W1zzardThe mobile apps haven't been updated for a long time, and are somewhat buggy atm. Not sure what to do with them. The site works perfectly in your mobile browser, so why duplicate the effort/maintenance/etc?

I use the app for just the forums, was unaware it was even broken elsewhere, lol.

trom89just mark the ones that are not confirmed as "LEAK", "WRITER OPINION", "RUMOR" and inside the article post in big red font "This information is NOT confirmed by the company bla bla bla."

I like the idea. "Opinion" is covered by "Editorial" already. "Leak"/"Rumor" are effectively the same. I guess "Rumor" could be best option?

How about an option to support for $1 monthly (or any custom amount) just for the sake of supporting and not having any extra benefits?

RahnakYou don't need anyone to tell you the 2080 Ti is on a very different level from the RX 570, no. But what if you're torn between buying the 1660 Super or the 1660 Ti? Do you know how big the gap is and if it's work the extra money to you? Or between the 1660 Ti and 5600XT? That's what these reviews aim to tell you, so you can get the best bang for your buck.
Figuring out how a particular GPU will perform on your system will then require some extrapolation. Is it perfect? No. Does it work? Yes.

I don't really know if the performance gap is useful if I don't even know if I'll be able to even run the game properly... or at all like some charts point out.

CheeseballThat is the point of the Benchmark Test & Performance Analysis articles. To see how good current video cards would work at the highest settings that is playable by the most powerful card. From the information you gather there, you can already conclude yourself that some of the settings in a game is not worth it. You can also conclude if the card you're interested is good compared to others on that list.

If you want real world usage, it would be best for you to experiment with the settings in the game itself. These are not purely game reviews.

So they are, indeed, useless.

bogamiAnd so we lost the internet! everyone makes membership fees, including antivirus upgrades, every year! Abuse of the position is nothing new !!! The pigs are multiplying, and it will be bursting somewhere ...

Why should everything on the internet be free? Back in the day, if you wanted this sort of content, you had to buy a PC magazine. The medium changed but there are still real people making the content that need to be paid. Luckily there are other ways for them to get revenue so you/we don't have to pay. But that's not always enough or not always the best way.

W1zzardThe mobile apps haven't been updated for a long time, and are somewhat buggy atm. Not sure what to do with them. The site works perfectly in your mobile browser, so why duplicate the effort/maintenance/etc?

If they don't have any extra functionality over the site, imo you should drop 'em.

ImoutoI don't really know if the performance gap is useful if I don't even know if I'll be able to even run the game properly... or at all like some charts point out.

How so? You have the 3 more common resolutions. Pick the one you use. If you like to play at High, you know the fps is going to be higher. If your CPU is older/slower, subtract a few.

FreedomEclipseI think there's some fundamental things that need to be addressed as far as news reporting before one decides to support the site.

Im talking about clickbait and 'sensationalist' articles and of course who could forget the hit peice on GamerNexus earlier in the year...

Do you know how badly that knocked TPUs reputation?

It's not the first time that this certain individual has been called out about his writing. He did it with an MSI article when there was a rumor floating around that they were locking off older boards from being compatible with the newer ryzen series. These are just two examples off the top of my head but I'm sure there are a few more....

TPU has got to evolve and bring the standards of writing back up. Some reviews occasionally feel pretty lazy and uninspiring. I understand that often the reviewers might not have have all the time in the world to pour into the reviews but sometimes going above and beyond that really puts the quality out there.

There's a lot TPU could learn from GN rather than write articles accusing people of being paid shills.

GN hasn't struggled for backers on patreon the same way that TPU has since they started up their patreon page.

Make TPU Great Again

Steve is a shill at times and is a rabid fanboy at times.

He does sensationalist articles/videos.
He does click bait articles/videos.
He doesn't disclose where and when he gets 'gifts' or massive 'discounts' by vendors/suppliers/etc.
He will beat around the bush, saying he 'aquired' and the item is 'worth'. If he actually buys it he says so.
He is self interested and is a massive divo.
He takes cash and items directly from manufacturers and vendors.

