Why are a lot of pc benchmarks in Ultra settings

[Wisman]

I see plenty of benchmarks where with low-mid range hardware where if it doesn't achieve 60 fps Ultra 1080p it is dismissed as a bad GPU. The worst part is usually a combination of medium and high is enough to push 1080p 60fps on even older cards and still have it look just about as good as it would in ultra. I think most modern games on low-mid range cards should be on medium or high settings.

[Jaunitta 🌸]

Agree its its a bit over rated.
But there are players who own 4K monitors who need this quality to enjoy the game.

[Snow]

I do absolutely agree with you. Often Ultra over Very High or High have little to no visual difference at all, but require way better graphics card. But those are benchmarks, if everyone tests graphics cards at some settings he prefer - it might be hard to compare cards from different reviews.

[Bad 💀 Motha]

Because that's what matters. If it can do ultra well it can do lesser settings at higher fps.

You can often look at sites like nvidia or others where they go in depth and shows fps per gpu, per various game visual settings.

[AbedsBrother]

Ultra settings benchmarks & gameplay make the gpu the limiting factor during testing. Settings that are too low run the risk of allowing other factors to impair performance, which wouldn't provide a good idea of how a gpu performs.

[Monk]

Originally posted by AbedsBrother:

Ultra settings benchmarks & gameplay make the gpu the limiting factor during testing. Settings that are too low run the risk of allowing other factors to impair performance, which wouldn't provide a good idea of how a gpu performs.

This is precisely it.

When the goal is to test hardware you want to use easily repeatable settings that will fully utilise said hardware (its the same reason why cpu's are tested at 1080p with with the most powerful cards, so the cpu is then the limiting factor), which is why wacking everything to maximum is the best approach and will give a clearer guide as to the performance between various gpu's.

[Tiberius]

It's easier to standardize the test with all of the graphic option being on or at the highest setting

[Wichtelman]

a good benchmark shows you all settings and the difference next to eachother...

[Washell]

The failure isn't in the type of benchmark, the failure is in calling it a bad card. They could just label it a decent budget model for people who game on medium. Like most sites I visit do. I think you're frequenting some elitist, biased reviewer who couldn't recognize good journalism if it hit him on the rear.

Also you need to use the same benchmark across the board, otherwise you have no data to compare the different models with.

[_I_]

Originally posted by 123:

a good benchmark *results* show you all settings and the difference next to eachother...

if you do not use the same settings on each bench test, its results will be skewed and offer no useful information

[Wichtelman]

Originally posted by _I_:

Originally posted by 123:

a good benchmark *results* show you all settings and the difference next to eachother...

if you do not use the same settings on each bench test, its results will be skewed and offer no useful information

just as example

https://www.youtube.com/watch?v=zvAFSJgmMBU

[_I_]

thats not good for comparing across different builds

but good for testing to see what settings would be most suitable for that build

[onlyHuman]

I use high-end with lowest settings.

[Bad 💀 Motha]

Just like when we here are sharing Unigine Heaven benches, we ask everyone run it on a certain preset, rather then picking various options, as when picking a preset this chooses the settings and screen res and thuswe can assure that scores make sense based on hardware if everyone uses the same preset for the test.

[Red™]

Most games for the past few years, you can barely tell the difference between high and low setting, since the game are so detailed. Sure, there is some element like grass or reflection will be weird looking in low setting, but other than that, it just look fine.

[SeriousCCIE]

oh based on what I've seen on other forums is that based on the makeup of the crowd on those sites, it depends on what you want out of the game and what sort of ego and epeen envy one wants to either display or cause to others.

Those people are cavorting with clowns, though. I think that if you don't fall into the all-ultra-all-the-time tripe, then a strategy I follow is... If you buy something in the top 3 rated cards in a given 'hardware cycle', they typically will last you 3 to 5 years provided they are coupled with a system that also can last that long.

If you disregard what many people suggest and don't get an i3 or i5 processor, and instead get an i7 or i9 or any of the threadripper and other higher end consumer chips--you're stand a much better chance (using absolutely no specfic part numbers as a basis for reference in my example here) to have a system that can run everything really well, and *keep doing so* even when you have to replace th GPU later.

If you get an emachine with some hot rod colored video card with flame decals on it, you may not have the same experiences, and end up posting somewhere later that the fps isn't so great despite having a "good computer." Yes, it's good for the price, but perhaps not good for the game settings chosen.

Anyway, the heaven and other benchmarks are really good ways to determine what you have, what other people have (what's common) and then for your own purposes, custom settings to see what it is that you have and how far it can take you with the settings you want.

So, mr. wisman -- I think you're right. One has to approach a card with the actual desires one has. If you dismiss a toyota prius as being a bad car because it isn't as fast off-the-line as the corvette it is compared against, you'd be considered a bad reviewer that isn't taking into account the various differences in the cars and focusing on the strong points of the highest end card. It is somewhat worse when you consider that many review sites receive hardware for free in exchange for positive reviews.

If you find a site that buys their own stuff, they won't make mistakes like what you cite--they will tailor the review for the audiences that would be actually buying in the price range of that card, and further, would provide a comparison of higher and lesser priced cards as a means of comparison, and not as a means of belitting the person that can't run everything on ultra.