Despite having a $130 higher MSRP, the RTX 5060 Ti 16GB provides a better price-to-performance ratio than its 8GB counterpart, even at 1080p. At 1080p ultra, the 16GB provides 7% greater FPS per dollar
Assuming you can find an RTX 5060 Ti 16GB at its $429 MSRP, there's only a $50 difference between that card and the 8GB version.
This math seems fishy.
MSRP: $379 (8 GB), $429 (16 GB)
Street: Newegg is actually throwing a $20 "rebate card" (never heard of that) alongside a $10 promo code, so you could get 8 GB for about $350. 16 GB model can be obtained for $450.
So straight up, it's not a $130 higher MSRP, it's $50 as stated later. The street pricing gap isn't $130 either. I'm not sure where that line came from. Edit: It was probably supposed to be "13%" instead of "$130".
16 GB model is 13% more expensive (MSRP) or about 18-29% more expensive (street). That's 18% if you get the 8 GB for $380, 29% if you get it at $350.
1080p ultra
21 game geomean = +12% avg, +16% lows
15 game raster geomean = +5% avg, 8% lows
6 game raytracing geomean = +37% avg, +44% lows
So with the MSRPs and the 21 game geomean average (with 1080p ultra being a better case for 16 GB than 1080p medium), the 5060 Ti 16 GB's performance is not better than the price difference (+13% money for +12% performance).
Raytracing on the 8 GB model is being dragged down significantly by the Indiana Jones result. It could be argued that's fair, as it indicates that the 8 GB is not enough. It's a small sample size though.
If you are willing to turn down settings (e.g. 1080p medium), the 8 GB card copes better. Performance is being left on the table at higher resolutions, but again, Indiana Jones is dragging it down. For example, 1440p raster average is only 11% higher (however, lows are 21% higher).