Tom's Hardware: ATI cheating?

[Wblane]

Did anyone see the Tom's Hardware article that shows how ATI might be cheating
on benchmarks? The article quotes some games developers that also point out ATI
is possibly inflating benchmarks by substituting lower quality textures even
when high quality textures are selected. I remember Nvidia got slammed here a
while back for optimizing it's drivers but it looks like ATI is playing the
same game now.
-Bill (remove "botizer" to reply via email)

 

[Andrew]

Did anyone see the Tom's Hardware article that shows how ATI might be cheating
on benchmarks?

What is the direct link to the article?

 

[LBJGH]

I think you mean Nvidia cheating no?

Did anyone see the Tom's Hardware article that shows how ATI might be cheating
on benchmarks? The article quotes some games developers that also point out ATI
is possibly inflating benchmarks by substituting lower quality textures even
when high quality textures are selected. I remember Nvidia got slammed here a
while back for optimizing it's drivers but it looks like ATI is playing the
same game now.
-Bill (remove "botizer" to reply via email)

 

[Inglo]

Did anyone see the Tom's Hardware article that shows how ATI might be cheating
on benchmarks? The article quotes some games developers that also point out ATI
is possibly inflating benchmarks by substituting lower quality textures even
when high quality textures are selected. I remember Nvidia got slammed here a
while back for optimizing it's drivers but it looks like ATI is playing the
same game now.
-Bill (remove "botizer" to reply via email)

In case you were worried, vague innuendo is totally acceptable on usenet.

 

[Skid]

In case you were worried, vague innuendo is totally acceptable on usenet.

And no matter how high you build the battlements, trolls still crawl in
through the sewers.

 

[JAD]

cheating in benchmarks OMG your kidding..!!!!!!! Cheating is a human past time.... kid goes and spends dads 50 bucks on a game
and immediately logs on to a cheat site and blows through the game in one night. Advertising and cheating \ lying go hand in
hand..... BTW who really gives a CRAP about benchmarks....just another way to hype and >>>>>ADVERTISE<<<<!!!!!!

 

[PB]

I think you mean Nvidia cheating no?

They both cheat, if you don't believe it, then you need to rejoin the real
world.

Oh, and for Wblane, that article is old news.

DD

 

[This space left intentionally blank]

cheating in benchmarks OMG your kidding..!!!!!!! Cheating is a human past time.... kid goes and spends dads 50 bucks on a game
and immediately logs on to a cheat site and blows through the game in one night.

Yea, but lots of games make up for weak AI by, you guessed it,
CHEATING!

 

[Gandolf_One]

Snip

.. Advertising and cheating \ lying go hand

you forgot to add politicians to that

 

[LBJGH]

That goes without saying... Here in Ontario the bone-heads elected the Lying
Liberal party... [insert Homer voice] stupid people.

 

[Rob G]

Motherboard manufacturers are also cheating by secretly raising the bus
frequencies. MSI
did it with their 'dynamic overclocking' which was a hidden feature until
Tom's Hardware
discovered it. Now it's user selectable in the bios. Even so, the default
200 fsb setting is really
203.

 

[Wblane]

This is the link to one of the questionable aspects of ATI's recent drivers:
http://www.tomshardware.com/graphic/20031023/nvidia-nv38-nv36-20.html

This article also quotes two different games publishers who also comment on
ATI's interesting driver problems/cheats.

Personally, I'm kinda pissed off that the only reason I bought ATI (i.e. better
performance) might all be a sham.

