AMD 10.12 vs 11.1 vs 11.1a tested ... interesting read

Page 8 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Wreckage

Banned
Jul 1, 2005
5,529
0
0
You just cherry picked one benchmark link, I posted more than 3 reviews from different sites.

Actually I posted 2 links. Stop posting misinformation. Just because my links prove you wrong does not constitute "cherry picking".

Xbit tested with Aliens Vs Predator a newer DX11 game. So I guess you are just validating my links.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,110
1,260
126
istockphoto_1732046-cherry-picker.jpg


Here's a good one to keep on hand for the bizarre posts. :)
 

Skurge

Diamond Member
Aug 17, 2009
5,195
1
71
Wow, some people here don't know much about electrical engineering huh.

Wrekage, that post wasn't for you, it was for copenhagen69 who asked how the other member knew the extra 2 pins where ground wires.

Apopin, Im sure you know a ground wire can't provide more current, those are basics. Even in your last post, you quoted someone saying it was just for reduced resistance. So a 6pin can provide as much power as an 8pin, it would just get hotter.
AMD added it just so they could comply with PCI-E specs, as the card is rated for 250W with powertune set to +20%.

When GTX570s are overclocked to be as fast as 580s and use as much, if not more power, I don't here people going "oh, it going to fail cause it doesn't have an 8pin". It fails for other reasons.
Your PCI-E cables would melt long before a 6950/570 fails from using a 6pin instead of an 8pin.
Now that I think about it, how on earth will a card fail if it isn't getting enough power? Thats just moronic reasoning.
 

load81

Member
Jan 21, 2011
105
0
0
When GTX570s are overclocked to be as fast as 580s and use as much, if not more power, I don't here people going "oh, it going to fail cause it doesn't have an 8pin". It fails for other reasons.

A 570 needs to be clocked to 900 core to match a stock 580 in benchmarks. Mine does that at 1.05 volts. Check this thread plenty of others do also and have no problems http://www.overclock.net/nvidia/891918-gtx-570-overclocking-thread.html The people who fried their cards unlocked the voltage limit something most people don't do anyways. Also there are people who are at 1.2 volts on a 570 and their cards didnt fry. No body knows why those few cards died could of been a faulty batch of vrms. All it takes to brick a 6950 is a 10-15% performance increase to a 6970 :biggrin:
 
Last edited:

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
Apopin, Im sure you know a ground wire can't provide more current, those are basics. Even in your last post, you quoted someone saying it was just for reduced resistance. So a 6pin can provide as much power as an 8pin, it would just get hotter.
AMD added it just so they could comply with PCI-E specs, as the card is rated for 250W with powertune set to +20%.

When GTX570s are overclocked to be as fast as 580s and use as much, if not more power, I don't here people going "oh, it going to fail cause it doesn't have an 8pin". It fails for other reasons.
Your PCI-E cables would melt long before a 6950/570 fails from using a 6pin instead of an 8pin.
Now that I think about it, how on earth will a card fail if it isn't getting enough power? Thats just moronic reasoning.
Since you are addressing me; i have been attempting to point out that there is a REASON for the "extra" two pins - they are not "useless" as MrK6 would have us believe.

In fact, you hit the nail on the head. WithOUT the grounds, it gets HOTTER ,,,, and the HD 6950 does not have even have the pins that the HD 6970 does. It is for stability .. and thus .. longer life. The HD 6950 is not engineered to run as a HD 6970(period).
:cool:

Please don't mention 570s running at GTX 580 speeds. The 570 is not flashed into a GTX 580 as the HD 6950 is flashed into a 6970. :p
 
Last edited:

evolucion8

Platinum Member
Jun 17, 2005
2,867
3
81
Actually I posted 2 links. Stop posting misinformation. Just because my links prove you wrong does not constitute "cherry picking".

Xbit tested with Aliens Vs Predator a newer DX11 game. So I guess you are just validating my links.

Nope, you are the one who's been banned several times by dirty antics, deceiving and false posts, not me. I posted lots of links for your delightment and yet, you just cherry pick two sites with questionable results. This Anandtech forum, and hence, I posted a link from Anandtech which it has been known to be reliable, there's no need to post links from unreputables sites like you do.

