Rumor: Price Cuts on GTX660Ti series coming next week

[blackened23]

Personally, I would find it hilarious if anyone claimed nvidia to be a saint regarding prices. Particularly from the 8800-GTX 500. They're a corporation though intent on making money so whatever...

 

[Arzachel]

I think people should just buy what they want at the price they are willing to pay. I never saw a problem with an alternative that can be similar value but less money.

This hobby isn't cheap and people tend to disregard logic instead opting to go on belief when larger sums of money are involved. Suffice to say people hate to see their beliefs being challenged because they don't tend to hold up in reality.

At least this forum actually facilitates some discussion even with the frequent us vs them, the ones that split up the boards in Amd and Nvidia sections manage to turn into circlejerks pretty quickly.

 

[3DVagabond]

Dubious? So you are saying to ignore the graph you just posted, putting near the top of all 28nm cards .

System power is a poor way to measure one component power usage. TPU's direct measurement of the device in question is a far better way of doing it.

Also, with the way voltages fluctuate from one card to another it's getting real difficult to know precisely how much power any particular model is using. I think we have a pretty good idea by now about power usage in general. Efficiency for playing games, Pitcairn>GK104>Tahiti. With Tahiti being well down on the other 2 designs.

Keep in mind as well that measurements can vary quite a bit between different review systems. There are too many variables to just point at one graph and declare it the absolute standard. Taken overall they do present an accurate picture.

 

[Magic Carpet]

After looking over several different site's reviews and power draws for various cards, I think by far the best way to measure power draw is by measuring the entire system. If card X draws 20 less watts of power than card Y, but card X in turn forces the rest of the system to draw 18 more watts than if the system had card Y, then the for all intents and purposes card X is NOT more efficient. The total power being sucked out of the outlet, that which shows up on the electric bill, is what counts. No one plays a computer game with the CPU, PSU, hard drive, and ram turned off.

I disagree. This would create a complete mess. It's better (as TPU does) to measure a specific component (in this case video card), so system builders can quickly meet a specific box power/thermal requirements. The proposed accuracy is good enough.

All video card results were obtained on this exact system with exactly the same configuration.

 

[RussianSensation]

You can't use AT's power consumption numbers to accurately isolate for GPU power consumption without looking at power supply inefficiency.

1) They are at the wall (do not account for PSU inefficiency. What if I have a Bronze and you have a Platinum PSU)? The more inefficient your PSU is, the larger the delta between the worst and the least offender. But real world power consumption between the 2 GPUs didn't change. The more efficient your PSU is, the lower the delta is between the worst and the least offender. This is mathematics. Unless they account for PSU inefficiency, we are not getting the real power usage of the components. What we are getting is the real power consumption of a particular system with a particular PSU efficiency in that test bed.

2) They do not isolate for GPU specifically. Since PSU efficiency is less than 100%, right away the delta between 2 GPUs is less than their #s show. This is mathematics.

However, by far the most obvious inconsistency is using HD7950B reference card @ 1.25V for the GPU to represent power consumption usage. It has been stated multiple times on our forums that HD7950 after-market cards do not use 1.25V BIOS:

Sapphire Vapor-X @ 0.97V for 950mhz GPU clock
http://www.techpowerup.com/reviews/Sapphire/HD_7950_Vapor-X/32.html

I am pretty sure 50W extra use over the 660Ti is not a big deal when the performance for $330 is like this:

Dubious? So you are saying to ignore the graph you just posted, putting near the top of all 28nm cards

I would take it a lot more seriously if you actually meant what you are saying. Right now I find it extremely ironic that a gamer with GTX460s in SLI is complaining about the power usage of a reference 7950 B card (that's not representative of after-market 7950s to begin with) and yet a single GTX460 @ 875mhz uses as much power as an overclocked HD7970!!!!!!!

GTX460 @ 900mhz uses as much power as an HD7970 TOXIC 6GB @ 1100mhz!

vs.

vs.

MSI TwinFrozr III 7950
1025mhz / 1775 MHz @ 1.093V : 167W
1200 / 1775 MHz @ 1.275V : 251W

@ 1025mhz, HD7950 already outperform GTX670 despite using just 167W

vs.

