Intel Arc Pro B390 vs NVIDIA RTX PRO 6000 Blackwell Max-Q Comparison

Where Each One Wins

The two products occupy entirely separate segments of the graphics market, and the recorded data reflects that split clearly. The Intel Arc Pro B390 is an integrated graphics processor (IGP) built into the Panther Lake platform, with no discrete memory, no standalone board, and a power envelope of 80 W. Its role is to provide display output and basic acceleration for portable devices, with system-shared memory and system-dependent bandwidth. The NVIDIA RTX PRO 6000 Blackwell Max-Q is a dual-slot, 300 W discrete workstation card with 96 GB of dedicated GDDR7 memory, a 512-bit bus, and 1.79 TB/s of bandwidth.

The Intel part wins on integration and power efficiency. It draws 80 W, uses no power connectors, and requires no suggested PSU rating, making it suitable for compact, low-power systems. Its base clock of 300 MHz and boost of 2500 MHz are tuned for thermal and power constraints typical of integrated designs. The NVIDIA card, by contrast, requires a 700 W suggested PSU and a 16-pin power connector, and its physical footprint is 267 mm in length, 111 mm in height, and 40 mm in width.

The NVIDIA part wins on raw compute and memory capacity. Its FP32 throughput of 109.7 TFLOPS dwarfs the Intel part's 7.680 TFLOPS, and its 96 GB memory pool is orders of magnitude larger than any system-shared allocation. The NVIDIA card also delivers 437.8 GPixel/s pixel rate versus 60.00 GPixel/s for the Intel IGP, and 1,714.6 GTexel/s texture rate versus 120.0 GTexel/s. The data confirms that the Intel part is for basic integrated duties, while the NVIDIA part is for heavy rendering, AI, and large dataset workloads.

Architecture Differences

The Intel Arc Pro B390 uses the Xe3-LPG architecture on a 3 nm process from Intel, built on the Panther Lake chip. It integrates 1536 shading units, 48 texture mapping units, 24 ROPs, and 12 ray tracing cores. It has no dedicated tensor cores, and its FP16 throughput of 15.36 TFLOPS is achieved at a 2:1 ratio relative to FP32. The memory subsystem is entirely system-shared, with no dedicated VRAM, no fixed bus width, and bandwidth that varies with the host system. The graphics API support includes DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The NVIDIA RTX PRO 6000 Blackwell Max-Q uses the Blackwell 2.0 architecture on a 5 nm process from TSMC, built on the GB202 chip. The die measures 750 mm² and contains 92,200 million transistors, giving a transistor density of 122.9M per mm². It has 24,064 shading units, 752 TMUs, 192 ROPs, 188 ray tracing cores, and 752 tensor cores. Its FP16 throughput matches FP32 at 109.7 TFLOPS with a 1:1 ratio, indicating equal rate for both precisions. The memory is 96 GB of GDDR7 on a 512-bit bus, running at 1750 MHz with 28 Gbps effective speed, yielding 1.79 TB/s bandwidth. The card uses a PCIe 5.0 x16 interface and provides four DisplayPort 2.1b outputs.

The transistor counts alone tell a significant story. The NVIDIA chip packs 92,200 million transistors into 750 mm², while the Intel part's transistor count is listed as unknown in the database. The process nodes differ by 2 nm, with Intel at 3 nm and NVIDIA at 5 nm, though the NVIDIA die is substantially larger. The Intel part has no dedicated memory clock, instead relying on system memory, while the NVIDIA memory clock is fixed at 1750 MHz. Both support DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, so API compatibility is not a differentiator.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between the Intel Arc Pro B390 and the NVIDIA RTX PRO 6000 Blackwell Max-Q. The wins counter shows 0 for both parts, and the head-to-head benchmark array is empty. However, the NVIDIA card has a recorded benchmark score from 3DMark Steel Nomad DX12, which is absent for the Intel part.

The NVIDIA RTX PRO 6000 Blackwell Max-Q scores 11,088 in 3DMark Steel Nomad DX12. Its nearest rivals in the database are the NVIDIA RTX PRO 6000D Blackwell Max-Q with an identical average score of 11,088 and a delta of 0%, the AMD Radeon RX 550 with an average score of 11,075 and a delta of 0.1%, the NVIDIA GeForce GTX 1650 SUPER with an average score of 11,047 and a delta of 0.4%, and the AMD FirePro W4300 with an average score of 11,225 and a delta of -1.2%. The data shows the RTX PRO 6000 Blackwell Max-Q is effectively tied with its sibling RTX PRO 6000D, marginally ahead of the RX 550 and GTX 1650 SUPER, and slightly behind the FirePro W4300.

