TechPowerUp Impact: MSI Issues "OC Mode by Default" BIOSes

[Prima.Vera]

I wonder why nobody screamed since 3 years ago when they released the 780Ti using EXACTLY the same practice...

 

[Mussels]

So many are looking for something nefarious here, but if the same OC Mode clocks are available to the public by enabling it there is nothing foul about wanting reviewers to test the OC Mode because that's what they would have done had it been a toggle switch.

"end users can overclock their CPU by 500Mhz if they want to, so the review models are clocked 500Mhz higher"
"end users can upgrade the engine to a V8 themselves if they want to"

yeah... no. small or not, its NOT the same product that is being sold so it literally is false advertising.

 

[Bjørgersson]

Am I the only one not giving a damn about these 30-40 MHz differences?

 

[btarunr]

Am I the only one not giving a damn about these 30-40 MHz differences?

I think john_ summed this up nicely:

"Should I buy the MSI card that manages 100.3 fps, or the Gigabyte that scores only 99.9 fps?"
...
" Hi. I want to order an MSI."

 

[Bjørgersson]

I think john_ summed this up nicely:

He did, indeed.

 

[Ubersonic]

Just a note, the TPU review of the MSI Gaming 980ti also appears to be using an "OC mode by default" BIOS so they have been at this for some time.

Do they have any plans to release the "OC mode by default" BIOS for older cards? As I would like one for my 980ti as I don't want to use their app.

 

[refillable]

The same old response like ASUS. Not good at all, these are some shady business which definitely needs to stop. Tells that these companies do in fact cheat, not that they did this by accident.

I wonder why nobody screamed since 3 years ago when they released the 780Ti using EXACTLY the same practice...

Yep, but at least I'm thankful that this practice hasn't gone unnoticed, even though it took 3 years for someone to realize.

 

[Ungari]

"end users can overclock their CPU by 500Mhz if they want to, so the review models are clocked 500Mhz higher"
"end users can upgrade the engine to a V8 themselves if they want to"

yeah... no. small or not, its NOT the same product that is being sold so it literally is false advertising.

I'm not sure I understand. You are saying that it is not the same product only because the default setting for review samples is OC Mode, the same OC Mode that can be selected by any user at will?

 

[The Quim Reaper]

Once again, an issue blown out of all proportion by the internet rage monkeys.

 

[Ubersonic]

Once again, an issue blown out of all proportion by the internet rage monkeys.

How so? just because the cheating was minor doesn't change the fact it happened and had real world benefits for the manufacturers concerned. Nobody says "Lance Armstrong may have done performance enhancing drugs but it's not like he had his brain transplanted into a robotic superbody so it's all fine".

 

[Ungari]

How so? just because the cheating was minor doesn't change the fact it happened and had real world benefits for the manufacturers concerned. Nobody says "Lance Armstrong may have done performance enhancing drugs but it's not like he had his brain transplanted into a robotic superbody so it's all fine".

How is it cheating when the reviewers would have chosen to test the card at it's highest factory OC rate regardless, or at least stated which setting was used to benchmark these cards?

 

[Mussels]

I'm not sure I understand. You are saying that it is not the same product only because the default setting for review samples is OC Mode, the same OC Mode that can be selected by any user at will?

that is completely correct. They are advertising the product with different specs than it comes with - there is a difference between them.

Whether its a small difference or not is irrelevant, its a very dangerous precedent to allow, as that will only grow further. "Cherry picked" review samples have long been mocked in the industry.

 

[Ungari]

that is completely correct. They are advertising the product with different specs than it comes with - there is a difference between them.

Whether its a small difference or not is irrelevant, its a very dangerous precedent to allow, as that will only grow further. "Cherry picked" review samples have long been mocked in the industry.

Except that Jaytwocents review sample while given the OC Mode as default turned out to be a poor silicon chip that did not manually overclock to the level of his Founders Edition card.
I could understand if the default Bios OC for review samples was superior to the OC Mode setting that was available to the retail customers, but this is not the case, it is merely that the OC Mode was selected by default.

The advertised clocks for each factory preset Mode were correctly stated by these companies, so how is it a different spec?

MSI Gaming X 1080

• Core Clock 1708 MHz (OC Mode) 1683 MHz (Gaming Mode) 1607 MHz (Silent Mode)
• Boost Clock 1847 MHz (OC Mode) 1822 MHz (Gaming Mode) 1733 MHz (Silent Mode)

 

[Mussels]

it doesnt matter if they list 20 modes - if the default mode changes between products sold under the same name, are they really the same product? You're saying you see no difference with a core clock that could be any of those 6 numbers, depending on which box you open?

 

[EarthDog]

Sorry... what crime did they commit?

 

[rtwjunkie]

The advertised clocks for each factory preset Mode were correctly stated by these companies, so how is it a different spec?

MSI Gaming X 1080

• Core Clock 1708 MHz (OC Mode) 1683 MHz (Gaming Mode) 1607 MHz (Silent Mode)
• Boost Clock 1847 MHz (OC Mode) 1822 MHz (Gaming Mode) 1733 MHz (Silent Mode)

I have for years noticed that my MSI cards default to the Gaming mode. I have also noticed that W1zzard's GPU-Z screenshots show the OC mode for default. I guess I always figured he had set it there himself, because if he is testing the card overclock ability, then he should start at the factory limits. The factory limits for MSI Gaming cards are the OC mode.

So, it never bothered me that he was doing testing at about 30Mhz higher than I received a card. The little bit hardly makes a huge difference; maybe 1 FPS. What his benchmarking shows is a representation, not what I can necessarily expect to get. Reviews are just to give you an idea what the card is capable of and how it will likely perform.