Asus DCUII 7970
1150 / 1775 MHz @ 1.174V : 237W

vs.

MSI HD7970 Lightning
1100 / 1775 MHz @ 1.174V : 225W

That's an epic post right there trying to claim that HD7900 generation is a power hog and using 2x GTX460s @ 875mhz in SLI...I don't even know what to say to that. Would you please explain to everyone in this thread why then you didn't buy the HD6850 or HD5850 if power consumption was so important?

Let me get this straight just so I understand it correctly:

GTX460 @ 900mhz delivers ~ HD5870 level of performance (flagship videocard performance for less $) and becomes a legendary price/performance card, despite using 230-240W of power at load
HD7950 @ 1025mhz delivers > GTX670 performance and uses 167-170W of power and is a power hog, but costs $70-80 less
HD7950 @ 1150-1200mhz delivers > GTX680/7970 GE performance, uses 240-250W of power and is a power hog, overpriced, etc.

GTX460 @ 875mhz is awesome though but uses 236W. :thumbsup:

 

[3DVagabond]

they are scrambling to keep up and it's obvious

something a company in desperation mode does

Nothing desperate. Just business as usual. Capitalism at work.

You can't use AT's power consumption numbers with accuracy since:

1) They are at the wall (do not account for PSU inefficiency. What if I have a Bronze and you have a Platinum PSU)? The more inefficient your PSU is, the larger the delta between the worst and the least offender. The more efficient your PSU is, the lower is the delta between the worst and the least offender. This is mathematics. Unless they account for PSU inefficiency, we are not getting the real PSU consumption of the components. What we are getting is the real power consumption of a particular system with a particular PSU efficiency in that test bed.

2) They do not isolate for GPU specifically. Since PSU efficiency is less than 100%, right away the delta between 2 GPUs is less than their #s show.

The most obvious inconsistency is using HD7950B reference card @ 1.25V for the GPU to represent power consumption usage. It has been stated at least 10x on our forums that HD7950 after-market cards do not use 1.25V BIOS:

Sapphire Vapor-X @ 0.97V for 950mhz GPU clock
http://www.techpowerup.com/reviews/Sapphire/HD_7950_Vapor-X/32.html

I am pretty sure 50W extra use over the 660Ti is not a big deal when the performance for $330 is like this:

The overclocking results listed in this section were achieved with the default fan and voltage settings as defined in the VGA BIOS.

That O/C'd figure was stock volts and stock fan profile! When this gen was first released we were concerned that the performance increase was too marginal. Now that the process has matured though we are seeing harvested chips perform at the same level as the dual GPU cards of the last generation. Throw in GCN's improved compute performance and 28nm is actually quite impressive.

Pricing is still a bit on the high side, IMO. Even for the AMD cards. Get the 7970 to ~$350 - $370 and let the rest fall in around that price and we'll be where we should be perf/$.

 

[RussianSensation]

That O/C'd figure was stock volts and stock fan profile!

But I thought you need 1.25V for a 7950 card to get 925mhz GPU boost because Reference 7950B card is what enthusiasts use to gauge how much power they will use @ load and is reprentative of what overclocking voltages are required for after-market 7950s to reach those speeds?

Only GTX460 OC is allowed to be an amazing price/performance card. Even though HD7950 OC ~ GTX670 OC for less $, this doesn't count because the 7950 would use 500W more power. Price/performance and OCing only matter if NV makes the card.

 

[notty22]

Once again a thread has turned in to a rambling strewn together mish mash of arguments and assertions that hardly make any sense.
I have 40nm cards, and I was responding to a comment on the gtx 660ti, a card this thread was about. All this other nonsense is barely coherent.

 

[SirPauly]

But I thought you need 1.25V for a 7950 card to get 925mhz GPU boost because Reference 7950B card is what enthusiasts use to gauge how much power they will use @ load and is reprentative of what overclocking voltages are required for after-market 7950s to reach those speeds?

But it is AMD after all that feels the 1.25v is needed though over-all. They can't afford to cherry pick the cores over-all, imho!