The Intel Arc Pro B390 has no benchmark entries, no average score, and no nearest rivals listed. Its percentile rank against all GPUs is 50, which is the same percentile as the NVIDIA card. This percentile indicates a median position in the overall database distribution, though the actual score of 0 for the Intel part suggests a lack of recorded performance data rather than a literal zero-performance result. The NVIDIA card's average benchmark score is 11,088, confirming it is a functioning, measured product.

Without direct head-to-head results, the comparison relies on the architectural specifications. The NVIDIA card's FP32 rate of 109.7 TFLOPS is approximately 14.3 times higher than the Intel part's 7.680 TFLOPS. The texture rate difference is even larger: 1,714.6 GTexel/s versus 120.0 GTexel/s, a factor of about 14.3 as well. The pixel rate difference is a factor of about 7.3, with 437.8 GPixel/s versus 60.00 GPixel/s. These ratios are derived directly from the recorded specifications in the database.

The Verdict

The data supports a clear separation of roles. For integrated graphics in portable or low-power systems, the Intel Arc Pro B390 is the only option between the two, as it is an IGP with no slot width, no power connectors, and an 80 W TDP. It uses system-shared memory, making it dependent on the host platform for bandwidth, and it provides display outputs that are portable device dependent. Its 1536 shading units and 12 ray tracing cores are adequate for basic acceleration, but the 7.680 TFLOPS FP32 rate and 60.00 GPixel/s pixel rate place it firmly in the entry-level segment.

For professional workstation workloads, the NVIDIA RTX PRO 6000 Blackwell Max-Q is the clear choice based on measured performance. Its 3DMark Steel Nomad DX12 score of 11,088 is recorded and verified, while the Intel part has no scores at all. The NVIDIA card's 96 GB of GDDR7 memory, 1.79 TB/s bandwidth, and 109.7 TFLOPS FP32 throughput are specification-level advantages that no integrated part can match. The 752 tensor cores and 188 ray tracing cores also indicate substantial acceleration for AI and ray-traced workloads, features entirely absent from the Intel IGP.

The launch MSRP for the NVIDIA RTX PRO 6000 Blackwell Max-Q is 8,565 USD. The Intel Arc Pro B390 has no launch MSRP listed, consistent with its role as an integrated component rather than a standalone retail product. The NVIDIA card's predecessor is Workstation Ada, while the Intel part's predecessor is HD Graphics-WM, further confirming the different market positions. Users seeking a discrete workstation card with large memory and high compute should select the NVIDIA part; users requiring only integrated graphics for a portable device should rely on the Intel IGP.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The NVIDIA RTX PRO 6000 Blackwell Max-Q delivers 109.7 TFLOPS FP32, while the Intel Arc Pro B390 delivers 7.680 TFLOPS FP32, a difference of roughly 14.3 times.

Q: Does the Intel Arc Pro B390 have dedicated video memory?

A: No. The Intel part uses system-shared memory with a system-dependent bandwidth, and its memory type and bus width are both listed as system shared.

Q: What is the measured benchmark score for the NVIDIA RTX PRO 6000 Blackwell Max-Q?

A: It scores 11,088 in 3DMark Steel Nomad DX12, with an average benchmark score of 11,088. Its nearest rival, the RTX PRO 6000D Blackwell Max-Q, scores the same 11,088.

Q: What is the power consumption of each GPU?

A: The Intel Arc Pro B390 has a TDP of 80 W, while the NVIDIA RTX PRO 6000 Blackwell Max-Q has a TDP of 300 W and requires a suggested PSU of 700 W.

Q: How much memory does the NVIDIA RTX PRO 6000 Blackwell Max-Q have?

A: It has 96 GB of GDDR7 memory on a 512-bit bus, with a bandwidth of 1.79 TB/s and a memory clock of 1750 MHz (28 Gbps effective).

Q: Does the Intel Arc Pro B390 have tensor cores?

A: No. The Intel part lists no tensor cores, while the NVIDIA card has 752 tensor cores. The Intel part does have 12 ray tracing cores, compared to 188 on the NVIDIA card.