Look, I found more links!! Its like 6 or more against 2 of yours, which clearly shows something.

http://www.hardwareheaven.com/revie...50-graphics-card-review-power-temp-noise.html

"In terms of overall experience from the cards the reference model findings still apply. The Radeons use less power, the GTX 570 runs quieter"
 

Skurge

Diamond Member
Aug 17, 2009
5,195
1
71
A 570 needs to be clocked to 900 core to match a stock 580 in benchmarks. Mine does that at 1.05 volts. Check this thread plenty of others do also and have no problems http://www.overclock.net/nvidia/891918-gtx-570-overclocking-thread.html The people who fried their cards unlocked the voltage limit something most people don't do anyways. Also there are people who are at 1.2 volts on a 570 and their cards didnt fry. No body knows why those few cards died could of been a faulty batch of vrms. All it takes to brick a 6950 is a 10-15% performance increase to a 6970 :biggrin:

I was just using that to show that there are a bunch of other things that would cause a card to fail. I seriously doubt insufficient power is one of them.

Also, wasn't on of the reasons the lack of temp monitoring on the vrms of the 570?
 

MrK6

Diamond Member
Aug 9, 2004
4,458
4
81
You really don't get it - at all. According to your posts, you would really have us believe that the 2 extra pins are "useless"
... and now you are quoting forum posters on another forum.
o_O

However, they actually got it right even though you didn't:
Where did I ever say they were "useless?" Go quote it. Stop putting words in my mouth in some lame attempt to try to save face after you were so clearly proven wrong. Wreckage tried to push some FUD and you backed him on it, but you're both wrong and haven't the first clue about electrical engineering.

Wow, some people here don't know much about electrical engineering huh.

Wrekage, that post wasn't for you, it was for copenhagen69 who asked how the other member knew the extra 2 pins where ground wires.

Apopin, Im sure you know a ground wire can't provide more current, those are basics. Even in your last post, you quoted someone saying it was just for reduced resistance. So a 6pin can provide as much power as an 8pin, it would just get hotter.
AMD added it just so they could comply with PCI-E specs, as the card is rated for 250W with powertune set to +20%.

When GTX570s are overclocked to be as fast as 580s and use as much, if not more power, I don't here people going "oh, it going to fail cause it doesn't have an 8pin". It fails for other reasons.
Your PCI-E cables would melt long before a 6950/570 fails from using a 6pin instead of an 8pin.
Now that I think about it, how on earth will a card fail if it isn't getting enough power? Thats just moronic reasoning.

Since you are addressing me; i have been attempting to point out that there is a REASON for the "extra" two pins - they are not "useless" as MrK6 would have us believe.
Still haven't quoted it. Looks like you've resorted to outright lying now.

In fact, you hit the nail on the head. WithOUT the grounds, it gets HOTTER ,,,, and the HD 6950 does not have even have the pins that the HD 6970 does. It is for stability .. and thus .. longer life. The HD 6950 is not engineered to run as a HD 6970(period).
:cool:
Where's the proof that this is what's causing the failures? Since you've blasted everyone that disagrees with you, one would think you have some definite proof.

Please don't mention 570s running at GTX 580 speeds. The 570 is not flashed into a GTX 580 as the HD 6950 is flashed into a 6970. :p
It still draws more power over the lines and the same principles apply, not that it matters since you don't seem to understand them anyway.

Sorry, you're not getting out of this one. Note to self - alienbabeltech is a bad joke :rolleyes:
 

evolucion8

Platinum Member
Jun 17, 2005
2,867
3
81
A 570 needs to be clocked to 900 core to match a stock 580 in benchmarks. Mine does that at 1.05 volts. Check this thread plenty of others do also and have no problems http://www.overclock.net/nvidia/891918-gtx-570-overclocking-thread.html The people who fried their cards unlocked the voltage limit something most people don't do anyways. Also there are people who are at 1.2 volts on a 570 and their cards didnt fry. No body knows why those few cards died could of been a faulty batch of vrms. All it takes to brick a 6950 is a 10-15% performance increase to a 6970 :biggrin:

Insanity Check. . . .

Kinda weird that you feel good about bricking an HD 6950, a lifeless product with no feelings or soul, from a company who doesn't even know who you are. Akward, specially when such cards comes with a Dual BIOS switch, only people like you can really "brick" a card.
 

load81

Member
Jan 21, 2011
105
0
0
Insanity Check. . . .

