracing cores, but the NVIDIA part additionally includes 48 tensor cores, while the Intel part has no tensor cores listed. This difference in tensor core presence suggests the NVIDIA Switch 2 GPU has dedicated hardware for AI and tensor workloads, which the Intel Arc Pro B390 lacks.

The memory architecture differs fundamentally. The Intel Arc Pro B390 uses system shared memory, where the size, type, bus width, and bandwidth are all system dependent, meaning they vary based on the host platform. The NVIDIA Switch 2 GPU has a fixed 12 GB of LPDDR5X memory on a 128-bit bus, providing 102.4 GB/s of bandwidth. The NVIDIA part also lists a memory clock of 800 MHz with 6.4 Gbps effective transfer rate.

Clock behavior also differs. The Intel Arc Pro B390 has a base clock of 300 MHz and a boost clock of 2500 MHz, while the NVIDIA Switch 2 GPU has a base clock of 561 MHz and a boost clock of 1400 MHz. The Intel part has a much wider clock range and a significantly higher boost clock, contributing to its higher compute rates.

Power consumption is markedly different. The Intel Arc Pro B390 has a TDP of 80 W, while the NVIDIA Switch 2 GPU has a TDP of 40 W. The NVIDIA part is designed for lower power operation, consistent with its console application, while the Intel part operates at double the power envelope.

The Verdict

The data shows a clear performance advantage for the Intel Arc Pro B390 in raw compute and fill rate metrics. The Intel part delivers 7.680 TFLOPS of FP32 compute compared to 4.301 TFLOPS for the NVIDIA Switch 2 GPU, representing a 1.79 times advantage. The pixel rate of 60.00 GPixel/s versus 22.40 GPixel/s gives the Intel part a 2.68 times advantage in pixel throughput. The texture rate of 120.0 GTexel/s versus 67.20 GTexel/s provides a 1.79 times advantage in texture fill.

The NVIDIA Switch 2 GPU counters with a lower power draw of 40 W versus 80 W, and it includes tensor cores that the Intel part lacks. The NVIDIA part also has a fixed 12 GB memory allocation, whereas the Intel part relies on system shared memory that may vary. Both GPUs occupy the same 50th percentile ranking among all GPUs, indicating they are positioned at a similar performance tier in the overall distribution.

The choice between these two depends on the application context. For systems where integrated graphics are required and higher compute throughput is needed, the Intel Arc Pro B390 provides superior raw performance based on the recorded specifications. For console applications where power efficiency and fixed memory allocation are priorities, the NVIDIA Switch 2 GPU offers a more constrained but self-contained solution. The presence of tensor cores in the NVIDIA part suggests capability for AI-accelerated workloads, while the Intel part relies on its higher raw compute rates.

Specification Differences

The two GPUs differ in nearly every hardware specification recorded in the database.

Process and Chip: The Intel Arc Pro B390 uses the Panther Lake chip with Xe3-LPG architecture at a 3 nm process node from Intel. The NVIDIA Switch 2 GPU uses the GA10B chip with Ampere architecture at an 8 nm process node from Samsung. The NVIDIA part has a die size of 200 mm², while the Intel part has no die size listed.

Clocks: The Intel Arc Pro B390 has a base clock of 300 MHz and a boost clock of 2500 MHz. The NVIDIA Switch 2 GPU has a base clock of 561 MHz and a boost clock of 1400 MHz. The memory clock for the Intel part is listed as system shared, while the NVIDIA part has a memory clock of 800 MHz with 6.4 Gbps effective.

Memory: The Intel Arc Pro B390 uses system shared memory with system dependent size, type, bus width, and bandwidth. The NVIDIA Switch 2 GPU has 12 GB of LPDDR5X memory on a 128-bit bus with 102.4 GB/s bandwidth.

ROPs: The Intel Arc Pro B390 has 24 ROPs, while the NVIDIA Switch 2 GPU has 16 ROPs.

Tensor Cores: The Intel Arc Pro B390 has no tensor cores listed, while the NVIDIA Switch 2 GPU has 48 tensor cores.

Pixel and Texture Rates: The Intel Arc Pro B390 has a pixel rate of 60.00 GPixel/s and a texture rate of 120.0 GTexel/s. The NVIDIA Switch 2 GPU has a pixel rate of 22.40 GPixel/s and a texture rate of 67.20 GTexel/s.

FP32 and FP16: The Intel Arc Pro B390 delivers 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16 (2:1). The NVIDIA Switch 2 GPU delivers 4.301 TFLOPS FP32 and 8.602 TFLOPS FP16 (2:1).

TDP: The Intel Arc Pro B390 has a TDP of 80 W, while the NVIDIA Switch 2 GPU has a TDP of 40 W.

Physical Form: The Intel Arc Pro B390 has an IGP slot width, no power connectors, and an IGP bus interface. The NVIDIA Switch 2 GPU has dimensions of 272 mm by 116 mm by 14 mm, no display outputs, and no slot width or power connector information.

