Intel Arc B390 vs NVIDIA GeForce RTX 5060 Ti 8 GB Comparison

The Verdict

The recorded database positions these two parts in entirely different performance strata. The NVIDIA GeForce RTX 5060 Ti 8 GB delivers a 3DMark Steel Nomad DX12 score of 4019, while the Intel Arc B390 scores 1482. That is a 63.1% deficit for the Intel part in the only shared benchmark, making the NVIDIA card the clear choice for any workload represented by that test. The Intel Arc B390 sits at the 9th percentile of all GPUs in the database, with nearest rivals like the NVIDIA GeForce GT 520MX (1463 score, 1.3% delta) and the NVIDIA GeForce 800M (1460 score, 1.5% delta). The RTX 5060 Ti 8 GB, by contrast, ranks at the 29th percentile, with its nearest rivals being the AMD FirePro W5130M (4904 score, -0.1% delta) and the NVIDIA GeForce GTS 450 (4893 score, 0.2% delta). For a user seeking discrete desktop graphics performance, the RTX 5060 Ti 8 GB is the only defensible selection from this data. The Arc B390, as an integrated graphics processor, serves a fundamentally different role.

Architecture Differences

The two products diverge sharply at the architectural level. The Intel Arc B390 uses the Xe3-LPG architecture, built on an Intel 3 nm process, and is part of the Arc Graphics-M (Panther Lake) generation. It is an integrated graphics processor (IGP) with no dedicated slot width, no power connectors, and a bus interface of IGP. Its memory subsystem is entirely system shared, meaning it draws from main system memory rather than a dedicated VRAM pool. The chip contains 1536 shading units, 48 texture mapping units, 24 raster output units, and 12 ray tracing cores. It carries no dedicated tensor cores. The base clock is 300 MHz with a boost clock of 2500 MHz. Its FP32 throughput is 7.680 TFLOPS, and FP16 throughput is 15.36 TFLOPS using a 2:1 ratio.

The NVIDIA GeForce RTX 5060 Ti 8 GB uses the Blackwell 2.0 architecture on a 5 nm process from TSMC, with the GB206 chip. It is a discrete, dual-slot card with a PCIe 5.0 x8 interface. The chip contains 21,900 million transistors on a 181 mm² die, giving a transistor density of 121.0M per mm². It has 4608 shading units, 144 TMUs, 48 ROPs, 36 RT cores, and 144 tensor cores. Base clock is 2407 MHz, boost clock is 2572 MHz. FP32 throughput is 23.70 TFLOPS, and FP16 is 23.70 TFLOPS at a 1:1 ratio. The RTX card uses 8 GB of GDDR7 memory on a 128-bit bus, with 448.0 GB/s bandwidth and a memory clock of 1750 MHz (28 Gbps effective). The Intel part's memory bandwidth is system dependent, with no fixed figure. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RTX 5060 Ti 8 GB lists a 180 W TDP with a 1x 8-pin power connector and a suggested PSU of 450 W. The Arc B390 lists an 80 W TDP.

Head-to-Head Benchmarks

The database includes exactly one shared test between the two: 3DMark Steel Nomad DX12. In that benchmark, the NVIDIA GeForce RTX 5060 Ti 8 GB scores 4019, while the Intel Arc B390 scores 1482. The delta is -63.1% for the Intel part, meaning the RTX 5060 Ti 8 GB is roughly 2.7 times faster in this workload. This is not a marginal gap; it is a decisive margin that places the two in different performance classes. The RTX 5060 Ti 8 GB also has additional benchmark entries in the database: PassMark DirectX 10 (142), DirectX 11 (220), DirectX 12 (82), DirectX 9 (228), G2D (1169), G3D (21981), and GPU Compute (11365). The Arc B390 has no corresponding entries in those tests, so no comparison is possible beyond the single 3DMark result. The wins tally reflects this: the NVIDIA card wins 1 head-to-head comparison, the Intel part wins 0.

