GPU award criteria

[Selaya]

Hello,

currently any and all GPUs in a certain series (ex: 4080 and 4090) are awarded the exact same awards (in these cases: the Editor's Choice), irrespective of the AIB model in question's merit.

My suggestion: Change how awards are awarded for GPUs. Each individual AIB card should receive its own award, without inheriting any from the series itself. Those should be awarded to the reference design; individual AIB cards should be judged on their own merits and receive their independent awards, if any at all.

 

[W1zzard]

They are, these are independent awards, the cards are just extremely similar

Gigabyte 4070 Ti is only Recommended

 

[Selaya]

I see.
In that case, I'd like to suggest giving the Editor's Choice for the reference design only and be more uh conservative w/ it otherwise; it makes no sense to give an otherwise poor AIB model the same Editor's Choice when they're clearly underwhelming compared to the reference design (either through sheer ineptitude or, more commonly bad pricing) where you'd be far better off purchasing the reference model instead.

At the very least that's just my (general) feeling looking at some of these reviews

 

[cvaldes]

One important note: reference design cards are frequently difficult to source and nearly impossible in some markets. Even in the USA, the Founders Edition of the Ampere cards (30 Series) were almost impossible to buy from the approved retail channels (mostly Best Buy).

It makes sense for the Editor's Choice award to be given to worthy AIB partner models as an alternative to the nearly mythical reference models especially for TPU readers who live in areas where the latter are not marketed.

 

[Selaya]

oh, im not suggesting to deny AIBs said award altogehter, i am merely suggesting a higher threshold for it than currently (basically, it'll have to be an epicfail of an AIB design to not receive said award rn) - instead of like, >90% of all 4080s automatically receiving the Editor's Choice bc the 4080's a good gpu (ignoring pricing yes, yes) rather maybe make at least half of them receive it

basically, if an aib's like $200 more expensive than the reference, on top of being louder and having a middling cooling solution - that's not editor's choice material

 

[W1zzard]

it makes no sense to give an otherwise poor AIB model the same Editor's Choice when they're clearly underwhelming compared to the reference design

Which card are you talking about?

basically, if an aib's like $200 more expensive than the reference, on top of being louder and having a middling cooling solution - that's not editor's choice material

Correct, that's why Gigabyte only got Recommended

 

[Selaya]

zotac 4080 (and colorful 4080 to some lesser extents - $50 extra for a cooler actually worse than FE)

 

[chrcoluk]

I would like to see awards been much harder to get. But yeah, I agree with the OP getting an award because the Nvidia part of the product shouldnt be what its about, AIB reviews are on the non chip quality side of the card. Cooling, noise, bios, software, build quality, power delivery etc. If every card gets the award, it devalues the award itself. Hopefully my feedback is considered helpful.

 

[pavle]

I would envision a "Good buy" award for the "similar cards" (in different parameters as @W1zzard describes above) and "Editor's choice" award only for one variant that really impresses with its price or performance. I do realise reviewing and having access to numerous good cards makes you feel you like a child in a candy store (you like them all).

 

[ThrashZone]

Hi,
Thought this was all about the awards deal

About TechPowerUp Awards

https://www.techpowerup.com/awards/ The plan is to put a link to this page after the award(s) listing on the conclusion page of all reviews. Thoughts? Feedback? Questions?

www.techpowerup.com

 

[W1zzard]

i am merely suggesting a higher threshold for it than currently (basically, it'll have to be an epicfail of an AIB design to not receive said award rn) - instead of like, >90% of all 4080s automatically receiving the Editor's Choice bc the 4080's a good gpu (ignoring pricing yes, yes)

Why the double standards? Wouldn't that mean it'll effectively be easier for FE to get an award? What if there is no FE? What if the AICs tell NVIDIA "plzplzplz dont send TPU a FE, or we won't be getting any awards"

I think you're looking at this from the perspective of someone who knows the product in and out and who has read a dozen reviews about it. A substantial (majority?) of our readership comes to use, possibly via Google, with limited technical information and knowledge and just wants to know "is this good?"