 

[3DVagabond]

Once again a thread has turned in to a rambling strewn together mish mash of arguments and assertions that hardly make any sense.
I have 40nm cards, and I was responding to a comment on the gtx 660ti, a card this thread was about. All this other nonsense is barely coherent.

Agreed. It's simple, the 660ti isn't good value at ~$300 and nVidia is lowering the price. Somehow that logic escapes some people. It's really quite simple.

Hopefully, AMD will be pressured to follow.

 

[sze5003]

I'm just going to watch as they will also need to drop the price of the 670 and 680. I don't know if AMD will also drop prices more on the 7970 series too. I don't think they would do that.

 

[3DVagabond]

I'm just going to watch as they will also need to drop the price of the 670 and 680. I don't know if AMD will also drop prices more on the 7970 series too. I don't think they would do that.

I was wondering that myself. Typically the top nVidia card is isolated from perf/$, but it's usually the fastest. That's not the case now. Although I believe a lot of people still think it's the fastest and that might be enough to protect it's $500 price point.

 

[RussianSensation]

But it is AMD after all that feels the 1.25v is needed though over-all.

False. Only needed for 7950 Reference B cards that get sorted the worst 7950 chips on the market. If you happen to buy 1 such card out of 20 other 7950 cards, it's your own fault for not doing research. Plus, if you actually intend to overclock the 7950, you are not buying this reference card since the cooler is inadequate. Would you overclock a Core i5 2500K to 4.5ghz on the stock Intel cooler or buy a Cooler Master 212+, Noctua NH-D14, etc.? So why are you talking about the reference design 7950 card? Please see below the type of heatsink it has.

They can't afford to cherry pick the cores over-all, imho!

Many AMD's AIB partners on Newegg can and they do. Windforce 3x 7950 900mhz, PowerColor PCS+ 7950, MSI TF3, Sapphire Dual-X 950mhz are all pre-binned with higher quality 7950 chips. None of those cards uses the 7950 B Bios either despite 880-950mhz GPU clocks (unless it says so specifically on the SKU --> like so vs. without). All those crappy 50-60% ASIC 7950s go into those reference 7950B cards. Buy an MSI TF3 7950 right now, and if you personally get 1.25V out of the box in GPU-Z VDCC or HWInfo64 @ 99% GPU load, I'll buy you a game on Steam. If you get lower, you'll buy me a Steam game. Friendly advice, don't take the bet because this is not possible since the top ASIC quartile of 7950 chips go into the TF3 6+8 pin 7970 PCB edition. :thumbsup:

This has been discussed for 2 months in every 7950 overclocking owner's thread and has been 100% ignored, no matter how many times we tell you guys. Remember how all GTX470/480 cards supposedly hit 90-92*C because HardOCP showed it to be true? My 470 card hit 760mhz at barely broke 81*C. Why? Because I researched what I need to do to lower temperatures with a TIM swap, high flow bracket and get a proper case for sufficient airflow so my Fermi cards don't sound like jet engines.

If you research the card you are buying for overclocking, you don't get a 7950 card with 1.25V bios on a reference cooler. It's simple unless you want 90*C temperatures.

If you happen to "accidentally" miss the Gigabyte Windforce 3x, Sapphire Dual-X 950mhz, Sapphire Vapor-x 7950, MSI TF3 7950, HIS IceQ 7950 and buy this instead in hopes of getting 1100mhz, it's your fault.

vs.

http://www.techpowerup.com/reviews/Sapphire/HD_7950_Vapor-X/4.html

or

http://www.techpowerup.com/reviews/Sapphire/HD_7950_Flex/4.html

 

[sze5003]

I was wondering that myself. Typically the top nVidia card is isolated from perf/$, but it's usually the fastest. That's not the case now. Although I believe a lot of people still think it's the fastest and that might be enough to protect it's $500 price point.

For some people that at least have a card now its not bad waiting at all. For me, I returned the HIS 7970 because it was horrible with coil whine even on the desktop so no card for me and the odd thing is just as i am about to buy something, something like this that requires time happens lol.