Kinda weird that you feel good about bricking an HD 6950, a lifeless product with no feelings or soul, from a company who doesn't even know who you are. Akward, specially when such cards comes with a Dual BIOS switch, only people like you can really "brick" a card.

lol why get mad and insult me? Jeeze relax a little you dont even know me or what I do. I dont feel good about bricking a card I'm not the one who brought it up. I smiled because it can happen with any card not just the 570/6950. I dont think I'm the one who needs an insanity check whoa
 

evolucion8

Platinum Member
Jun 17, 2005
2,867
3
81
lol why get mad and insult me? Jeeze relax a little you dont even know me or what I do. I dont feel good about bricking a card I'm not the one who brought it up. I smiled because it can happen with any card not just the 570/6950. I dont think I'm the one who needs an insanity check whoa

Insulting you? There's any word or adjective in my post like calling you names and stuff? Nope. So it isn't. So relax and enjoy the thread! It was a misunderstanding anyways.

Have a cookie:
cookie-monster.jpg
 

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
Talk about your personal attacks. You are the one that is dishing them out non-stop.

You have provided no proof of ANYTHING you have said.

The 6950 is not engineered to run at the higher specs of the 6970, but it isn't due to it lacking an 8-pin power connector.
Give an explanation a try at least. :p

You also said this bit of self-contradictory nonsense and refused to clarify anything further:
The 8-pin connector, which only adds two ground wires, doesn't somehow magically enable the same 12V lines to somehow safely carry more power - the truth is they already can.
The truth is that the "extra" 2 pins are designed to ensure the stability of the higher power requirements of the HD 6970 over the HD 6950.



Where did I ever say they were "useless?" Go quote it. Stop putting words in my mouth in some lame attempt to try to save face after you were so clearly proven wrong. Wreckage tried to push some FUD and you backed him on it, but you're both wrong and haven't the first clue about electrical engineering.



Still haven't quoted it. Looks like you've resorted to outright lying now.

Where's the proof that this is what's causing the failures? Since you've blasted everyone that disagrees with you, one would think you have some definite proof.

It still draws more power over the lines and the same principles apply, not that it matters since you don't seem to understand them anyway.

Sorry, you're not getting out of this one. Note to self - alienbabeltech is a bad joke :rolleyes:
 
Last edited:

Outrage

Senior member
Oct 9, 1999
217
1
0
The truth is that the "extra" 2 pins are designed to ensure the stability of the higher power requirements of the HD 6970 over the HD 6950.

The truth is that the 8 pin connector on the 6970 is there to comply with pci sig.
 

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
The truth is that the 8 pin connector on the 6970 is there to comply with pci sig.
So you are suggesting that the 8-pin PCIe connector is a ridiculous invention that serves no useful purpose whatsoever over the 6-pin connector and that AMD and Nvidia's engineers are stupid for using it at all on their high-end cards?
:rolleyes:
 

Outrage

Senior member
Oct 9, 1999
217
1
0
So you are suggesting that the 8-pin PCIe connector is a ridiculous invention that serves no useful purpose whatsoever over the 6-pin connector and that AMD and Nvidia's engineers are stupid for using it at all on their high-end cards?
:rolleyes:

Amd and nvidias enginers dont have anything to do with this, you even suggesting that just shows how out of the loop you are.
 

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
Amd and nvidias enginers dont have anything to do with this, you even suggesting that just shows how out of the loop you are.
i understand it quite well and i also realize that i am wasting my time here.

AMD and Nvidia's engineers have to meet a PCIe specification. What you are saying is that the specification itself is stupid and that you and a few vocal forum posters know better than the engineers who designed the PCIe spec


Goodbye
 

load81

Member
Jan 21, 2011
105
0
0
Insulting you? There's any word or adjective in my post like calling you names and stuff? Nope. So it isn't.

Ok so I guess the insanity check was for you. I thought you were calling me insane. And the "people like me" thing must of ment someone who flashes a bios instead of someone who is more safe and just overclocks it for the same gains. Gotcha o_O
 

minddripper

Junior Member
Feb 18, 2008
6
0
0
From MrK6's link (http://www.tomshardware.com/forum/274631-28-power-spec-power-plug)

Johnny & the group,

Have learned several interesting facts about the PCI-E plugs and power after researching it this past week, that is worth sharing

The PCI-E connecter has 6 pins. They are not all used for power. For the 6 pin connector, Pins 1 & 3 are 12V & each can carry 8 Amps. Pin 2 by spec is not connected, although some PSU manufacturers do add a 12V line there. Pins 4 & 6 are Com return lines. Pin5 is Com for sensing. Using 2 lines, you get 12V*8A*2= 192Watts, much over the required 75Watts. With an 8 pin PCI-E connector, 2 Com lines are added (4&8) not a 12V & Com. There, Pins 1,2,3 are 12V, Pin 4 is a Com for the 8 pin connector sensing, Pin 5,7,8 are Com return lines, & pin 6 is for the 6 pin connector sensing. Using that config, 12V*8Amp*3=288 Watts, much over the required 150Watts.