Release Date: The Intel Arc Pro B390 has a release date of 2026-01-26, while the NVIDIA Switch 2 GPU has a release date of 2025-06-04. The NVIDIA part has a launch MSRP of 449 USD, while the Intel part has no launch MSRP listed.

Head-to-Head Benchmarks

The database records no head-to-head benchmark results for these two GPUs, and neither has individual benchmark scores or nearest rival data. The analysis therefore relies on the specification-derived rates and the percentile rankings.

The Intel Arc Pro B390 shows a decisive lead in compute throughput. Its FP32 rating of 7.680 TFLOPS is 1.79 times the 4.301 TFLOPS of the NVIDIA Switch 2 GPU. This means for any FP32 workload, the Intel part has the potential to process nearly twice as many floating-point operations per second, assuming the clocks are sustained at the listed boost values.

The pixel fill rate advantage is even larger. The Intel Arc Pro B390 achieves 60.00 GPixel/s, which is 2.68 times the 22.40 GPixel/s of the NVIDIA Switch 2 GPU. This suggests a substantial advantage in rasterization-heavy scenes where pixel output is the limiting factor. The texture rate of 120.0 GTexel/s versus 67.20 GTexel/s gives the Intel part a 1.79 times advantage in texture mapping operations.

The NVIDIA Switch 2 GPU has a lower base clock of 561 MHz compared to 300 MHz for the Intel part, but the Intel part has a much higher boost clock of 2500 MHz versus 1400 MHz. The NVIDIA part has a narrower clock range, suggesting a more consistent performance profile, while the Intel part relies on aggressive boosting to reach its rated output.

The memory bandwidth situation favors the NVIDIA Switch 2 GPU in one respect: it has a fixed 102.4 GB/s of bandwidth, while the Intel Arc Pro B390 has system dependent bandwidth that could be higher or lower depending on the host memory configuration. However, the Intel part also has a higher pixel rate and texture rate, which may compensate in many workloads.

Both GPUs are listed at the 50th percentile among all GPUs, which places them at the median of the performance distribution. This percentile is identical for both, suggesting that in overall performance terms they are expected to land in the same tier, despite the specification differences favoring the Intel part in raw rates.

Where Each One Wins

The Intel Arc Pro B390 wins in scenarios that demand high raw compute throughput. Its FP32 rating of 7.680 TFLOPS is 1.79 times higher than the NVIDIA Switch 2 GPU, making it the stronger choice for compute-heavy applications such as general-purpose GPU computing, physics simulations, or any workload that can use the shading units at full capacity. The pixel rate of 60.00 GPixel/s, which is 2.68 times higher, gives it an edge in high-resolution rendering or scenes with heavy overdraw. The texture rate of 120.0 GTexel/s, 1.79 times higher, benefits texture-heavy rendering paths.

The NVIDIA Switch 2 GPU wins in power-constrained environments. Its TDP of 40 W is exactly half of the 80 W TDP of the Intel Arc Pro B390. For battery-powered portable devices or systems with limited cooling, the NVIDIA part operates at lower power draw while still delivering 4.301 TFLOPS of FP32 compute. The presence of 48 tensor cores gives the NVIDIA part a dedicated hardware path for tensor operations, which the Intel part lacks entirely. This could make the NVIDIA Switch 2 GPU more suitable for AI inference or machine learning tasks that leverage tensor cores.

The NVIDIA Switch 2 GPU also wins on memory determinism. It has a fixed 12 GB of LPDDR5X memory with 102.4 GB/s bandwidth, while the Intel Arc Pro B390 relies on system shared memory with system dependent characteristics. In a scenario where the host system has limited memory bandwidth, the Intel part could be disadvantaged, whereas the NVIDIA part has a guaranteed memory configuration.

The Intel Arc Pro B390 wins on ROP count with 24 ROPs versus 16 ROPs for the NVIDIA part, reinforcing its pixel fill advantage. It also has a higher boost clock of 2500 MHz versus 1400 MHz, which contributes to its higher compute rates. The Intel part is built on a 3 nm process node compared to 8 nm for the NVIDIA part, which may contribute to its higher clock speeds and lower power per transistor, though the NVIDIA part still achieves a lower overall TDP.

For integrated systems where the GPU must share memory with the CPU, the Intel Arc Pro B390 benefits from higher clock-driven compute rates. For console-style systems where the GPU has dedicated memory and operates within a strict power budget, the NVIDIA Switch 2 GPU provides a balanced package with tensor core support and half the power draw. The release dates differ, with the NVIDIA part released on 2025-06-04 and the Intel part on 2026-01-26, making the NVIDIA part earlier to market.

Both GPUs support the same API set: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. This means game and application compatibility is equivalent at the API level. The deciding factors come down to compute throughput, power efficiency, memory configuration, and the presence of tensor cores. The Intel Arc Pro B390 leads in raw performance metrics, while the NVIDIA Switch 2 GPU leads in power efficiency and dedicated tensor hardware.