 

[3DVagabond]

For some people that at least have a card now its not bad waiting at all. For me, I returned the HIS 7970 because it was horrible with coil whine even on the desktop so no card for me and the odd thing is just as i am about to buy something, something like this that requires time happens lol.

Get a card that has solid chokes. 7970 DCII or Sapphire Vapor X I believe do. Double check, don't just take my word for it. I'm working from memory. Both of those cards take up more than 2 slots though.

Another consideration is the PSU. Some of them will whine when pushed hard. Or, they will make the card whine because they have poor filtering.

 

[sze5003]

Get a card that has solid chokes. 7970 DCII or Sapphire Vapor X I believe do. Double check, don't just take my word for it. I'm working from memory. Both of those cards take up more than 2 slots though.

Another consideration is the PSU. Some of them will whine when pushed hard. Or, they will make the card whine because they have poor filtering.

Yeh I am about to buy a 7970 Ghz Vapor X. The DC2's didn't look like they reviewed well on Newegg. And they are not the Dc2T anymore. My psu is a Corsair 750w Enthusiast series its usually pretty quiet on its own. Even with the last card pushing heavely, my 560Ti would only make fan noise not squeeling.

 

[SirPauly]

False. Only needed for 7950 Reference B cards that get sorted the worst 7950 chips on the market. If you happen to buy 1 such card out of 20 other 7950 cards, it's your own fault for not doing research. Plus, if you actually intend to overclock the 7950, you are not buying this reference card since the cooler is inadequate. Would you overclock a Core i5 2500K to 4.5ghz on the stock Intel cooler or buy a Cooler Master 212+, Noctua NH-D14, etc.? So why are you talking about the reference design 7950 card? Please see below the type of heatsink it has.



Many AMD's AIB partners on Newegg can and they do. Windforce 3x 7950 900mhz, PowerColor PCS+ 7950, MSI TF3, Sapphire Dual-X 950mhz are all pre-binned with higher quality 7950 chips. None of those cards uses the 7950 B Bios either despite 880-950mhz GPU clocks (unless it says so specifically on the SKU --> like so vs. without). All those crappy 50-60% ASIC 7950s go into those reference 7950B cards. Buy an MSI TF3 7950 right now, and if you personally get 1.25V out of the box in GPU-Z VDCC or HWInfo64 @ 99% GPU load, I'll buy you a game on Steam. If you get lower, you'll buy me a Steam game. Friendly advice, don't take the bet because this is not possible since the top ASIC quartile of 7950 chips go into the TF3 6+8 pin 7970 PCB edition. :thumbsup:

This has been discussed for 2 months in every 7950 overclocking owner's thread and has been 100% ignored, no matter how many times we tell you guys. Remember how all GTX470/480 cards supposedly hit 90-92*C because HardOCP showed it to be true? My 470 card hit 760mhz at barely broke 81*C. Why? Because I researched what I need to do to lower temperatures with a TIM swap, high flow bracket and get a proper case for sufficient airflow so my Fermi cards don't sound like jet engines.

If you research the card you are buying for overclocking, you don't get a 7950 card with 1.25V bios on a reference cooler. It's simple unless you want 90*C temperatures.

If you happen to "accidentally" miss the Gigabyte Windforce 3x, Sapphire Dual-X 950mhz, Sapphire Vapor-x 7950, MSI TF3 7950, HIS IceQ 7950 and buy this instead in hopes of getting 1100mhz, it's your fault.

vs.

http://www.techpowerup.com/reviews/Sapphire/HD_7950_Vapor-X/4.html

or

http://www.techpowerup.com/reviews/Sapphire/HD_7950_Flex/4.html

Amd raised the default clocks and volts -- not me.

 

[blackened23]

Amd raised the default clocks and volts -- not me.

It was you! Get the pitchforks boys! Just kidding

 

[3DVagabond]

Amd raised the default clocks and volts -- not me.

The Sapphire 7950 Vapor -X (Which is a 7950 Boost model) runs at .97v @ 950MHz and O/C's to 1135/1635 with no voltage modification or changes to the stock fan profile.

