I don't remember which specific videos.
Value of Hardware Unboxed benchmarking
- Thread starter Phantom88
- Start date
Any reviewer that’s trying the “objective” approach should show their test scenes so people can compare to their own pc, or replicate data in other reviews.
As far as I see it HUB has biases and also knows the biases of their audience, so they cater to that, which is totally fine. Digital Foundry is biased towards ray tracing which is fine. HUB is biased against. Pick your poison. Not every reviewer has to be everything to everyone.
As far as I see it HUB has biases and also knows the biases of their audience, so they cater to that, which is totally fine. Digital Foundry is biased towards ray tracing which is fine. HUB is biased against. Pick your poison. Not every reviewer has to be everything to everyone.
It's sad how commonly it's just running built in benchmarks, rather than actual gameplay
You should keep in mind the amount of AMD cards they have recommended over their Nvidia counterparts during the last 10 years can probably be counted on the fingers of 1 hand.
Why? I don't care. People know the parameters of what they're testing, and they can look at the data and decide if it's a source they want to incorporate into their decision making or not.
Because accusing an outlet that rarely recommends the purchasing of AMD GPUs as being biased is odd.
Wtf are you talking about? They have biases against ray tracing. They'll even pretty much say as much. They do polls of their fans and their fans do not really care about ray tracing either. So they don't focus on it. I'm not accusing them of anything malicious. Digital Foundry biases their reviews towards ray tracing. Again, not malicious. They just have more interest in what they see as forward looking technology. Both are absolutely fine.
The premise behind this thread is that HUB is biased towards AMD and possibly even puts out fake/altered benchmarks to make Nvidia look worse.
D
Deleted member 2197
Guest
AFAIK I don't think I read anywhere the benchmarks were fake or altered, but results that produced a watered down analysis that catered to his subscriber base when compared to other reviews.
HUB definitely slant (sometimes heavily) towards AMD IMO... however I absolutely do not believe that they produce unreliable data. I believe that when it comes to the numbers they present, they are factual and objectively what they observe during their testing. I also believe their methodologies are sound and believe they have integrity. I don't always agree with them, however there is definitely as much value in what they do as any other source.
I believe that they do what most other sites do which have inherent biases... and that is present the information in the best light possible for the side they lean towards when possible. And what I mean by that is going out of their way to rationalize why X might not be so bad compared to Y... if you take Z into account as well.
I watch both AMD leaning reviewers and sites, as well as Nvidia leaning reviewers and sites... despite my own preference being Nvidia. It's important not to limit yourself to just reviewers and sites you agree with, or that you know will hold the same opinion as you do because you're less likely to allow yourself to accept that others can be right to have a different viewpoint and that yours can also be wrong.
I believe that they do what most other sites do which have inherent biases... and that is present the information in the best light possible for the side they lean towards when possible. And what I mean by that is going out of their way to rationalize why X might not be so bad compared to Y... if you take Z into account as well.
I watch both AMD leaning reviewers and sites, as well as Nvidia leaning reviewers and sites... despite my own preference being Nvidia. It's important not to limit yourself to just reviewers and sites you agree with, or that you know will hold the same opinion as you do because you're less likely to allow yourself to accept that others can be right to have a different viewpoint and that yours can also be wrong.
Agreed. Trust no one reviewer. It’s not that hard to make an informed decision based on a plethora of reviews. If I’m buying a $800-$2000 GPU, I’m using a ton of sources to make sure I’m doing my due diligence.
HUB is one of a few sites that I visit for GPU reviews. TPU and DF are the others. My main problem with HUB is that their raytracing coverage seems disingenuous. It’s perfectly valid to live in the present and say that there aren’t enough RT games in 2018 during Turing launch. It’s also perfectly valid to say that the performance cost of RT is still very high in 2023. Their agenda is laid bare though when they choose a handful of RT enabled titles to include in a review and they choose some of the absolute worst RT implementations when they do so.
