Techspot, 7900XTX vs 4080 Revisited

256 bits limitation seem to show up at 4k, quite impressive for a 380mm-256 bit card to keep up that much with 529mm of silicon with a 384 bit bus...
 
Nice! I hope this prompts Nvidia to release that driver they’ve been saving to boost the 40xx series.
 
4saken said:
and my point stands.
Click to expand...

Only if you intentionally look at with green colored glasses. Both cards have their strengths and weaknesses. It’s not like the 4080 doesn’t take a massive hit in performance when RT is enabled as well. Both cards end up just fine with native RT rendering at 1080p and both end up suffering at 4K. Both cards need super sampling to maintain good performance at RT 4K. Outside of a couple of outliers, both end up playable with scaling. And even with those outliers the average difference only rises to 11% when using both RT and super sampling above 1080. Everything else is at a point where the average difference is either negligible or within margin of error.

With a larger sampling of games I wouldn’t be surprised if things end up being even closer. Both cards come out of this test looking good, outside of their prices, there’s really need reason to suck the dick of one brand over the other.
 
Not much has really changed since release. You really only consider the price of the 4080 being worth it if you play a lot of 2077 or DL2.
 
It's nice to see a re-review down the line that takes updated drivers into account, but as usual, no VR re-benchmarks, which is one of the big reasons I traded my RX 7900 XTX for an RTX 4080 to begin with. (The other reason being that damn vapor chamber defect at launch.)

In DCS and NMS back in December and January, it wasn't even close - the 4080 mopped the floor with the 7900 XTX in VR, well worth the trade. New drivers allegedly rectified the VR performance, but I'm not about to go drop serious dosh on a new card just to sate my curiosity.

Further complicating things is that DCS got a long-awaited multi-threaded CPU update a few months back, which may or may not improve frame times depending on how well it's implemented.

With all that aside, if AMD had this kind of performance (and no underfilled vapor chambers) at launch, RDNA 3 would've been off to a much better start, but they fumbled at a very bad time considering NVIDIA's utter dominance and unchecked greed right now. Not everyone's going to check follow-up reviews and see how much the gap has been closed and then some post-launch.
 
You must log in or register to reply here.
Back
Top