He may not have a direct link to Nvidia as he raged about in the anon outing response. Where he basically shamed and refused to accept the apology from TPU.

I think his 'reporting' is blogger level crap. He acts like he discovered the 2060 KO having a failed 2070 die, but rushes out how it's got a 40+% uplift in workstation performance. Then he stalls on the video about how NV and EVGA neutered it's overclocking, how it uses more power, and runs hotter... TPU had their article up with those details but other Youboob channels only mentioned GN because...

He hammered the Radeon VII because it was meh at gaming, he couldn't complain enough about that disgusting thermal pad.

Fine he's gaming focused... Oh wait what? Only CUDA is good enough to get rendering and compute benchmarks? Why he scared that his overpriced Quadro RTX 5000 getting destroyed in video rendering would look bad... Radeon VII is twice as fast as a 2080 Ti in Blender last I looked... Now that Premiere uses OpenCL I bet it would really tear his precious a new one.

Steve isn't innocent, he's just opaque and he only said he didn't get sponsored by Nvidia directly... He was so upset because it hit so close.

So I hope TPU doesn't go down that road. It would be utterly gutting if they started to be influenced by the money.

I will support TPU as best as I can when I can afford to. I wish there was a little $2 tier.

Edit: I want to clarify, I have questioned GNs supposed lack of outside influence for quite some time. I'm sorry but not just any Joe Blow gets behind the scenes tours of manufacturing plants, inside looks at how board vendors design their custom boards, direct access to high ranking folk in those companies.

Take his very boldly planted Thermal Grizzly sponsored materials. Does he disclose every video, no... Then he uses a tube of Hydronaut, but refers only to it as thermal paste when he says in a video shortly after that it turned to rock in the tube in a video with thermal results... Why not use Kryonaut? Is it because like Buildzoid mentioned he and others found that it seems to scratch dies. Nope just silently he changes paste...

He definitely doesn't bite the hand that pays his bills. I wouldn't hate so much but I hate how righteously he acts when he's very much like a politician saying those donations aren't bribes... They might not be directly but humans are susceptible to influence and that stuff influences his opinion.

In my honest non-professional opinion. I get it won't be popular to view PC Jesus in such a skeptical light.

I honestly want to Congratulate Steve on supporting such awesome causes. I tried to bring myself to buying something from his store, but I couldn't. So I directly donated to the causes. But no $20 signatures for me... Heh

Edit 2:

Please set the minimum tier at $2 month, because I think I can afford to swing that right now. It also lets the site get a real donation.

arczi19How about an option to support for $1 monthly (or any custom amount) just for the sake of supporting and not having any extra benefits?

I think that's already possible at Patreon, give it a try

This looks like it is starting up again. Unless I missed it (and I certainly might have) do those of us who are already Patreon supporters of TPU need to sign up again, or does it just carry over?

RahnakHow so? You have the 3 more common resolutions. Pick the one you use. If you like to play at High, you know the fps is going to be higher. If your CPU is older/slower, subtract a few.

That is what I'm getting at. Why present useless data and leave the information relevant to the end user up to their imagination when you can do much better?

I'm sorry but I can't even fathom how this is useful for anything but fueling the war between loyalists.

rtwjunkieor does it just carry over?

Check your inbox from Patreon, I sent out messages to existing subs before the news post announcement. I ported your permissions to the new tier levels already, but you might want to cancel and resub at the new lower pricing

W1zzardI think that's already possible at Patreon, give it a try

After a membership level is selected, there is a field where one can see the pledge, and edit it. However, it keeps reverting back to the pre-determined amount, after changing it to 1.

NRANMAfter a membership level is selected, there is a field where one can see the pledge, and edit it. However, it keeps reverting back to the pre-determined amount, after changing it to 1.

www.patreon.com/join/TechPowerUp?
Does that link have a "make a custom pledge" button below the 3 pledges?

