• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Should AT include OC'd Cards in Reviews?

Page 7 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Are OCed variants of graphics boards appropriate for reviews

  • It is neither ethical nor fair to ever include such in a product review

  • It is unfair, but does not cross ethical bounds; many nv/ati skewed sites do it

  • It would be only fair with prominent caveats in the review, and clearly visible differentiations

  • It can be fair, depending on the market conditions, and the OCed variants availability

  • It can be fair, but not with a large OC such as on the evga FTW

  • It is only fair when comparing OCed variants from both camps

  • It is normally fair, but not on initial product reviews like the debut of 6800s

  • It is only fair when the OCed variant is the highest volume of the model

  • It is such a minutia, I don't understand why most even care


Results are only viewable after voting.
Last edited:
Thanks, good to know; I have found people saying Afterburner will oc it but I could not find anything anywhere about overvolting it with Afterburner. I am aware that the Galaxy software that comes with the card allows only oc, not overvolting.

All Galaxy cards (gtx460,470,480'S) have the proper hardware to overvolt.
Your good buddy.
 
The oc card should have been included as a comparison with an overclock of the cards being reviewed. You should only compare stock with stock and oc with oc.

No overclock of the AMD cards was attempted which frankly I find highly lacking and so I believe that the Evga card should not have been there. IF a further article was in the works to compare overclocking save it for that spot.
 
Then the newer relevant info like same priced factory pre-OCed AMD cards should most definitely be taken into consideration. When AMD cut prices on HD5770 when GTS450 was launched, that's relevant new info we all wanted to know. 😀

The thing Ryan pointed to its disclaimer wasn't something like this.

He wondered, and I guess Anand as well, exactly how many EVGA GTX460 FTW will be available and for how long.

Remember cards like the X850XTPE or whatever that was virtually impossible to get their hands on?

I'm not saying that is the case in here but I wonder if this doesn't open the door to such tactics. And in the case of the X850XTPE you could call it a paper launch or a ghost model.

I think a simple reference to the fact that there are factory OC models and then a link to demonstrate how that card scales with overclocking would still be informative (actually more informative) and would prevent such tactics of taking place, where a company would send cherry picked limited/small editions, just to top the competitor on the charts at launch.
 
I'm okay with factory overclocked cards in reviews, so long as the same benchmarks are ran on all applicable cards in the write up. Both stock and OC'd cards are available for consumers to buy, so information about the performance of both is required for a consumer to make an informed purchase decision.

Now, aftermarket OCing of the video card can get sticky. In my opinion, its okay, so long as the OC'd cards are both shown as stock speeds as well. If you benchmark a stock card against an aftermarket OC'd card, thats not really a fair competition. Not everyone overclocks, and for many, that aspect isn't something they care about in the slightest.
 
Or GTX470 and 6870 trade blows and trade blows by api and I can link a few more reviews if you want where the GTX470 trades blows with a 6870.

Performance justifications are not necessary because a GTX470 has another 150mhz-200mhz of overclocking headroom at 1.087V. Personally, I don't even consider a GTX470 to be a good competitor to the HD6870 simply because of the 90W extra power consumption and a much louder stock cooler. However, the factory pre-overclocked GTX460 FTW runs quieter at load than an HD6870 and has a free game in the bundle, plus a 90-day step-up.
 
Last edited:
I don't get why is it such a big issue. Factory oced cards have been reviewed countless of times by most of the hardware sites out there.

There is fairness and merit in reviewing BOTH camps with oced units and I suppose the timing of this particular review is also in discussion, but I wouldn't worry that much about it. I doubt the FTW edition is less expensive than the 6870 outside US.

Seems to me more of a discussion between stock holders from both companies.

So AMD had a less than stellar launch performance wise with the 6850-70, so what? If the rates get lower at the retailers then we are all winners. Things are not peachy for either AMD or Nvidia when this stuff happens to them? Are they losing money in the process? I seriously doubt that but even if it's true then that serves them right for baiting us into buying ever less performance increase than we did before.

Regarding the AT review, if there's even ONE AMD card factory oced out there then I guess there's bias from their side for not adding it to the review. Getting it delivered to their doorstep or not, it makes no difference, the site should do this out of integrity if nothing else.

Just my 2c.
 
I voted yes. Unlike cpu's, in which some models overclock rather nicely, this is different. They are selling you a overclocked card from the factory, which for all intents and purposes means it's the default for that particular brand of card. If it doesn't run as advertised it's faulty. If AMD brand gpu's want to provide the same thing then it should be fair to compare overclocked AMD's to overclocked nvidias. Obviously the overclocked AMD's aren't out the day of a launch so it's slightly biased.
I would have been happy enough comparing an overclocked AMD card with an overclocked nvidia card, which was done. I say let it slide. We are enthusiasts, we overclock stuff 🙂
 
i think it's fine to include because the question is: what can i buy with X dollars in the competitive price point bracket?

the evga card in this review fits in nicely here because, well, why would you buy any other gtx460 model after seeing this one at its price?

in any case, stock vs stock comparisons are fine, but honestly, who here in this community ever runs anything at stock? (personally, i'd like it better if the review articles contained the "average" oc for the popular cards in addition to the stock ones) and hell, throw in a factory oc model too if the price is right.
 
Why is this even an issue? Stock clocked GTX 460 1GB performance numbers are included in the charts. As for the overclocked model, seems like just about all Nvidia partners sell cards factory overclocked to similar levels and with warranty.
 
The problem with this analysis lies on the PRICE of the heavily OCed card.

At one point, Nvidia is marketing massive price drop of 460, at the same time, pressing review sites to compare the 6xxx series using massively Oced cards that also charge a premium.