Nobody's disagreeing that the modified bios that AMD sent to reviewers set the voltage at 1.25v. It was silly of them because nVidia Fanbois will jump all over anything they can and run with it. In actuality it's very hard to buy a card with that voltage specification. There's only one reference 7950B listed on Newegg and it's OoS.

Again, nobody's disagreeing with what you are saying. Just that it's irrelevant.

 

[RussianSensation]

Amd raised the default clocks and volts -- not me.

Can you please not quote such large posts?

AMD raised default clocks and volts on 7950B cards. How are you ignoring that AIBs sell 10+ SKUs of non-7950B cards that come undervolted with 880-950mhz GPU clocks? Are you that ignorant? Among those at least 4-5 versions overclock to 1050-1125mhz on 1.175V or less. You tend to ignore those cards that enthusiasts buy and focus on a reference card because taking those other 7950 cards into consideration undermines your entire argument.

This is the most pointless argument for enthusiasts since not a single overclock would buy a reference 7950B card in hopes of getting 1100mhz out of it. I just told you that people looking to buy an overclocked 7950 will buy other models. Then I linked you performance data where MSI TF3 @ 1025mhz > GTX670 and that card used 167W. You still ignored that. Glorious. Ok then, I guess after-market videocards do not count. Next stop, HD7970 GE Reference card is hot and loud and therefore all HD7970 GE cards are hot and loud and run at 1.25V. You and notty22 have been repeating this non-stop how we should only look at reference cards. I am sorry. I didn't know I was forbidden to buy non-reference GPUs. I am sorry that you cannot buy a reference 7970 GE card on the market.....but it's being used over and over and over as evidence of how hot and loud it is.

I find it amusing you won't admit that HD7950 OC = GTX670 OC > GTX660Ti and that there are at least 5 after-market 7950s which do not use 1.25V unless they have GPU Boost BIOS.

Curious, what videocards did you buy from Fall 2009 to March 2012? Where any of them by NV? The reason I ask is because NV hasn't made a single power efficient card vs. the competition in the last 3 years so you must have gone with the AMD solution then, right? Or was Fermi's power consumption worth putting up with because of PhysX? I love it, power consumption didn't matter for 3 years and now 30-40W extra is a major deal-breaker on a 350-400W enthusiast rig with an overclocked i5/i7s and overclocked GPUs. Keep this up. I need a laugh on a Sunday night. :thumbsup:

It's pretty funny how you and notty22 continue to use 1 particular reference design of 7950/7970/7970 GE to represent the entire HD7900 series. I was not aware that after-market 7900 series cards do not exist but after-market 660Ti cards do, especially when they are GPU boosting to 1200mhz in launch reviews against the worst 7950 reference card on the market.

For example this 7970 GE Visiontek card that gets lower power consumption than a reference 680 doesn't NOT exist! LMAO!

How about this, you go out and buy an MSI TF3 7950 6+8 PIN card from Newegg since you are so confident 7950s have 1.25V bioses. If it has GPU-Z voltage / HWInfo64 @ 1.25V at 99% GPU load, I'll send you the $ for the card. If you lose, you donate the card to someone on this forum who needs a new GPU upgrade from this thread.

Last 2-3 NV generations GPGPU compute mattered, overclocking mattered (GTX460/470), performance/$ mattered (8800GT, GTX460), power consumption didn't matter (all GTX400/500 series). This round compute is worthless and performance/watt is the most important metric, overclocking is blasphemy and enthusiasts should forget about it without raising power consumption to nuclear reactor levels.

At least if you are going to pretend to be objective, be consistent in your message.

 

[Hypertag]

Why? Because I researched what I need to do to lower temperatures with a TIM swap, high flow bracket and get a proper case for sufficient airflow so my Fermi cards don't sound like jet engines.

Now come on, that is just stupid. Is the 6990 a quiet card now because you can make it quiet if you buy a 3rd party cooler? You love to throw around the argument "no user buying a x dollar graphics card doesn't spend hours fine-tuning overclocking setting!". Now that claim is obviously inaccurate, but you have some point. There is a reasonable percentage of users that overclock their cards.

