Comparing Graphics Cards

Hi All, I’m wondering if anyone has stumbled upon a reliable, independent and objective site or forum where graphics cards are evaluated, vetted and/or discussed?

This would be for a mobile workstation (laptop) for video editing.

I read specs for mobile workstations from Lenovo, Dell and HP and various models of the Nvidea card are listed. I’m trying to ascertain what more do you get when you upgrade from a Quadro T1000 to an RTX 2000 to an RTX 3000 to an RTX 5000? More cores? More threads? Is there a certain minimal core and/or thread count to render a video in a certain time (e.g. 30 minutes, 1 hour?)

It’s my understanding (perhaps mistakenly) that the Geofroce is fine for regular computer applications (e.g. Word, Libre Writer, Excel, Calc, Power Point, etc.) but for video editing (especially for rendering) you really need at least the Nvidea Quadro but which model?

Thank you!

Critical question… do you intend to edit with programs other than Shotcut? Which ones if so? There are lots of variables and nuances here.

I will also be editing with Da Vinci Resolve (free) required by work. Thank you.

There are a lot of things to consider here one I typically use as refrence material is going to be the nvenc/nvdec matrix

This highlights some of the encode/decode limits on certain cards such as the T1000 being limited to 3 encodes, or the 1650 not having HEVC B frame support(something the T1000/T2000 shouldn’t either actually since TU117 uses the older volta block but it looks like they made a mistake when they added the 3000 series cards and reformated the matrix)

Beyond that the largest difference for most things is going to be raw gpu power and ram, the 3000/4000/5000 are all going to have more of both but will require more power(thicker laptop/bigger powerbrick) depending on the complexities of your projects now or in the future there’s likely going to be a “big enough” without going overkill, once you’re using nvenc/nvdec it’s really on the cpu to keep the gpu fed since that block is typically the same on a family of gpu’s with a few exceptions(the RTX 4000 actually has multiple decode blocks something the larger 6000 doesn’t)

This topic was automatically closed after 90 days. New replies are no longer allowed.