W1zzardwww.patreon.com/join/TechPowerUp?
Does that link have a "make a custom pledge" button below the 3 pledges?

It does. It wasn't there before, I think.

NRANMIt does. I wasn't there before, I think.

Yeah, note the ? at the end of the URL. Not sure why they're doing it like that. Probably they want to hide the reduced cost option by default

@arczi19 this should work for you, too

ImoutoI don't really know if the performance gap is useful if I don't even know if I'll be able to even run the game properly... or at all like some charts point out.

Your argument is quite flawed here. If a card runs a game at 23 FPS at 4K on maximum settings, you can already conclude that it would probably run a lot better with a lower resolution and settings. The reviews already did the job with the currently common resolutions here.

There is no need to run game benchmarks at medium or low settings when you can already determine what will happen at max (or near max).

ImoutoI don't really know if the performance gap is useful if I don't even know if I'll be able to even run the game properly... or at all like some charts point out.

So they are, indeed, useless.

ImoutoThat is what I'm getting at. Why present useless data and leave the information relevant to the end user up to their imagination when you can do much better?

I'm sorry but I can't even fathom how this is useful for anything but fueling the war between loyalists.

If you don't know what to do with the data, then that's up to you.

Data like that (especially the 1080p and 1440p charts) is helpful in determining which card is worth upgrading to. If I'm upgrading to a 2080 Ti (I'm not since I have a 2080 Super), I would like to know how it runs at maximum settings. If it can't handle it, then I would simply lower the settings.

ImoutoThat is what I'm getting at. Why present useless data and leave the information relevant to the end user up to their imagination when you can do much better?

I'm sorry but I can't even fathom how this is useful for anything but fueling the war between loyalists.

Open up a word doc and type out the configurations you'd use to test out all current gpus available. You'll see why what you propose is unfeasible.

NRANMIn my opinion paid reviews are always a bad idea if a website has a reputation for being trustworthy and objective. The moment I see a paid review (text, video, whatever), I immediately consider it worthless. I also don't trust any articles or videos that may be sponsored by Company X, even if they are not directly related to the subject matter, like for example "Let's build a gaming PC. This video is sponsored by be quiet!." And then the case, PSU, CPU cooler and case fans are all be quiet! ones. That isn't a "let us help you build a well balanced gaming PC (for a certain budget)" video; this is "let me advertise these components to you" video.

The way I see it, services like Patreon actually help reviewers who want to be objective to remain that way. When one has independent funding, one is much less reliant on the benevolence of hardware manufacturers. It is what allows Gamers Nexus to still review hardware if they are unhappy with how a manufacturer is handling NDAs or cannot take criticism and blacklist them, etc. "You don't want to provide a sample for review because you are butthurt over something? No problemo, we are going to purchase it with our own money, and do whatever we want."

GamersNexus works with Linus all the time, yet Linus does a lot of paid reviews and still gives honest feedback. I agree with you I don't want that to become TPU the more I think about it, but loads of people still trust and watch Linus, and know that there are certain reviews he does to take with a grain of salt, but he still gives honest brutal opinion a lot of times while still raking in the money. One of the reasons Linus could even talk about retiring in his early 30s recently is because of how rich he has become, can't lie, I envy that and would retire in a heartbeat if I had as much money in savings as I assume he does. yolo

TheLostSwedeI guess you've never 1. worked as a tech journalist and 2. never had your own business?

Both things are true and I admit my view my be a bit too narrow, but that's why I added that I hope that the Patreon subscribers could help with the opportunity to write such reviews and, in my opinion, draw more readers as I'd say a review of a Seasonic PSU or Lian Li case will bring more readers than a review of the umpteenth mouse Thermaltake has released. May I ask where your supporter badge is?

W1zzardI like the idea. "Opinion" is covered by "Editorial" already. "Leak"/"Rumor" are effectively the same. I guess "Rumor" could be best option?

change editorial to Opinion or Satire, these words are more recognized by the majority of the people, and definitely add Rumor.

