Asus cheating for performance stats?

[zoomee]

Ahead of the weekend a bit of a controversy was stirring with both Asus and MSI being accused of sending out Nvidia GTX 1000 Series graphics cards that weren't exactly the same as the shipping retail versions. In a news post published last Thursday, TechPowerUp said that both Asus and MSI had employed "questionable tactics" to show their new GTX 1000 Series cards in a favourable light.

http://hexus.net/tech/news/graphics...1000-series-review-cards-use-faster-defaults/

Whats to say that Asus doesn't do the same for its routers? - From past experience (with 2 Asus routers) I've generally found performance is great to start with but waters down as further more reliable and stable firmware is provided.......

No accusations here - just a thread to discuss.

 

[System Error Message]

compared to reference boards for a GPU there are many things one could add that can help for extra performance. Better power system, better cooling system helps for a better overclock. A hotter CPU uses more watts than a colder one at the same temperature. Nvidia probably doesnt want its launch of products to have different variations other than what it wants at the start. Usually in the past custom versions (like those with double VRAM, more fans, etc) usually come out later.


In the case of routers, not much you can do here, only overclock the hardware as all the software and hardware are the same. The libraries are the same, they are all linux based. Perhaps ASUS clocked the ASIC higher for faster throughput?

 

[RMerlin]

Unlike video cards, routers aren't exactly marketed along with benchmark numbers, so I don't see what "tactics" that they might have used. The CPU is clocked at the rated speed, radio power output is validated by regulation bodies to be within specifications, and they don't provide any comparative numbers or test results in their marketing material.

 

[zoomee]

I'm thinking more for wireless performance gains - tweaks here and there which aren't made into official firmware which allow Asus to come ahead of the pack when sites like SNB review them in batches.

Thats basically what Asus and MSI have done here - they've incorporated software overclocking into the firmware to "make it easier for the reviewer" - in effect this is a form of cheating as it shows they are different firmware from what we as normal consumers are getting.

Question is do SNB use manufacturer supplied routers for their testing or retail units picked off the shelf?

 

[RMerlin]

I'm thinking more for wireless performance gains - tweaks here and there which aren't made into official firmware which allow Asus to come ahead of the pack when sites like SNB review them in batches.

Thats basically what Asus and MSI have done here - they've incorporated software overclocking into the firmware to "make it easier for the reviewer" - in effect this is a form of cheating as it shows they are different firmware from what we as normal consumers are getting.

Question is do SNB use manufacturer supplied routers for their testing or retail units picked off the shelf?

A lot of Tim's review have been done on devices he got himself. Also, he typically download and install the latest firmware available, so that would overwrite any customized firmware that might have been there.

There's simply no evidence so support such a theory. Let's not put our tin foil hats just yet...

 

[sfx2000]

A lot of Tim's review have been done on devices he got himself. Also, he typically download and install the latest firmware available, so that would overwrite any customized firmware that might have been there.

There's simply no evidence so support such a theory. Let's not put our tin foil hats just yet...

There might be ways to "game" the test plan, but Tim's test coverage would show up any "optimizations" - and honestly, I haven't seen any - and devices have been retested, and they've been consistent, and strong correlation with reports on the forums from end-users/customers.

It would be very hard to cheat... and it's a competitive enough environment that other vendors would "call" a cheater...

 

[Fizix]

. . . . Let's not put our tin foil hats just yet...

Awwwwwww - you're no fun‼

 

[paraplu]

Asus started cheating a couple of years back with their n66u. That device blew all other devices away in reviews. As it happened, they turned up the tx power above the allowed, to get outstanding reviews. Netgear a.o. complained and won the lawsuit eventually.

 

[TonyH]

http://hexus.net/tech/news/graphics...1000-series-review-cards-use-faster-defaults/

Whats to say that Asus doesn't do the same for its routers? - From past experience (with 2 Asus routers) I've generally found performance is great to start with but waters down as further more reliable and stable firmware is provided.......

No accusations here - just a thread to discuss.

Just like automobile manufacturer's claim on MPG figures. So many similar things going on so many different products.
Starbuck coffee is sued for underfilling ....

 

[sfx2000]

Asus started cheating a couple of years back with their n66u. That device blew all other devices away in reviews. As it happened, they turned up the tx power above the allowed, to get outstanding reviews. Netgear a.o. complained and won the lawsuit eventually.

And they were caught - which makes it even more clear these days, the community is watching, and so are the vendors - pretty hard to cheat these days...

 

[RMerlin]

Asus started cheating a couple of years back with their n66u. That device blew all other devices away in reviews. As it happened, they turned up the tx power above the allowed, to get outstanding reviews. Netgear a.o. complained and won the lawsuit eventually.

That's not the same thing as what the OP was referring to:

Asus and MSI being accused of sending out Nvidia GTX 1000 Series graphics cards that weren't exactly the same as the shipping retail versions.

Asus weren't sending special versions of the RT-N66U to reviewers, the product they actually sold had higher output levels than what was allowed by the FCC. So in short, what reviews indicated were the exact same things you'd expect at home, so it wasn't as deceptive as what happened with those video cards, or what Samsung did with some of their Android products that would detect the presence of a benchmarking software, and boost performance to skew its results. Same thing nVidia did a few years ago as well, skewing results with "custom" optimizations whenever you ran 3DMark on their cards.

 

[sfx2000]

Same thing nVidia did a few years ago as well, skewing results with "custom" optimizations whenever you ran 3DMark on their cards.

And let's not forget the skullduggery around Intel, AMD, and Intel's compiler and running AMD in a less than optimal code path in certain benchmark software...

 

[System Error Message]

And let's not forget the skullduggery around Intel, AMD, and Intel's compiler and running AMD in a less than optimal code path in certain benchmark software...

This is why i use both intel and AMD's compiler so i compile for specific CPUs using their own compilers instead.

 

[paraplu]

That's not the same thing as what the OP was referring to:



Asus weren't sending special versions of the RT-N66U to reviewers, the product they actually sold had higher output levels than what was allowed by the FCC. So in short, what reviews indicated were the exact same things you'd expect at home, so it wasn't as deceptive as what happened with those video cards, or what Samsung did with some of their Android products that would detect the presence of a benchmarking software, and boost performance to skew its results. Same thing nVidia did a few years ago as well, skewing results with "custom" optimizations whenever you ran 3DMark on their cards.

Yes true, but the damage was already done. The (initial review) word got out, and happened to be one of the main reasons of this website asus' interest success. They rectified their firmware afterwards but still many old forks floating around, particularly on this forum.