zotac 4080

The cooler is slightly better than FE, you get an OC out of the box, you get good RGB, the fan is a bit louder so it can handle the heat output from the OC, with similar temps as FE. None of those are enough imo to take away the Editor's Choice award, it's still a RTX 4080, which is one of the best cards ever made (perf, efficiency, rt, dlss 3) and you will be an extremely happy gamer with it, if you can afford it

only for one variant that really impresses with its price or performance

Not possible unless I build a time machine first, which I could if I could travel back to the past with my time machine and bring the plans

makes you feel you like a child in a candy store (you like them all)

While I love what I do I certainly don't feel that way. These cards are work items and have no further personal value, I can only have one card installed in my work PC, and I don't have time to play games anyway. I certainly don't like them all, but it's my job to be impartial and fair .. I think I'm doing ok on that front after having reviewed more than 500 graphics cards

 

[Selaya]

Why the double standards? Wouldn't that mean it'll effectively be easier for FE to get an award? What if there is no FE? What if the AICs tell NVIDIA "plzplzplz dont send TPU a FE, or we won't be getting any awards"
[ ... ]

Well, perhaps I should elaborate a bit more on this:
The FE would get an award (or not) purely on the basis of whether this series is well, good or not (short of obvious, critical failures in the FE's design, ig).
Alternatively, bifurcate the awarding process - one award for the series as a whole shared by all cards, and then another for individual standout designs

[ ... ]
I think you're looking at this from the perspective of someone who knows the product in and out and who has read a dozen reviews about it. A substantial (majority?) of our readership comes to use, possibly via Google, with limited technical information and knowledge and just wants to know "is this good?"
[ ... ]

True, maybe the latter suggestion above might be a better idea in that case

[ ... ]
The cooler is slightly better than FE, you get an OC out of the box, you get good RGB, the fan is a bit louder so it can handle the heat output from the OC, with similar temps as FE. None of those are enough imo to take away the Editor's Choice award, it's still a RTX 4080, which is one of the best cards ever made (perf, efficiency, rt, dlss 3) and you will be an extremely happy gamer with it, if you can afford it
[ ... ]

in a vacuum, yes obviously
the main reason that award kinda didn't sit right w/ me is the fact that none of these are worth the $200 markup - if it were MSRP or maybe $50 at most, sure - but $200? for an otherwise lackluster card? meh.

[ ...]
Not possible unless I build a time machine first, which I could if I could travel back to the past with my time machine and bring the plans
[ ... ]

You're reviewing multiple cards for the launch - I agree, the cards that you're not receiving/testing on launch cannot be factored in but - you usually receive a healthy amount of AIBs for a bigger launch (like, 5 at least) and release reviews for all of them simultaneously. Why not just use those as the reference group, Editor's Choice the best/most standout of them (dito w/ Great Value, Innovation should that be applicable), and then for later launches/samples, use that existing group as uh benchmark?

 

[ThrashZone]

Hi,
This is not the first time someone noticed the same editor choice award given out
Might be At one time it was on every product reviewed I haven't read every review so I'll leave that conclusion to others.

Last I read it's all about google search rankings.

Edited

 

[pavle]

While I love what I do I certainly don't feel that way. These cards are work items and have no further personal value, I can only have one card installed in my work PC, and I don't have time to play games anyway. I certainly don't like them all, but it's my job to be impartial and fair .. I think I'm doing ok on that front after having reviewed more than 500 graphics cards

You are doing well, I never really got the impression of bias or anything, the many editor's choice awards prompted that thinking is all. Looks like the cards these days are really that good, so we can leave it at that.
And yeah one can really only have one card in the system, so no matter how big a collection, the card doesn't matter much, when put on a shelf.

 

[oxrufiioxo]

I think @W1zzard covers the bases better than most. His reviews have helped me pick out my last 3 models and I've been happy with all 3.

All 3 were/are terrible from a value perspective but due to reasons I can afford them and they fit my performance needs.

There is enough information in each review to inform me if somthing is worth buying at the same time I wouldn't touch a zotac card regardless of how many awards it gets.

A good example is the Strix 4090 1949 usd vs my Gaming OC 4090 1699 usd both got an editors choice awards all that tells me is if I want the strix card and I'm ok paying the premium it's a very good card. I personally don't think the Strix is worth 250 usd more but I still think it's a nice card and worth it for someone who is dead set on it.

The gpu reviews here are the primary reason I'm a supporter.