How many users buying a x dollar graphics card do you honestly believe are going to remove the stock GPU cooler, delid the damned die, remove the stock TIM, and replace the stock TIM? Just a note, this shouldn't be a high number. We know how many actually do this due to Intel's choice to use the crappiest TIM on the planet with Ivy Bridge processors. If this practice is as "common" for graphics cards as you are claiming, then why isn't this widespread with the 3770ks and 3570ks? The answer is that this is NOT a common practice, or even an uncommon practice. It is an exceptionally rare practice, and it is beyond ludicrious to expect users to replace stock TIM inside a gpu die.




This is what the 7970 GHz edition will always be remembered as, no matter how much cherry picking you do.

 

[tviceman]

System power is a poor way to measure one component power usage. TPU's direct measurement of the device in question is a far better way of doing it.
Also, with the way voltages fluctuate from one card to another it's getting real difficult to know precisely how much power any particular model is using. I think we have a pretty good idea by now about power usage in general. Efficiency for playing games, Pitcairn>GK104>Tahiti. With Tahiti being well down on the other 2 designs.

Keep in mind as well that measurements can vary quite a bit between different review systems. There are too many variables to just point at one graph and declare it the absolute standard. Taken overall they do present an accurate picture.

I disagree. This would create a complete mess. It's better (as TPU does) to measure a specific component (in this case video card), so system builders can quickly meet a specific box power/thermal requirements. The proposed accuracy is good enough.

snip

You all completely skipped over the point I was trying to make. If X gpu is more efficient when isolated, that's fine and dandy and neat to point out, but if X gpu causes the rest of the system to work harder and draw more power than the less efficient Y gpu then in my opinion it's just as important to have that information and to measure the total power draw of the system. Total system power is what shows up on the electric bill, so in my opinion that trumps anything when it comes to measuring power and efficiency.

 

[3DVagabond]

You all completely skipped over the point I was trying to make. If X gpu is more efficient when isolated, that's fine and dandy and neat to point out, but if X gpu causes the rest of the system to work harder and draw more power than the less efficient Y gpu then in my opinion it's just as important to have that information and to measure the total power draw of the system. Total system power is what shows up on the electric bill, so in my opinion that trumps anything when it comes to measuring power and efficiency.

Do you really think that a video card makes any appreciable contribution to system power draw besides the power it uses directly? Is there any testing that's ever been done that shows that?

 

[RussianSensation]

Now come on, that is just stupid. Is the 6990 a quiet card now because you can make it quiet if you buy a 3rd party cooler? You love to throw around the argument "no user buying a x dollar graphics card doesn't spend hours fine-tuning overclocking setting!". Now that claim is obviously inaccurate, but you have some point. There is a reasonable percentage of users that overclock their cards.

I never said a reference HD6990 is quiet. Could you have purchased a non-reference HD6990? I don't remember such card. Therefore, a reference 6990 will forever be considered a loud videocard out of the box. That's a fair statement. Is HD7970 GE a loud card? Yes, but only in Review sample reference design form. The reality is that a 7970GE card is not loud or hot since you cannot buy a reference 7970 GE in retail. Feel free to find me one. I am all ears. Furthermore, ignoring after-market 7950 cards (or any after-market videocards) is supporting a position that no alternatives are present in the retail channel. Since this is not true, your position is only theoretical in nature.

Our forum is not BestBuy, and therefore overclocking and fine-tuning is a consideration for many enthusiasts. This has applied to Celeron 300A, Q6600, Core i5 2500k. If for example you choose to spend $500-1000 on high-end CPUs as well, and do not overclock, we can all respect such a position. However, you cannot disregard overclocking and how it can be beneficial. That's how we save $ and it's a fun hobby. If overclocking is not your thing, please feel free to you ignore it and focus on whatever metrics you choose. Since this is an enthusiast forum where we do overclock our GPUs and CPUs, you implications against overclocking are actually against many of the wishes of this forum's members. No one forces you to overclock but then you have ignored other factors discussed in this thread related to GTX660Ti's price/performance against HD7870 and even against the GTX670. If you have an aversion for overclocking, you still have no provided any reasonable explanation for why a GTX660Ti would not benefit from a $20 price drop and how that is not beneficial to the consumers.