I agree it was deceptive on their part. I'm glad it was brought to light and they are changing their stance. It's not a world-stopping event, though. @Mussels has a great, even-keeled attitude about it. The internet "rage" over this, as if they just killed all their employees and hired replacements immediately without a second thought is...bewildering.

 

[Ubersonic]

The advertised clocks for each factory preset Mode were correctly stated by these companies, so how is it a different spec?

MSI Gaming X 1080

• Core Clock 1708 MHz (OC Mode) 1683 MHz (Gaming Mode) 1607 MHz (Silent Mode)
• Boost Clock 1847 MHz (OC Mode) 1822 MHz (Gaming Mode) 1733 MHz (Silent Mode)

Basically, the retail card has a 1683MHz core and comes with software that can raise it to 1708MHz for more power, or reduce it to 1607MHz for less noise. The review card on the other hand comes with a 1708MHz core and comes with software that can lower it to 1683MHz or 1607MHz.

It was done to mislead reviewers and customers about out of box performance (if it wasn't supposed to mislead then they would have admitted to it sometime in the last three years). The will be many many customers who bought the card based on it's review numbers with the expectation that it would do those numbers out of the box plus higher if they used the software to activate the OC mode. There will also be many customers who bought the cards instead of other cards which appeared slower in reviews but were actually faster in practice.

 

[Mussels]

Sorry... what crime did they commit?

deceptive advertising, aka false advertising. They are getting third parties to test and verify their products, which are set to (miniscule) higher speeds than the models being sold, out of the box.
No one here is saying the speed change is a big deal, the fact they ship with a customised bios is the big deal. whats next, the review models have a higher TDP/wattage limit, or a more aggressive fan curve? next time will it be 50MHz, 100Mhz, 200? Will the put in a hidden voltage offset to make the cards OC better than retail models?

as someone who used to do hardware reviews, nothing is worse than finding out the product you got and recommended has been changed for the worse, in the retail models. It makes a liar out of the people who reviewed them.

 

[EarthDog]

it doesnt matter if they list 20 modes - if the default mode changes between products sold under the same name, are they really the same product? You're saying you see no difference with a core clock that could be any of those 6 numbers, depending on which box you open?

Yes, they are the same product.

First of all, those "6" numbers are really just 3... the boost clock, in the end, is what is important. And yes, the app can change between them. That doesn't mean its a different card. There are variations on each level of card and those clocks.

EDIT: As far as making a liar out of the reviewer, no... just no. It doesn't. If one can't explain their way out of some troll calling a reviewer a liar for this, that person probably shouldn't be writing reviews (and takes things too seriously).

EDIT2: I have also mentioned this point in my reviews. What I did miss in all of that is the BIOS was doing it and not the app I had installed.

 

[Ungari]

I have yet to see a reviewer that did not state which preset Mode they were testing the card with, and in the case of a card that ships with multiple presets I never saw a reviewer that ignored the highest factory Boost clock preset in their review.
If a user buys a card based on a review that shows OC factory preset benchmarks and fails to select this preset for himself to get similar results, it is not the fault of the manufacturer.

 

[Ubersonic]

Yes, they are the same product.
[/snip]
That doesn't mean its a different card. There are variations on each level of card and those clocks.

The only difference between an EVGA reference 980 and an EVGA reference 980 SC is the clocks and the price, nobody claims they are the same card.

It's the exact same for the HD7970 and the HD7970 GHz edition.

They tricked reviewers into recommending a different product to what customers received and they hid the fact they were doing it.

 

[EarthDog]

You missed my point. I agree with you on the goalposts you setup, however, I'm on a different field kicking at that.

Let me explain a bit further. MSI 1070 Gaming X comes with a choice of 3 clockspeeds. Review samples set to the highest it can go. Its not a different card because of that. However for the sake of clarity, it would be best if the samples sent were the same as retail... but it is NOT a different card.

 

[jsfitz54]

deceptive advertising, aka false advertising. They are getting third parties to test and verify their products, which are set to (miniscule) higher speeds than the models being sold, out of the box.
No one here is saying the speed change is a big deal, the fact they ship with a customised bios is the big deal. whats next, the review models have a higher TDP/wattage limit, or a more aggressive fan curve? next time will it be 50MHz, 100Mhz, 200? Will the put in a hidden voltage offset to make the cards OC better than retail models?

as someone who used to do hardware reviews, nothing is worse than finding out the product you got and recommended has been changed for the worse, in the retail models. It makes a liar out of the people who reviewed them.

I like the way you said this.

 

[Ubersonic]

You missed my point. I agree with you on the goalposts you setup, however, I'm on a different field kicking at that.

Let me explain a bit further. MSI 1070 Gaming X comes with a choice of 3 clockspeeds. Review samples set to the highest it can go. Its not a different card because of that. However for the sake of clarity, it would be best if the samples sent were the same as retail... but it is NOT a different card.

The cards are physically identicle, the only difference is they have a differen't BIOS flashed. That's the same difference as EVGAs reference card, their reference SC card, their reference FTW card. All of which are considered different cards.

The same was true of the HD7970/HD7970ge/280X, the HD7950/HD7950 Boost, the GTX680/GTX770, etc.

 

[EarthDog]

Repeating yourself doesn't help me reach the same conclusion. Thanks though.

Edit: Do the tiers/models of these cards overlap with the silent/gaming/oc modes? Evga likely would since they will have like 8 models, but they don't do this. What about asus and msi? The meager increases are not bumping it up to the next tier in their lineup.