Intel Arc B390 vs NVIDIA RTX PRO 6000 Blackwell Comparison

Head-to-Head Benchmarks

The recorded data contains a single direct comparison between the Intel Arc B390 and the NVIDIA RTX PRO 6000 Blackwell: the 3DMark Steel Nomad DX12 test. In this benchmark, the NVIDIA RTX PRO 6000 Blackwell delivers a score of 16408, while the Intel Arc B390 reaches only 1482. The delta is a decisive 91 percent deficit for the Intel part, which places it far behind in absolute rendering throughput.

The NVIDIA card’s score of 16408 places it in the 59th percentile among all GPUs in the database. Its nearest rivals include the AMD Radeon PRO W7500 at 16415 (a 0 percent delta), the AMD Radeon RX 5700 XT at 16361 (0.3 percent ahead of the RTX PRO 6000), and the AMD Radeon Pro 5600M at 16351 (0.4 percent ahead). Notably, the NVIDIA GeForce RTX 5090 D V2 scores 16504, which is 0.6 percent higher than the RTX PRO 6000, meaning the workstation card trails that consumer flagship by a narrow margin. These figures indicate that the RTX PRO 6000 Blackwell sits in a competitive cluster near the top of the database’s measured results, trading places with other high-end parts within a fraction of a percent.

The Intel Arc B390, by contrast, scores 1482 and sits in the 9th percentile. Its nearest rivals are older, low-end parts: the NVIDIA GeForce GT 520MX at 1463 (1.3 percent behind the B390), the NVIDIA GeForce 800M at 1460 (1.5 percent behind), the NVIDIA GeForce GT 625 OEM at 1446 (2.5 percent behind), and the NVIDIA GeForce GT 710 at 1443 (2.7 percent behind). The B390 leads this group, but the margins are small, and the group itself occupies the lower tail of the performance distribution. The 91 percent gap between the two units is not merely a generational difference; it reflects a fundamental positioning gap between an integrated graphics solution and a dedicated workstation accelerator.

FAQ

Q: Which GPU wins the only recorded benchmark?

A: The NVIDIA RTX PRO 6000 Blackwell wins the 3DMark Steel Nomad DX12 test with a score of 16408 versus the Intel Arc B390’s 1482, a 91 percent advantage.

Q: How does the Intel Arc B390 compare to its nearest rivals?

A: The B390 leads its nearest recorded rivals by slim margins: 1.3 percent over the GeForce GT 520MX, 1.5 percent over the GeForce 800M, 2.5 percent over the GeForce GT 625 OEM, and 2.7 percent over the GeForce GT 710.

Q: Where does the RTX PRO 6000 Blackwell sit among its nearest rivals?

A: The RTX PRO 6000 Blackwell scores 16408, which is essentially tied with the AMD Radeon PRO W7500 (16415, 0 percent delta), slightly ahead of the AMD Radeon RX 5700 XT (16361, 0.3 percent) and AMD Radeon Pro 5600M (16351, 0.4 percent), and 0.6 percent behind the NVIDIA GeForce RTX 5090 D V2 (16504).

Q: What is the percentile ranking for each GPU?

A: The Intel Arc B390 ranks in the 9th percentile among all GPUs, while the NVIDIA RTX PRO 6000 Blackwell ranks in the 59th percentile.

Q: What is the memory configuration of the RTX PRO 6000 Blackwell?

A: The RTX PRO 6000 Blackwell uses 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: What is the memory configuration of the Intel Arc B390?

A: The Intel Arc B390 uses System Shared memory, meaning its memory size, type, bus width, and bandwidth are all system-dependent; the recorded data lists bandwidth as system dependent.

Architecture Differences

The two GPUs come from entirely different design philosophies. The Intel Arc B390 is built on the Xe3-LPG architecture, using the Panther Lake chip, and is fabricated on a 3 nm process at Intel’s own foundry. It functions as an integrated graphics processor (IGP), which means it shares system memory and has no dedicated VRAM. Its transistor count and die size are recorded as unknown, reflecting its integrated nature and the lack of discrete measurement data.

The NVIDIA RTX PRO 6000 Blackwell uses the Blackwell 2.0 architecture on the GB202 chip, fabricated by TSMC on a 5 nm process. It packs 92,200 million transistors on a 750 mm² die, yielding a transistor density of 122.9 million transistors per square millimeter. This is a discrete, dual-slot card with a 304 mm length, 137 mm height, and 40 mm width. It connects via PCIe 5.0 x16 and draws power from a single 16-pin connector, with a 600 W TDP and a recommended 1000 W power supply.

The compute resources differ massively. The Intel Arc B390 has 1536 shading units, 48 texture mapping units, 24 raster output pipelines, and 12 ray tracing cores. The NVIDIA RTX PRO 6000 Blackwell has 24064 shading units, 752 TMUs, 192 ROPs, 188 ray tracing cores, and 752 tensor cores. The RTX PRO 6000 also supports FP16 at a 1:1 ratio with FP32, both at 126.0 TFLOPS, whereas the Arc B390 delivers FP16 at a 2:1 ratio, reaching 15.36 TFLOPS versus 7.680 TFLOPS for FP32. Pixel rate for the NVIDIA card is 502.5 GPixel/s versus 60.00 GPixel/s for Intel; texture rate is 1,968.0 GTexel/s versus 120.0 GTexel/s.

Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The NVIDIA card outputs to 4x DisplayPort 2.1b, while the Intel part’s display outputs are listed as portable device dependent. The Intel B390 has no power connectors and an IGP bus interface, while the NVIDIA card uses a dual-slot form factor.

Specification Differences

The specification sheet separates the two units across nearly every field. The Intel Arc B390 has a base clock of 300 MHz and a boost clock of 2500 MHz; the NVIDIA RTX PRO 6000 Blackwell has a base clock of 1590 MHz and a boost clock of 2617 MHz. The Intel part’s memory is system shared, while the NVIDIA part uses 96 GB of GDDR7 with a 512-bit bus and 1.79 TB/s bandwidth.

Compute resources: the Arc B390 has 1536 shading units, 48 TMUs, 24 ROPs, and 12 RT cores, with no tensor cores listed. The RTX PRO 6000 has 24064 shading units, 752 TMUs, 192 ROPs, 188 RT cores, and 752 tensor cores. FP32 throughput is 7.680 TFLOPS for Intel versus 126.0 TFLOPS for NVIDIA. FP16 throughput is 15.36 TFLOPS (2:1) for Intel versus 126.0 TFLOPS (1:1) for NVIDIA. Pixel rate is 60.00 GPixel/s versus 502.5 GPixel/s, and texture rate is 120.0 GTexel/s versus 1,968.0 GTexel/s.

Power and physical specifications differ drastically. The Arc B390 consumes 80 W, is an IGP, has no power connectors, and requires no separate power supply rating. The RTX PRO 6000 consumes 600 W, is dual-slot, uses a 16-pin connector, and requires a 1000 W power supply. The Intel part’s release date is 2026-01-26, while the NVIDIA part’s release date is 2025-03-17. The NVIDIA card has a launch MSRP of 8,565 USD, which can be stated once as a reference point. The Intel part has no launch MSRP recorded. The NVIDIA card’s predecessor is listed as Workstation Ada; the Intel part has no predecessor recorded.

The Verdict

The benchmark data makes the choice unambiguous for raw 3D performance. The NVIDIA RTX PRO 6000 Blackwell delivers a 3DMark Steel Nomad DX12 score of 16408, which is 91 percent higher than the Intel Arc B390’s 1482. The NVIDIA card sits in the 59th percentile, surrounded by peers within a 0.6 percent band, indicating it competes at the high end of the database’s results. The Intel Arc B390 sits in the 9th percentile, grouped with legacy low-end discrete parts from over a decade ago.

For any workload that relies on the measured DX12 rendering path, the RTX PRO 6000 is the only viable option between the two. The Arc B390’s integrated nature, system-shared memory, and 80 W power envelope place it in a completely different performance class. Its 9th percentile ranking means it trails the vast majority of GPUs in the database, not just the RTX PRO 6000. The 91 percent delta is not a close contest; it is a categorical separation.

The RTX PRO 6000’s 59th percentile ranking, however, is not a top-of-the-chart position. It trades within 0.6 percent of the GeForce RTX 5090 D V2 and essentially ties the Radeon PRO W7500. This suggests that while the RTX PRO 6000 is far ahead of the Arc B390, it is not the absolute fastest GPU in the database; it sits in a competitive cluster where small score differences separate peers.

Where Each One Wins

The Intel Arc B390 wins in exactly one category from the recorded data: proximity to the bottom of the performance distribution among its nearest rivals. It leads the GeForce GT 520MX, GeForce 800M, GeForce GT 625 OEM, and GeForce GT 710 by margins between 1.3 and 2.7 percent. That group represents the lowest tier of the database, so the B390’s wins there are strictly relative to very weak hardware. The B390 also has the advantage of an integrated form factor, requiring no power connectors and consuming 80 W, which makes it suitable for compact portable systems where a discrete card cannot be installed.

The NVIDIA RTX PRO 6000 Blackwell wins the only head-to-head benchmark and therefore wins the only direct comparison available. It also wins on every absolute performance metric in the specification sheet: FP32 throughput (126.0 TFLOPS versus 7.680 TFLOPS), FP16 throughput (126.0 TFLOPS versus 15.36 TFLOPS), pixel rate (502.5 GPixel/s versus 60.00 GPixel/s), texture rate (1,968.0 GTexel/s versus 120.0 GTexel/s), and memory bandwidth (1.79 TB/s versus system dependent). It has 96 GB of dedicated GDDR7 memory versus shared system memory, and it provides 752 tensor cores where the Arc B390 lists none.

For a workstation context, the RTX PRO 6000’s 96 GB capacity, 512-bit bus, and 1.79 TB/s bandwidth are the data-backed reasons it would be chosen for memory-intensive rendering or large model workloads. The Arc B390’s system-shared memory means its bandwidth is dependent on the host system, which is a structural disadvantage for any workload requiring sustained memory throughput. The RTX PRO 6000 also has a higher boost clock (2617 MHz versus 2500 MHz) and a higher base clock (1590 MHz versus 300 MHz), reinforcing its compute lead.

The only scenario where the Arc B390 appears preferable from the data is one where the integrated form factor, 80 W power draw, and lack of external power connectors are the deciding factors. That scenario does not involve the recorded benchmark, because the B390 loses that test by 91 percent. The data shows no benchmark where the Intel part wins, and no metric in the specification sheet where it leads the NVIDIA card. The verdict is therefore straightforward: the RTX PRO 6000 wins the performance comparison, and the Arc B390 wins only in the context of its low-power integrated design, which the database records but does not measure as a performance advantage.