Amd Volcanic Islands details

[boxleitnerb]

I'm not saying these numbers are not true. But if they are representative and can survive under the scrutiny of reviewers, that's the interesting question.

[PedantOne]

I'm not saying these numbers are not true. But if they are representative and can survive under the scrutiny of reviewers, that's the interesting question.

history told us answers already. All "marketing results" from manufacturers are inaccurate, different games locations, different settings, every reviewer has own settings = different results

[Lanek]

Its allready hard to find 2 reviews with same results... its clear everyone have his method, location, settings, games, when they are not just re use previous review results .... ( i dont say the results we are seeing now are accurate ).

[jam2k]

[Lanek]

Is it just a compilation of the number allready given / seen, or their numbers ?.. Ok i see its the compilation of the numbers seen in " the AMD guide ".

[cx-ray]

Someone just pulled out their secret weapon - G-SYNC

(In a nutshell, the GPU controls video sync. In other words at 40FPS your monitor refresh is at 40Hz, at 100FPS your monitor is at 100Hz. So, no more tearing. Requirement is a G-SYNC equipped monitor.)

[PedantOne]

in detail you are wrong, on 40fps will be 80Hz i think, BUT this is thread about AMD

[nossy23]

So Nvidia IS releasing a new card: GTX 780 TI in mid November. Things are getting more interesting!! We'll get a new competitive situation with the new AMD cards.

[Tim]

Some very nice improvements over the 780, but on several games the difference isn't that much oddly.

[Lanek]

So Nvidia IS releasing a new card: GTX 780 TI in mid November. Things are getting more interesting!! We'll get a new competitive situation with the new AMD cards.

Clearly this will be interesting end of the year.

[LordEC911]

Someone just pulled out their secret weapon - G-SYNC

(In a nutshell, the GPU controls video sync. In other words at 40FPS your monitor refresh is at 40Hz, at 100FPS your monitor is at 100Hz. So, no more tearing. Requirement is a G-SYNC equipped monitor.)

LoL. Didn't realize it was April 1st...

[Hornet331]

so the guru3d page says that those test systems in the article have been supplied by AMD... yet they are Intel 2011 chips... while I understand the reasoning am I the only one who finds it the least bit ironic?

no not really, AMD (ATI) nearly always used the best cpus to not bottleneck there cards, they have learnd from the past.

[i found nemo]

Someone just pulled out their secret weapon - G-SYNC

(In a nutshell, the GPU controls video sync. In other words at 40FPS your monitor refresh is at 40Hz, at 100FPS your monitor is at 100Hz. So, no more tearing. Requirement is a G-SYNC equipped monitor.)

so.. 3fps is... :S 3hz?

[zalbard]

PS. Numbers with UBER-MODE activated!

So it is essentially overclocked?

By the way, 2160p results are likely influenced by 512-bit bus (which is why AMD is pushing them). So they may be quite different at 1080p and 1600p.

[LordEC911]

So it is essentially overclocked?

Not really.

[Ace123]

Please elaborate so he understands what you mean with your response. He may need help to understand the concept fully

The card is designed to run within certain specifications, and on a refrence card it's not likely to deviate from that set of guidelines. Assuming the card has a dual bios, one with quiet mode and perhaps another with "uber" mode. Baisically uber mode would sacrifice silence and maintain clock speeds and power limits closer to its TDP or power consumption limits for longer periods of time, requiring quite a bit more cooling and higher sustained fan speeds, sacrificing silence for performance.

The card was not overclocked, it's just running at or near its limit all the time.
Pendant one was trying to discredit the benchmarks by making it seem as the the cards were running out of specification by offering no explanation for his comment or what "uber-mode" was if that's what it's called. This omission of information may create doubt in the minds of the uninformed that AMD wasn't playing fair or cheating somehow.

[zalbard]

From your description, uber-mode sounds like factory overclock to me.

Higher voltages to maintain higher clocks. Higher TDP and noise levels being the consequence of this.

Not much different from manually raising TDP limit, fan speed and frequency.

