We keep getting requests from Quora and occasional friends asking, what is the best AIB for____, or can AMD XXX do better than Nvidia YYY? Mostly there are lazy questions that could be asked by the questioner with a little effort and web searching. We never answer such questions, although there seem to be plenty of people within Quora who are willing to assert their authority and opinion on such matters. It reminds us of the brain-dead questions in Quora that start out as Can Trump get reelected. . .
We’ve tested a lot of AIBs over the years, and about 10 years ago we expanded the testing to include price and power consumption. Occasionally we also include loudness in our evaluation when we encounter a particularly obnoxious AIB.
This last week we collected a handful, a big handful, of AIBs and ran UL’s Timespy on them to get a relative merit of comparison. As you might have expected, Nvidia’s RTX2080 Ti was the unquestioned winner, and AMD’s RX580 came in last.
|Performance-based on Timespy of various AIBs|
If raw performance is all you care about, then you can stop reading now, and just look at the above chart.
But that’s not necessarily the whole story. If you compare price (we used the price the AIB was introduced at, not today’s price which may be even lower)you get a different winner, the popular GTX 1660 is the clear winner and the Radeon VII came in last.
|Price-performance of several graphics AIBs|
If you are ecologically concerned, and what to minimize power consumption because there’s enough carbon in the atmosphere already, then you should look at the Pmark. That analysis shows once again the Nvidia GTX 1660 as the best choice, and again the AMD Radeon VII in last place.
|Pmark comparison of popular graphics AIBs|
This suite of tests and analysis will hopefully offer a choice to the readers. The best doesn’t have to be the only choice. And if you already have one these AIBs you see how it stacks up (from our testing and analysis) to others.
As in all things in life, there’s no single right answer. It’s all a gradient and we fit on the curve at various places. What’s in your PC?
Scores vs. frame rate
The score may not give a user the information he or she is seeing, so we did the same analysis, using Timespy’s frame rate GT1 score (fps), and not surprisingly, the results were the same.
|FPS Pmark results for various graphics AIBs|
The conclusion is UL’s Timespy is a reliable and consistent test, and by comparing raw scores, price-performance scores, and Pmark scores, a user has a pretty good armament for making an informed decision.