What is the direct link to the article?
--

-Bill (remove "botizer" to reply via email)

 

[LeeB18509]

Give me a ****ing break...
http://www.gamersdepot.com/hardware/video_cards/ati_vs_nvidia/dx9_rel50/001.htm

 

[Anders E.D]

Did anyone see the Tom's Hardware article that shows how ATI might be
cheating
on benchmarks? The article quotes some games developers that also point
out ATI
is possibly inflating benchmarks by substituting lower quality textures
even
when high quality textures are selected. I remember Nvidia got slammed
here a
while back for optimizing it's drivers but it looks like ATI is playing
the
same game now.
-Bill (remove "botizer" to reply via email)

You should read the following two articles. It's a direct response to the
Tom's Hardware article and it examines the ATI drivers in a completely
unbiased way

http://elitebastards.com/page.php?pageid=2397&head=1&comments=1

http://elitebastards.com/page.php?pageid=2416&head=1&comments=1

( It was a rendering bug in the previous ATI drivers, but it has probably
been fixed in the latest drivers. In other words: No cheating )

 

[Skid]

This is the link to one of the questionable aspects of ATI's recent drivers:
http://www.tomshardware.com/graphic/20031023/nvidia-nv38-nv36-20.html

And that article is more than two months old -- a lifetime in these parts --
and has been thoroughly refuted on a half-dozen other websites. You'll find
the links in replies from other posters in this thread.

Bottom line, it's old news to us.

A good while back, ATI was caught optimizing driver code for Quake
benchmarks. Nvidia said it was cheating, ATI said it was an attempt to
improve performance for gamers.

More recently, Nvidia blasted 3DMark 03 as an unfair and useless benchmark
(because their cards sucked at it,) and then was caught rewriting drivers to
get higher scores. ATI said they were cheating, Nvidia said they were taking
advantage of special routines peculiar to their hardware.

The whole tempest in a teapot at Tom's was an abortive attempt to fan those
old flames once again. But as he often does, Tom got the facts wrong.

It's a cutthroat competition for video card sales, and gamers weighing the
choices unfortunately lend a lot of credence to benchmarks that can -- and
are -- manipulated to make one side or the other look better.

For the moment, ATI has the more potent hardware, drivers have vastly
improved over the tortures of the past, and sales are up. Does it really
surprise you that when Nvidia is caught red-handed, one of their defenses is
to leak to a favored webmaster an allegation that the other guys are
cheating too?

 

[K]

More recently, Nvidia blasted 3DMark 03 as an unfair and useless benchmark
(because their cards sucked at it,) and then was caught rewriting drivers to
get higher scores. ATI said they were cheating, Nvidia said they were taking
advantage of special routines peculiar to their hardware.

That's because 3DMark03 *is* a contrived and useless benchmark. Nvidia
shouldn't even bother cheating to get better scores in it.

As an R9600 pro user having gone from a GF3 I find that there is many
graphical anomalies that simply weren't there with my old card. I don't
think these are intentional cheating to inflate benchmark scores, it's just
that ATi's drivers are still ****ed. I have no doubt that that ATi has the
fastest hardware, it's just that they couldn't write a decent set of drivers
to save their own skins.

Did you know that Nvidia employs more software engineers than those working
on the hardware? It shows and Ati could learn a thing or two.

K

 

[FaxCap]

it's just that ATi's drivers are still ****ed. I have no doubt that that ATi has the
fastest hardware, it's just that they couldn't write a decent set of drivers
to save their own skins.

Did you know that Nvidia employs more software engineers than those working
on the hardware? It shows and Ati could learn a thing or two.

I agree 100%, K.....that's been ATI's problem for years now....flakey
drivers but great hardware.

 

[Darthy]

This is the link to one of the questionable aspects of ATI's recent drivers:
http://www.tomshardware.com/graphic/20031023/nvidia-nv38-nv36-20.html

This article also quotes two different games publishers who also comment on
ATI's interesting driver problems/cheats.

Personally, I'm kinda pissed off that the only reason I bought ATI (i.e. better
performance) might all be a sham.

As another article from another site pointed out...

ATI and Nvidia are two different types of technolgy... the FX and
9x000 series cards do things differently.

I looked at the comparision... its hard to say which WAY looks better
or worse... different yes.... but *IS IT WORSE?*

The Nvidia cheating involved were obvious... parts of the benchmark or
details of a game were missing... like not drawing smoke or bullets
etc... In the end, no heat from this OCT article.