The percentile data reinforces the same conclusion. The RTX 5060 Ti 8 GB sits at the 29th percentile of all GPUs, which places it above roughly a quarter of the database population. Its average benchmark score across all recorded tests is 4901. The Arc B390, with only one recorded benchmark, has an average score of 1482 and sits at the 9th percentile. The nearest rival list for the Arc B390 includes older, low-end NVIDIA parts: the GT 520MX (1463), the 800M (1460), the GT 625 OEM (1446), and the GT 710 (1443). All of these are within 2.7% of the Arc B390 score, confirming that the Intel part clusters with entry-level legacy hardware rather than with modern discrete GPUs. The RTX 5060 Ti 8 GB's nearest rivals are similarly close: the FirePro W5130M is 0.1% faster, the GTS 450 is 0.2% slower, and the R7 M265 and R7 M360 are 0.6% faster and slower respectively.

Specification Differences

The following fields differ between the two parts in the database:

• Architecture: Intel Xe3-LPG versus NVIDIA Blackwell 2.0
• Process node: 3 nm (Intel) versus 5 nm (TSMC)
• Transistors: unknown versus 21,900 million
• Die size: unknown versus 181 mm²
• Transistor density: not listed versus 121.0M / mm²
• Base clock: 300 MHz versus 2407 MHz
• Boost clock: 2500 MHz versus 2572 MHz
• Memory size: System Shared versus 8 GB
• Memory type: System Shared versus GDDR7
• Memory bus width: System Shared versus 128 bit
• Memory bandwidth: System Dependent versus 448.0 GB/s
• Memory clock: System Shared versus 1750 MHz (28 Gbps effective)
• Shading units: 1536 versus 4608
• TMUs: 48 versus 144
• ROPs: 24 versus 48
• RT cores: 12 versus 36
• Tensor cores: not listed versus 144
• Pixel rate: 60.00 GPixel/s versus 123.5 GPixel/s
• Texture rate: 120.0 GTexel/s versus 370.4 GTexel/s
• FP32: 7.680 TFLOPS versus 23.70 TFLOPS
• FP16: 15.36 TFLOPS (2:1) versus 23.70 TFLOPS (1:1)
• TDP: 80 W versus 180 W
• Slot width: IGP versus Dual-slot
• Power connectors: None versus 1x 8-pin
• Suggested PSU: not listed versus 450 W
• Bus interface: IGP versus PCIe 5.0 x8
• Display outputs: Portable Device Dependent versus 1x HDMI 2.1b, 3x DisplayPort 2.1b
• Dimensions: not listed versus 241 mm length, 111 mm height, 40 mm width
• Release date: 2026-01-26 versus 2025-04-15
• Predecessor: not listed versus GeForce 40
• Successor: not listed versus GeForce 60
• Launch MSRP: not listed versus 379 USD (stated once here as launch MSRP)

The Intel part has no fixed memory capacity, bus width, or bandwidth because it relies on shared system memory. The NVIDIA card has a fixed 8 GB GDDR7 allocation. The transistor counts and die sizes are only recorded for the NVIDIA part. Both support the same API versions.

FAQ

Q: Which GPU is faster in the shared 3DMark Steel Nomad DX12 benchmark?

A: The NVIDIA GeForce RTX 5060 Ti 8 GB scores 4019, while the Intel Arc B390 scores 1482. The delta is -63.1% for the Intel part.

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

A: The Arc B390's nearest rivals are the NVIDIA GeForce GT 520MX (1463 score, 1.3% delta), the NVIDIA GeForce 800M (1460, 1.5% delta), the NVIDIA GeForce GT 625 OEM (1446, 2.5% delta), and the NVIDIA GeForce GT 710 (1443, 2.7% delta). All are within a few percent of the Intel part's 1482 score.

Q: What are the memory specifications for each card?

A: The Intel Arc B390 uses system shared memory with a system dependent bandwidth. The NVIDIA GeForce RTX 5060 Ti 8 GB uses 8 GB of GDDR7 on a 128-bit bus with 448.0 GB/s bandwidth.

Q: Which part has more shading units and RT cores?

A: The NVIDIA GeForce RTX 5060 Ti 8 GB has 4608 shading units and 36 RT cores. The Intel Arc B390 has 1536 shading units and 12 RT cores.

Q: What is the thermal design power difference?

A: The Intel Arc B390 has a TDP of 80 W, while the NVIDIA GeForce RTX 5060 Ti 8 GB has a TDP of 180 W and requires a 1x 8-pin power connector with a suggested PSU of 450 W.

Q: What are the release dates in the database?

A: The Intel Arc B390 has a release date of 2026-01-26. The NVIDIA GeForce RTX 5060 Ti 8 GB has a release date of 2025-04-15.