 

[W1zzard]

Last I read it's all about google search rankings.

Not aware how the awards would affect ranking

Looks like the cards these days are really that good, so we can leave it at that.

They certainly are. As much as we all don't like it, NVIDIA's various approval programs make sure of that .. and those cards that get sent to me are from a subset that performs even better. I keep asking vendors for lower priced models, but they are all scared of bad reviews... and don't see the "we get free professional testing and advice that could help make our products better"

A good example is the Strix 4090 1949 usd vs my Gaming OC 4090 1699 usd both got an editors choice awards all that tells me is if I want the strix card and I'm ok paying the premium it's a very good card. I personally don't think the Strix is worth 250 usd more but I still think it's a nice card and worth it for someone who is dead set on it.

Perfectly valid thinking, I think yday I saw a thread here about someone who's super happy with his STRIX because he can afford it and he wanted the best, and didn't mind paying 10-15% more.
I want you guys to come to your own conclusions, based on your needs/constraint and the data/thoughts that I present, I'm not an influencer

The gpu reviews here are the primary reason I'm a supporter.

Thanks, much appreciated

 

[chrcoluk]

I do value the reviews and read them when I am considering a purchase, I agree with @pavle on moving on.

 

[HTC]

Alternatively, bifurcate the awarding process - one award for the series as a whole shared by all cards, and then another for individual standout designs

This seems like a good idea.

 

[Selaya]

I think @W1zzard covers the bases better than most. His reviews have helped me pick out my last 3 models and I've been happy with all 3.

All 3 were/are terrible from a value perspective but due to reasons I can afford them and they fit my performance needs.

There is enough information in each review to inform me if somthing is worth buying at the same time I wouldn't touch a zotac card regardless of how many awards it gets.

A good example is the Strix 4090 1949 usd vs my Gaming OC 4090 1699 usd both got an editors choice awards all that tells me is if I want the strix card and I'm ok paying the premium it's a very good card. I personally don't think the Strix is worth 250 usd more but I still think it's a nice card and worth it for someone who is dead set on it.

The gpu reviews here are the primary reason I'm a supporter.

absolutely, w1zz is probably like, the last person in the tech scene who's (still) doing in-depth (!) full series (ie, multi-/all AIB) gpu reviews

most sites these days review either just the FE and/or a single random AIB and call it a day

mind you, ofc it'll generate a lot of redundant data, a strix's not gonna magically be 30% faster than the FE or a ventus, but there's still a lot to be learnt from the cooling performance, noise levels et cetera
it's just a shitton of effort to do these kind of reviews, and most people, including reviewers are just lazy these days ig

 

[ir_cow]

I'm going to sneak in here and semi hi-jack (apologies), but goes with this thread.

My problem with giving awards in reviews is mainly the actual title. Editors Choice is a clear and cut one. Top dog, you think is great and personally buy it, everyone should buy it if they can afford to, etc etc. Recommended is a mixed bag. Says Highly on the title, the description allows it to fill multiple roles. It could be just one notch down for Editors Choice, simply because the reviewer doesn't consider it a product they want to buy it themselves due to personal preference. It can also be one tick above not getting a product award at all.

Great Value is hard to place as well. It can only be used for MSRP pricing and for one region. Combing with recommended make it look like a product with no equal. Cheap and good?!?! Now that's a Editors Choice.

These awards are all mind games. People who read the award descriptions paired with conclusion will understand which award was given and why. Those who skim and just look at the award will be confused. As @ThrashZone pointed out. At one point every review got a editors choice. Meaning the product was good and the reviewer would personally buy it. If awards are just given out for everything, than it loses it meaning entirely. I personally don't look at awards in reviews as a determining factor. In fact I am more incline to read reviews from sites that don't put on everything.

Using a example here. I just finished up the ASRock Z790 Sonic review. It is good for the targeted market, but so is pretty much all the motherboards at that price point. So either every product gets Highly Recommended award or none of them do. That is unless one really stands out of the crowed. Its hard to do that with a budget product. I would consider it "recommended" and that is technically what this award is for says "The reviewer would recommend this product to their friends, for specific usage scenarios", but HIGHLY really bugs me to the point where I'm thinking you can't have both in the same award. Not highly recommended. Just recommended. Is that now budget instead?