RahnakOpen up a word doc and type out the configurations you'd use to test out all current gpus available. You'll see why what you propose is unfeasible.

This is a mock up.

New game into the lineup:

- Check settings until the worst card does 60 FPS at 1080p. - Call it "Budget".
- Check settings until the worst ~$300 card does 60 FPS at 1440p. - Call it "Performance"
- Check settings until the worst ~$400 card does 60 FPS at 4K. - Call it "Premium"

- Get all the cards from the worst to ~$300 and test them with "Budget" settings. Congrats, now you have real world data.
- Get all the cards from ~$300 to ~$400 and test them with "Performance" settings. Congrats, now you have real world data.
- Get all the cards from ~$400 onwards and test them with "Premium" settings. Congrats, now you have real world data.

- If you want to go the extra mile get one or two cards that are borderline into another segment and test them with those settings. Use those reference points to link the three segments.

1234 | 4567 | 789

win / win / win

- You end up doing 4 more tests.
- If a card can't keep up with a resolution no matter the settings be honest about it and drop it from the chart.
- If the game isn't demanding at all feel free to drop one segment.
- Repeat for every manufacturer.
- Revisit the methodology once a year or six months if you are bored.

Again, only a first draft.

This is honestly the first time I've heard of buying a product based on awful performance numbers and crossing your fingers so it performs adequately when it gets home.

ImoutoI don't really know if the performance gap is useful if I don't even know if I'll be able to even run the game properly... or at all like some charts point out.

in the end, you really just don't know...

ImoutoThis is a mock up.

New game into the lineup:

- Check settings until the worst card does 60 FPS at 1080p. - Call it "Budget".
- Check settings until the worst ~$300 card does 60 FPS at 1440p. - Call it "Performance"
- Check settings until the worst ~$400 card does 60 FPS at 4K. - Call it "Premium"

- Get all the cards from the worst to ~$300 and test them with "Budget" settings. Congrats, now you have real world data.
- Get all the cards from ~$300 to ~$400 and test them with "Performance" settings. Congrats, now you have real world data.
- Get all the cards from ~$400 onwards and test them with "Premium" settings. Congrats, now you have real world data.

Yikes. No. Just...No. wow.

My dude, every PC player's goal (money notwithstanding) is to run ultra settings in games. Sadly, this isnt possible for many users, again due to money. But there are so many games and so many settings and resolutions, what is real to you, isnt real to the next person. Testing using ultra settings is designed to show a relative difference between the cards. The goal isnt to show you what fps you will get in game. Even using an ingame benchmark doesnt really show ingame fps. The point is to show relative performance bud. Everyone is going to set something different and you cant please everyone. If you see "awful" performance numbers at a given res in a given title, guess what, that card isnt good enough. The point is to run at the highest settings where possible, turning things down only when needed. God only knows what you, I, he, and she turn down to reach 60 fps.

I think this would be a great discussion to carry on in a thread of it's own.....

So I thought I'd add my 0.02$ as it were, if that is the correct phrase.

I decided to support TechPowerUp in Patreon very recently. I see a lot of concerns from other users in this thread and I understand some of them, but I will remain neutral on many of them.

I don't really know how to word this well but I have some bullet points from things I noticed in the thread here:

• TPU's graphics card reviews offer an excellent view of 'overall performance' of a graphics card's performance. The wide game suite gives us a better look at the product's performance from a top-level, more so than other publications who test (even if more detailed with % lows and stuff as mentioned earlier), because they test A LOT fewer games. I have been quoted as stating before now "2 or 3 benchmarks isn't indicative of a GPU's performance, you should take an average of maybe a dozen games that are popular and graphically demanding". And I stand by that. We have to remember W1zzard is "only" human too, lol. I absolutely love the GPU perf review "Summary" option of all of the games. If I want a more detailed look at 1 or 2 games some other pubs or YouTube has that covered. I see TPU's charts linked EVERYWHERE for overall product positioning and performance. Please don't change this! Just generally, I love the reviews detail & content about graphics architecture, and I the author's opinions and I feel the reviews are 100% solid and I'm willing to pay Patreon to support these.