Since Mar 2005, the molex pins are required to be "HCS" rather than "Std", which each carry a max of 11Amps. So properly made, an 8 pin PCI-e can supply 12V*11Amps*3lines=396Watts of power for the graphic cards.

Ther aren't 4 12V pins on the 8 pin connector, only 3, and 2 leads are used for sensing the connector type.

The other limiting factor of course is the width of the PCB lines on the graphics card, each typically carrying 1-2 Amps. That would also determine how much power it could carry to the graphics electronics.
Also worth noting Scott Meuller in his latest 19th Ed of "Upgrading and Repairing PCs" has this information incorrectly listed.

Hope that answers most of the questions regarding this unusually engineered connector.

John VK

So according to this the 6 pin connector is 2x12V and 2xGround with one wire used for detecting the plug, while the 8 pin is 3x12V and 3xGround with two wires used for detecting 6 pin or 8 pin plugs. I don't have many power supplies with 6 pin PCIe connectors, but all of mine have three 12V wires, so I found this strange. However, it seems very obvious that the specs on the PCIe connectors are extremely conservative. According to the quoted post, the max rated current for the 8 pin connector (12.5A for 150W) is barely over the max rating of a single HCS molex pin (11A).

Having more wires and leads reduces the resistance and heat but is the heat significant? The resistance of a 2ft long 18 gauge copper wire is 0.01277 ohms. For example, if three wires are supplying 12.5A and three are returning 12.5A, 1.33W of heat are generated by the wires if the cable is 2ft long. If there are only two wires supplying and two returning the heat goes up to 2W. The amount of heat generated by the connector itself should be much, much smaller.

Another thing is that 8 pin connectors are designed to accept 6 pin plugs and can detect the difference between the two. Video cards can be designed to operate differently depending on the type of plug connected, refuse to operate at all, or operate the exact same way. How does the 6970 react to a 6 pin plug? Does a flashed 6950 know that a 6 pin plug is connected? Does it make a difference?

From what little I've read it doesn't look like the connectors themselves really matter. It is clear that the 6 pin connector can handle more than 150W. It looks to me that any limiting factors should exist on the card after the power connector, and those details probably aren't available.
 

MrK6

Diamond Member
Aug 9, 2004
4,458
4
81
i understand it quite well and i also realize that i am wasting my time here.

AMD and Nvidia's engineers have to meet a PCIe specification. What you are saying is that the specification itself is stupid and that you and a few vocal forum posters know better than the engineers who designed the PCIe spec
No, you don't. How many people have to tell you you're wrong before you finally get it?
About time. You should have ducked out gracefully awhile ago instead of dragging this thread down to your level.
From MrK6's link (http://www.tomshardware.com/forum/274631-28-power-spec-power-plug)



So according to this the 6 pin connector is 2x12V and 2xGround with one wire used for detecting the plug, while the 8 pin is 3x12V and 3xGround with two wires used for detecting 6 pin or 8 pin plugs. I don't have many power supplies with 6 pin PCIe connectors, but all of mine have three 12V wires, so I found this strange. However, it seems very obvious that the specs on the PCIe connectors are extremely conservative. According to the quoted post, the max rated current for the 8 pin connector (12.5A for 150W) is barely over the max rating of a single HCS molex pin (11A).

Having more wires and leads reduces the resistance and heat but is the heat significant? The resistance of a 2ft long 18 gauge copper wire is 0.01277 ohms. For example, if three wires are supplying 12.5A and three are returning 12.5A, 1.33W of heat are generated by the wires if the cable is 2ft long. If there are only two wires supplying and two returning the heat goes up to 2W. The amount of heat generated by the connector itself should be much, much smaller.