How many users buying a x dollar graphics card do you honestly believe are going to remove the stock GPU cooler, delid the damned die, remove the stock TIM, and replace the stock TIM? Just a note, this shouldn't be a high number.

You don't need to do any of these things if you buy after-market SKUs of 670/680/7950/7970 cards. You ignored the existence of such products for reasons unknown. Forgive me please for assuming that enthusiasts who overclock on air to extract additional performance, full well knowing the risks involved, would not consider buying a reference 7950 card. Furthermore, forgive me please for assuming that enthusiast who aspire to have a quiet system would be required to buy a blower design fan reference GPU and subject themselves to unnecessary noise and high temperatures when overclocking.

Additionally, a prospective buyer cannot buy a reference 7970 GE card at all. As such, linking the noise levels for a reference HD7970 GE card as evidence of how "People will remember the HD7970 GE forever" is ignorant. Your reasoning is not sound since you are enforcing 2 constraints which do no exist:

1) The use of a reference design blower card is not a requirement for building an enthusiast gaming system, and it is not a requirement for overclocking in an enthusiast gaming PC;
2) Using a unicorn/mythical 7970 GE reference blower card that was used in reviews ONLY a review sample (as stated explicitly in nearly ever 7970GE launch review) is not representative of the 7970 GE SKU in retail that consumers will actually purchase.

Based on the above, all that you can say objectively is that a reference designed 7970 GE blower card is hot and loud. That is true but by extension you cannot say anything of this nature about retail HD7970 GE series since they do not exhibit such properties or characteristics.

Real world data reveals that it cannot be true that after-market 7950/7970/670/680 need be hot and loud. You ignoring the existence of those cards is not my fault, but only a constraint you have superimposed in your world. I find such artificial constraint to be amusing since after-market solutions from both AMD and NV are readily available for sale on the market. You also seem to conveniently ignore that after-market solutions can address shortcomings of the reference designs in terms of heat and noise levels.

While it is accurate to state that a reference blower air cooled GTX480 card ran hot and often loud, that does not preclude the existence of excellent after-market solutions of the GTX480 SKU that have resolved both of these problems:

If only 7950B/7970 GE reference air cooled cards had been on sale, as was the case for the air cooled reference HD6990s, your statement would have been valid. Given that it is impossible to buy a reference blower 7970 GE card in retail, the position which you present is illogical. On the contrary, the only logical way to think about the HD7970 GE SKU is to consider after-market solutions.

This is what the 7970 GHz edition will always be remembered as, no matter how much cherry picking you do.

I am not cherry-picking anything, actually you are. You cannot buy a reference 7970 GE card in retail. Only someone who is ignored will keep linking noise levels for a 7970 GE reference blower card. No matter how much you claim this to be true, such card does not exist in retail.

Since you have earlier opposed that enthusiasts are going to tinker with their components since it is outside of the scope of the average user's ability, you have eliminated the option of flagshing an HD7970 reference blower with the GE Bios. Thus, you cannot use this as an example to support your view as that contradicts your initial position, i.e., that hardly many people have the ability to tinker with parts by using such 'advanced' enthusiast techniques. Based on your imposed constraints, we must now look at only real world SKUs of the said 7970 GE videocard:

As it happens, the retail 7970 GE SKUs use after-market coolers and are therefore quiet:

http://www.newegg.com/Product/Produc...82E16814202001
http://www.newegg.com/Product/Produc...82E16814125439
http://www.newegg.com/Product/Produc...82E16814129265

In addition, please do tell every 7970 after-market owner that after-market 7970 cards cannot be bought in retail and that all of them use 1.25V voltage out of the box.

 

[Hypertag]

I never said a reference HD6990 is quiet.

You acclaimed a GTX 470 was quiet because after all of that work you got it to 75 Celsius on load (is that supposed to be good for a heavily modified card?). It's just a false point you are using to justify your purchase of triple SLI 470s. The reality is simpler. The 470 was a crappy card.