Ryzen 5 3600 + RTX 3070 Ti in Baldur's Gate 3

Every figure below is a measurement aggregated from published review charts — the CPU's frame-rate ceiling and the GPU's frame-rate ceiling in this game, combined. Settings: Ultra preset, DX11, native rendering, no upscaling, no frame generation. CPU ceiling measured at 1080p, Medium preset (reviewers' CPU-isolation method), flagship GPU, Act III.

1080p66–66 fps · Strong CPU limit1440p66–66 fps · Strong CPU limit4K59–66 fps · Balanced pairing

1080p

Strong CPU limit

Expected frame rate: 66–66 fps (measured range across sources).

CPU Ryzen 5 3600ceiling ~66 fps on its own
GPU RTX 3070 Ticeiling ~132 fps on its own

The CPU side is lower by about 50.3%: the pair runs at the lower ceiling, and upgrading the other part alone would not raise it.

That range suits a 60 Hz monitor; a high-refresh panel would be underused in this game.

Confidence: MEDIUM — based on 2 source measurements for these two parts.

1440p

Strong CPU limit

Expected frame rate: 66–66 fps (measured range across sources).

CPU Ryzen 5 3600ceiling ~66 fps on its own
GPU RTX 3070 Ticeiling ~98 fps on its own

The CPU side is lower by about 33.4%: the pair runs at the lower ceiling, and upgrading the other part alone would not raise it.

That range suits a 60 Hz monitor; a high-refresh panel would be underused in this game.

Confidence: MEDIUM — based on 2 source measurements for these two parts.

4K

Balanced pairing

Expected frame rate: 59–66 fps (measured range across sources).

CPU Ryzen 5 3600ceiling ~66 fps on its own
GPU RTX 3070 Ticeiling ~59–69 fps on its own

The two measured ceilings sit within 2.3% of each other — neither part wastes much of the other.

That range suits a 60 Hz monitor; a high-refresh panel would be underused in this game.

Confidence: MEDIUM — based on 3 source measurements for these two parts.

CPUs measured in this game that clear this GPU:

This pair in other measured games

Sources for Baldur's Gate 3

Full data: Baldur's Gate 3 benchmarks · how verdicts are computed.