So I guess, quiet mode (AMD) vs stock (Nvidia) and max OC (AMD) vs max OC (Nvidia) would still be the fair way to compare the cards.

[LordEC911]

From your description, uber-mode sounds like factory overclock to me.

Higher voltages to maintain higher clocks. Higher TDP and noise levels being the consequence of this.

Not much different from manually raising TDP limit, fan speed and frequency.

So I guess, quiet mode (AMD) vs stock (Nvidia) and max OC (AMD) vs max OC (Nvidia) would still be the fair way to compare the cards.

So you consider boost on reference/stock cards to be factory overclocked?

[Ace123]

An overclock describes a situation where a user has manually altered the speed and power of the device so that it is now running out of its original limits. Therefore there is no such thing as a factory overclock if those limits were not breached. Third party "factories" may produce overclocked cards with AMDs permission no doubt but they are in fact operating outside of the original specifications and limits, an original refrence card is not likely to do.

[bill_d]

From your description, uber-mode sounds like factory overclock to me.

Higher voltages to maintain higher clocks. Higher TDP and noise levels being the consequence of this.

Not much different from manually raising TDP limit, fan speed and frequency.

So I guess, quiet mode (AMD) vs stock (Nvidia) and max OC (AMD) vs max OC (Nvidia) would still be the fair way to compare the cards.

you just need to turn off boost , my 680 directcu ii's boost to 1238 under factory settings when they should only go to 1201 max

[[XC] hipno650]

no not really, AMD (ATI) nearly always used the best cpus to not bottleneck there cards, they have learnd from the past.

ahh well I understand the reason perfectly fine... i just had a little laugh on the inside when I saw it

[Sparky]

I figure it makes sure there is absolutely no reason for anyone to even suggest platform bias. Verifies even ground between the two cards.

[Ace123]

It makes perfect sense that the proving ground belong to a third party. Same reason the superbowl is played on a neutral field. Very astute anaylisis

[spursindonesia]

Please elaborate so he understands what you mean with your response. He may need help to understand the concept fully

The card is designed to run within certain specifications, and on a refrence card it's not likely to deviate from that set of guidelines. Assuming the card has a dual bios, one with quiet mode and perhaps another with "uber" mode. Baisically uber mode would sacrifice silence and maintain clock speeds and power limits closer to its TDP or power consumption limits for longer periods of time, requiring quite a bit more cooling and higher sustained fan speeds, sacrificing silence for performance.

The card was not overclocked, it's just running at or near its limit all the time.
Pendant one was trying to discredit the benchmarks by making it seem as the the cards were running out of specification by offering no explanation for his comment or what "uber-mode" was if that's what it's called. This omission of information may create doubt in the minds of the uninformed that AMD wasn't playing fair or cheating somehow.

What do you expect, mate ? That's exactly how OBR operate, in thread crapping everything AMD related, past, present, and in the future. after past banishment, he tried to make another (expected) comeback under another nick name & appear more nicer + sincere in providing leaks & infos, but old habit & true color always show themselves out in the end.

[tajoh111]

So you consider boost on reference/stock cards to be factory overclocked?

There are two instances I would consider it an overclock.

If its like awesome mode like on the 6990, where activating a switch gives higher clocks, gives more voltage, speeds up the fans and lastly voids the warranty, it would definitely be an overclock.

The second instance I would consider is based on Tahiti's reference form.

How quiet is AMD quiet mode? Tahiti based cards have been pretty loud using a reference cooler. If quiet more is more of a standard mode that matches the loudness of current tahiti cards, I would then have to consider the loud mode, even if it didn't void the warranty based on loudness

Referenced designed Tahiti cards are loud enough as it is. The max I would want from a stock card. If turning on loud mode, makes it loud to an undesirable level beyond that, then I would for consider it an overclock.

Reference design vs reference design, gk104 has been much quieter than tahiti, even with boost enabled. This was further emphasized with the reference cooler based 7970 ghz editions which were pretty much unusable with the reference cooler. Enough that card makers just skipped the reference cooler altogether since it couldn't cool enough without sounding like a jet engine.

http://www.anandtech.com/show/6025/r...-to-gtx-680/16