To the careful reader, they will notice it. To the non so careful one, they will be mislead to think that the low price 460 will match the performance of the more expensive 6870.

This is the same case when we see 460 768MB and 460 1G launched, and people reference to 460 768MB price AND link to 460 1G performance (which there is a difference due to reason other than RAM).

I'd wait to see the 475 review, see if AT will pitch OCed 6xxx cards vs the 475 to determine if AT is biased to Nvidia.
 
I think factory OC cards are inherently limited availability/product life cycle, have the advantage and disadvantage of already being OC'd, and provide some amount of product confusion due to varying OC's available. Because of this they should probably be reviewed in their own article devoted to the OC card and left out of articles devoted to baseline cards.

But I don't think it was bias in any way and I do want the benchmarks, more info is better. I just think there are too many caveats to an OC card to compare with immediacy to a stock card in an article about the stock card.
 
Last edited:
So AMD had a less than stellar launch performance wise with the 6850-70, so what?

I wouldn't say they had less than stellar launch. You now get nearly HD5870 performance for $110+ less. Besides the few factory overclocked GTX460 cards from MSI, Palit and EVGA, the HD6850/70 are the better cards. The GTX470 is inferior to the HD6870 in terms of noise and power consumption. $430 GTX480 seems ridiculously overpriced considering HD6870s in CF at $480 will mop the floor with it. Also, this launch made HD5850/70/5970 obsolete. So, overall I would say it was pretty good!!

The only spot where NV has something worthy of buying besides GTX460 1GB OCed is $150 GTX460 768mb. :awe: With G92/GTS250 cards soon to be gone and horrible effort with GT430 and GTS450, NV's product lineup is in the worst spot I have ever seen.
 
Last edited:
Performance justifications are not necessary because a GTX470 has another 150mhz-200mhz of overclocking headroom at 1.087V. Personally, I don't even consider a GTX470 to be a good competitor to the HD6870 simply because of the 90W extra power consumption and a much louder stock cooler. However, the factory pre-overclocked GTX460 FTW runs quieter at load than an HD6870 and has a free game in the bundle, plus a 90-day step-up.
where is the free game listed?
 
So when does an OC'ed card have the right to be in the review? How far does it need to be overclocked? Who decides how far it needs to be overclocked?

So are they going to include OC'ed cards in future reviews or not? If it AT didn't do it cause nV pressured them, why didn't they include the 5870 SOC or the 5770 HAWK? Or does it have be 25%?

Not all OC cards scale the same with OCs. So I don't know whats going to happen now.
 
Why is this even an issue? Stock clocked GTX 460 1GB performance numbers are included in the charts. As for the overclocked model, seems like just about all Nvidia partners sell cards factory overclocked to similar levels and with warranty.

I have to agree with Zap. I think it is fair to test the OCed cards as long as:

* The stock cards get tested too
* The OCed cards are widely available
* The difference in price between stock and OCed cards is clearly mentioned

Seems like this applied to the FTW GTX460 in Anand's review, I don't see what all the fuzz is about.
 
So when does an OC'ed card have the right to be in the review? How far does it need to be overclocked? Who decides how far it needs to be overclocked?

So are they going to include OC'ed cards in future reviews or not? If it AT didn't do it cause nV pressured them, why didn't they include the 5870 SOC or the 5770 HAWK? Or does it have be 25%?

Not all OC cards scale the same with OCs. So I don't know whats going to happen now.

Well including something representative of the majority of cards on the market seems fair.

Only 4 of the 26 GTX460 1GB's listed on Newegg carried the stock core clock.
This compares to 5 of 21 HD5770's being overclocked. And in percentage terms, the HD5770 overclocks are much smaller than the GTX460 ones typically are.

It's a totally different situation, in one case most cards are NOT stock.
In the other, most ARE stock. The GTX460 was almost made to be overclocked by AIB partners, so including cards which represent that is fair, it also more accurately reflects the state of the market for that card.

The choice of overclocked card might be a bit too "high end", since most overclocks are lower than the card they used, but it works because it pretty much gives you an upper and lower bound for GTX460 performance. Stock being the lower bound minimum of a GTX460, and the OC version they picked giving the maximum performance that can be expected from a factory GTX460. Most cards fall between these two extremes.
 
Last edited:
Copy and paste of my post in the other thread.

Well including something representative of the majority of cards on the market seems fair.

Only 4 of the 26 GTX460 1GB's listed on Newegg carried the stock core clock.
This compares to 5 of 21 HD5770's being overclocked. And in percentage terms, the HD5770 overclocks are much smaller than the GTX460 ones typically are.

It's a totally different situation, in one case most cards are NOT stock.
In the other, most ARE stock. The GTX460 was almost made to be overclocked by AIB partners, so including cards which represent that is fair, it also more accurately reflects the state of the market for that card.

The choice of overclocked card might be a bit too "high end", since most overclocks are lower than the card they used, but it works because it pretty much gives you an upper and lower bound for GTX460 performance. Stock being the lower bound minimum of a GTX460, and the OC version they picked giving the maximum performance that can be expected from a factory GTX460. Most cards fall between these two extremes.
 
As long as there is a reference in the mix why not? I think its of value to see what a fairly high factory OC'd card thats readily available and now cheaper than the new card being reviewed can do.
 
Last edited:
its funny that the tiniest of irrelevant minutiae is made into a question of ETHICS and blown out of proportion...

Anandtech was very fair in analyzing this with their disclaimer and in depth explanation of what and why. But its not like it is a big deal anyways.
 
Back
Top