• Ads: I never actually had an issue with the adverts on TPU, as the banners were almost always PC related and relevant. As for the "grey market" key ads, I won't lie, I don't feel they are appropriate for TPU but if the revenue is there then I can tolerate them.

• I won't comment on particular authors of articles as I haven't noticed anything that I immediately think is not OK (But please keep in mind I am highly autistic, I mainly read GPU reviews or news concerning GPU / CPU / processors and such).

I might think of some more things to add, but the way I feel is that no one does graphics cards (that is the bread and butter of TPU for me) reviews like TPU, and also making GPU-z, hosting some other great tools (like the custom Ryzen power plan!) & excellent community on the forums (bless you guys/gals <3) and generally I felt it was more than worth the Patreon support every month. I pay about the same for an Adobe Rush sub, and I'd argue I spend more time reading TPU than using Rush (this is a fact) so why not.

Going forward I would like to see TPU thrive and keep up the comprehensive & broad GPU reviews that give the (IMHO) the industry's best insights into a Graphics Card's "top-level" performance. Maybe you could hire another editor to do a more detailed look at a smaller sub-selection of games? This would allow you to cater to that, also. By the way, I enjoy reading Anandtech, too.

Maybe for certain games you could include the minimum/lows % FPS. For the games that flirt with 60 FPS this might be an idea but overall I like what you are doing @W1zzard keep it up!

Thanks for reading my long post :3

W1zzardSure, if I bring down the number of games to 5, at 2 resolutions, with 3 other cards compared.. and we'll be like everybody else.
27 comparison cards x 21 games x 4 resolutions = 2268 test runs currently

This. I'm personally tired of people screaming about .1% lows.
1 - I come to TPU to see a broad spectrum of games. Other then maybe PCgamer, I don't know of any site that uses as many games in their reviews and off the top of my head I think TPU still uses more.
2 - If I want .1% then I know which sites will show it. I also know which specific site will run it into the ground as the end all be all in order to justify those 18 cores you must have for your CPU.

I support TPU because I would like to support someone that push hard and keep cadence on what they love for decades - which I couldn't do myself now lol. Keep it up on the field you're keen on. No need to show up faces and do eyecandy RGB videos. Let me know if you need any localisation assistance from Thai sources.

Imouto- Check settings until the worst card does 60 FPS at 1080p. - Call it "Budget".
- Check settings until the worst ~$300 card does 60 FPS at 1440p. - Call it "Performance"
- Check settings until the worst ~$400 card does 60 FPS at 4K. - Call it "Premium"

- Get all the cards from the worst to ~$300 and test them with "Budget" settings. Congrats, now you have real world data.
- Get all the cards from ~$300 to ~$400 and test them with "Performance" settings. Congrats, now you have real world data.
- Get all the cards from ~$400 onwards and test them with "Premium" settings. Congrats, now you have real world data.

Well, how far do you lower the settings? Personally I don't like to go lower than high. Person X might be OK with normal. And if you're calling it real world data.. you can't pair the the GPUs with the same CPU as that's no longer realistic. You're introducing a lot of variables into testing that make the process a lot harder, more time consuming and more prone to errors. And how confusing is it going to be when people see cards in such different price points having similar fps numbers?

ImoutoThis is honestly the first time I've heard of buying a product based on awful performance numbers and crossing your fingers so it performs adequately when it gets home.

That's not how it works. The reviews show you performance at Ultra and relative performance compared to other GPUs. In simple terms, say I'm looking for a 1080p card, I'm going to go with the one that offers the most performance within my budget. It's that simple, buy the best you can afford. And without relative performance you don't know which one that is.