Intel Arc B580 vs NVIDIA GeForce RTX 5050

Every spec side by side, from each maker's published figures.

Better figure, where one is clearly better

Arc B580 compared with GeForce RTX 5050
IntelArc B580NVIDIAGeForce RTX 5050
Overview
ReleasedDecember 2024July 2025
Performance
Chip makerIntelNVIDIA
ArchitectureXe2 (Battlemage)Blackwell
Shader cores2,5602,560
Boost clock–2,570 MHz
Architecture20 Xe-cores in 5 render slices; 160 Xe Vector Engines; Intel lists a 2670 MHz graphics clock–
Manufacturing processTSMC N5–
Ray tracing20 second-generation ray tracing unitsFourth-generation RT Cores; 40 RT TFLOPS
AI cores160 Intel XMX engines; 233 TOPS INT8Fifth-generation Tensor Cores; 421 AI TOPS
Upscaling and frame generationXeSS 2 with XeSS Super Resolution, XeSS Frame Generation and Xe Low LatencyDLSS 4.5 with Super Resolution, DLAA, Ray Reconstruction, Frame Generation, Multi Frame Generation and Dynamic Multi Frame Generation
Base clock–2.31 GHz
Memory
Video memory12 GB GDDR6 (better)8 GB GDDR6
Memory bus192-bit (better)128-bit
Memory bandwidth456 GB/s–
Memory detailGDDR6 at 19 Gbps–
Connections and power
Board power190 W130 W (better)
InterfacePCIe 4.0 x8PCIe 5.0
Display outputsHDMI 2.1, DisplayPort 2.1 up to UHBR 13.5, DisplayPort 2.1 up to UHBR 10; up to 4 displays3 × DisplayPort 2.1b (UHBR20), 1 × HDMI 2.1b; up to 4 displays; HDCP 2.3
Maximum resolution7680 × 4320 at 120 Hz over HDMI, 7680 × 4320 at 60 Hz over DisplayPort4K at 480 Hz or 8K at 165 Hz with DSC
Power connector1 × 8-pin (Limited Edition)1 × PCIe 8-pin cable, or a 300 W or greater PCIe Gen 5 cable (adapter in the box)
Recommended power supply600 W550 W
Slot width2-slot (Limited Edition)–
Size (reference card)272 × 115 mm, 779 g (Limited Edition)–
Video encode and decode2 multi-format codec engines; H.264, HEVC and AV1 encode and decode; HDR10, HDR10+ Gaming and Dolby Vision1 × ninth-generation NVENC, 1 × sixth-generation NVDEC; AV1 encode and decode

More comparisons