My take on HUB is that they’re hardware enthusiasts but they aren’t graphics enthusiasts.
My take on HUB is that they’re hardware enthusiasts but they aren’t graphics enthusiasts.
Or worse, choose a number of games that support RT but test them with RT disabled on $1000+ GPU's that are more than capable of playing them with RT enabled.
An interesting perspective. It’s ironic? That some reviewer sites go one way while another goes the opposite.
In one instance, one sees hardware as to chase diminishing returns in frame rate and resolution. In the other one, one sees ray tracing as chasing diminishing returns in graphical fidelity.
But imo, it’s pretty clear differential between the two. Ray tracing will get you closer to the promise land than traditional t&l will.
But if what you enjoy is not the graphics but the potential that hardware has, sort of like saying, I like having tons of horsepower but not care how fast it goes around the track, then I sort of see why some would go that way. The only stats that matter to them are the 0-60 and the quarter mile and it’s the drivers fault if they cannot win the race. Whereas everyone is looking at prepping their car for the track and it’s everything together to get across the line.
D
Deleted member 2197
Guest
Their method of testing might have merit if they only test one IHV and draw a conclusion for that one IHV. But holding back strengths of another IHV and making a comprehensive conclusion for all cards is far from maintaining a neutral perspective, especially if it impacts the final score card.
And that’s a valid opinion since it’s pretty subjective and depends on which games you play. For every game that uses RT well there are 2-3 others where the benefits aren’t obvious. As a somewhat respectable reviewer though if you want to take a strong stance that RT isn’t worth your readers’ time or money the least you can do is make a good faith effort to evaluate the tech. That’s the missing ingredient.
It doesn’t help that there is a very high correlation between IHV preference and RT enthusiasm. Why are Nvidia fans more excited about RT than AMD fans? It makes no sense when you look at it from a pure graphics perspective. The same thing happened with async compute when AMD was ahead. Yet everyone today agrees async compute is a very good thing. One day everyone will agree RT is awesome too. We just have to deal with a few years of nonsense first.
Most of these things work themselves out in the end. I’m still sore about physics hardware acceleration though. People threw tons of hate at Nvidia for their heavy handed proprietary PhysX shenanigans. Well eventually they got their wish and hardware PhysX is now dead. Unfortunately it didn’t work out for us either since physics in games haven’t really improved since. We all win if we all lose right?
HUB isn't anti RT. It’s not that cut and dry. There have been implementations they approve of such as Metro, Cyberpunk and others. Steve is a gamer first and graphics enthusiast second. Online shooters that are highly competitive are his preferred gaming experience. It comes down to the visual return relative to performance hit. RT to him becomes something worthwhile at the 3080 performance level. In every review he does he reiterates Nvidia’s RT advantage and notes DLSS 2 is a huge feature and something AMD just doesn't have an answer for yet. If you watch their content they demonstrate a similar RT performance gap as other reviews and have included every noteworthy RT title outside of the Nvidia funded path traced games. Do you guys think if it was AMD that was faster in RT that his views would change? I don’t think the history of his content supports that as being likely.
Last edited:
I think HUB's reviews are generally thorough and almost always provide at least some interesting data points. Tim's stuff on monitors stands out in particular. Steve's reviews on GPUs are pretty good too, all things considered, but he's upfront about his primary gaming interest being multiplayer twitch shooters.
As such, eye candy is not as high on his list of priorities as it might be for others (like myself). I recall he's said more than once that ray tracing performance is about a 10% value add to him. So yeah, different strokes and all that, and most people probably want to take in some opinions from a variety of sources.
As such, eye candy is not as high on his list of priorities as it might be for others (like myself). I recall he's said more than once that ray tracing performance is about a 10% value add to him. So yeah, different strokes and all that, and most people probably want to take in some opinions from a variety of sources.
Maybe his preferred class is sniper
Similar threads