Another thing is that 8 pin connectors are designed to accept 6 pin plugs and can detect the difference between the two. Video cards can be designed to operate differently depending on the type of plug connected, refuse to operate at all, or operate the exact same way. How does the 6970 react to a 6 pin plug? Does a flashed 6950 know that a 6 pin plug is connected? Does it make a difference?

From what little I've read it doesn't look like the connectors themselves really matter. It is clear that the 6 pin connector can handle more than 150W. It looks to me that any limiting factors should exist on the card after the power connector, and those details probably aren't available.
Good write-up. Going back to the original point, that's why it's ludicrous to say that the absence of an 8-pin connector causes the flashed 6950's to fail, especially considering how very few of them do.
 

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
How many people have to tell you that you are out of your depth? What, engineering student? :p

i doubt that you have the qualifications to make your claims that fly in the face of reason.

You have no clue if the absence of an 8-pin connector has to do with premature failure of flashed 6950 yet you pretend you do. You are attempting to minimize the risk of flashing the 6950 to 6970 with no proof of what you are posting whatsoever.

i want to see proof of what you are saying .. or maybe you should get of the discussiion
:thumbsdown:


No, you don't. How many people have to tell you you're wrong before you finally get it?
About time. You should have ducked out gracefully awhile ago instead of dragging this thread down to your level.
Good write-up. Going back to the original point, that's why it's ludicrous to say that the absence of an 8-pin connector causes the flashed 6950's to fail, especially considering how very few of them do.
 

Wreckage

Banned
Jul 1, 2005
5,529
0
0
Let's just assemble the facts

  • The 6970 has an 8-pin + 6-pin connector
  • The 6-pin connector is designed to handle 75W, while the 8-pin 150W
  • The 6970 uses more power than a 6950
  • Several 6950 users have damaged their cards with the bios hack
  • Several users who have not damaged their cards have complained of other issues (artifacting, or zero success with the flash).

So it's clear that a.)There is risk in using the bios hack and b.) The 6970 is engineered different than the 6950.

EDIT: This should probably in its own thread as it's not relevant to the original topic. Here we go.
http://forums.anandtech.com/showthread.php?t=2142549
 
Last edited:

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
i am still waiting for one of these engineering students to demonstrate proof that the PCIe 2.0 spec and the 8-pin is just a stupid engineering design that serves no useful nor practical purpose over the 6-Pin.

i also asked AMD for clarification on the 6-pin vs. 8-pin

There is definite risk with flashing 6950 to 6970. i understand the risk and i took it for myself. i will not attempt to minimize it- nor overstate it.

Agreed on the thread hijack of an otherwise uninteresting thread. :)
-If AMD replies, i will start a brand new thread on it.
 
Last edited:

MrK6

Diamond Member
Aug 9, 2004
4,458
4
81
Let's just assemble the facts

  • The 6970 has an 8-pin + 6-pin connector
  • The 6-pin connector is designed to handle 75W, while the 8-pin 150W
  • The 6970 uses more power than a 6950
  • Several 6950 users have damaged their cards with the bios hack
  • Several users who have not damaged their cards have complained of other issues (artifacting, or zero success with the flash).

So it's clear that a.)There is risk in using the bios hack and b.) The 6970 is engineered different than the 6950.

EDIT: This should probably in its own thread as it's not relevant to the original topic. Here we go.
http://forums.anandtech.com/showthread.php?t=2142549
All good facts, but they lack significance without context. What is the percentage of 6950->6970 flashes that failed out of all those attempted? Furthermore, what percentage is actually due to some hardware failure (as opposed to user error)? Anyone who overclocks or flashes a video card knows there's a risk involved, so why state the obvious? What point are you trying to make?

i am still waiting for one of these engineering students to demonstrate proof that the PCIe 2.0 spec and the 8-pin is just a stupid engineering design that serves no useful nor practical purpose over the 6-Pin.

i also asked AMD for clarification on the 6-pin vs. 8-pin

There is definite risk with flashing 6950 to 6970. i understand the risk and i took it for myself. i will not attempt to minimize it- nor overstate it.

Agreed on the thread hijack of an otherwise uninteresting thread. :)
-If AMD replies, i will start a brand new thread on it.
I thought you left? Or are you just back to continue twisting the argument to remove the egg from your face? Once again, the argument was not what benefits/improvements the 8-pin PCIe connector offers over the 6-pin connector, it was that the 6950 being fed by two PCIe 6-pin adapters still gets all the power it needs. Learn to read.