X800 VIVO to XTPE kills the card?

Joined
Aug 29, 2004
Messages
21
Hey all,

I've been perusing both here and on the Overclocker's Forums, and a lot of people over there are saying that a bunch of X800 Pros that are flashed to XTPE's are completely dying within like a month or two of the flash.

Now, it seems that all the people saying this have NVidia cards listed in their sig, so I don't really trust them, but I thought I should ask here( I trust the answers on this forum a lot more anyway). Has anyone been having problems with the card working fine after the XTPE flash and then just randomly crapping out after a few weeks/months? I'm still planning to get one this friday(when I can get to a compusa), but I'm a little bit nervous, too. I cant afford to throw away $400. Any help is greatly appreciated. Put my fears to rest.
 
I've been hearing this too but I have yet to actually hear from someone who's actually had this happen to them.

all I've been hearing is a bunch of phrases like "I've heard that alot of the ATi VIVO cards have been dying after xxxxx days..." and stuff like that.

but again, I have yet to actually hear from a real life scenario.

thats not to say it cant or isnt happening cuz anything's possible.
 
While I won't say absolute hardware failure is non-existant, I will say it's uncommon.

In my history of building computers and testing hardware, I have never had a component completely fail out of the blue.

In all likelihood, enabling 4 extra pipelines should only cause artifacting or hardlocks in a worst case scenario, but it should be reversible through a BIOS flash.
 
EVERY SINGLE POST like that I've seen has come from someone who 'heard it' and I've seen like 2 posts on the various forums where it's actually happened, and even in those two cases there were other contributing circumstances. (ie extreme overclocking, fan failure related to ATITool, etc)

The X800 VIVO cards (at least in the case of BBA cards) seem to consist of cards that either did NOT pass ATI's rigorous testing procedures at 530/575 speeds or cards that passed those tests but were needed to fill backordered demand for the Pro cards. These cards are simply flashed with a BIOS that disables 4 of the rendering pipes and lowers the core and memory speed to Pro levels, and then sold as X800 Pro cards. Aside from this difference the card is basically an X800 XT.

Almost ALL of these BBA VIVO cards are able to easily run at XT-PE speeds and well beyond, but you do see the odd card now and then that won't quite reach XTPE speeds. Even in the case that someone gets one of these subpar XT's that were rebranded as Pro VIVO cards, it's not just going to 'die' from being run with all of it's pipes open and at higher than Pro speeds. Artifacting yes, sudden death NO. It's my opinion that any of these 'sudden death' issues that this guy or that guy 'heard about' that are actually true are cases where the card was defective and would have died anyway wether it remained a Pro or not.

My own CompUSA bought X800 Pro will do XT-PE speeds with all pipes open, totally rock solid stable. I can even go quite a bit higher on the core and a tiny bit higher on the memory before I start seeing artifacting. That's one hell of a bang for your buck. :)
 
I'm in the same boat as you ozziegn. Been hearing ppl say "Oh those fail a little while after the flash. A LOT of those modded X800s die right away." but I've yet to hear from anyone who has actually had a card fail outright on them after a successful softmod. I'd appreciate any other input on this, as I feel this is a problem/rumor that needs to be addressed/put to rest.
 
trakslacker said:
I'm in the same boat as you ozziegn. Been hearing ppl say "Oh those fail a little while after the flash. A LOT of those modded X800s die right away." but I've yet to hear from anyone who has actually had a card fail outright on them after a successful softmod. I'd appreciate any other input on this, as I feel this is a problem/rumor that needs to be addressed/put to rest.

thats exactly the reason why I started a thread just for this specific topic at the Overclockers forum and so far I've had zero people come forward and say that this has personally happened to them.

here's my thread.
 
Sapphire Technology has issued a press release warning exactly that - board death might happen if you mod vivo -> xt

http://www.sapphiretech.com/pressrelease/Hack-Bios-Warning.pdf

Sapphire Technology said:
In general there is a large margin for board failure when the XT 16 pipe BIOS is applied to the X800 PRO’s 12 pipe boards. The X800 PRO boards while inherently possessing the 16 pipe pathways are not all capable of maintaining stability with 16 pipes enable. For this reason Sapphire will not be able to support the replacement or repair of any of our products returned with this BIOS hack applied.

Pretty common sense - if the board was truly an x800xtpe, it would have been sold as an x800xtpe, not an x800pro - the demand for x800xtpe is far larger than that of x800pro.
 
tranCendenZ said:
Sapphire Technology has issued a press release warning exactly that - board death might happen if you mod vivo -> xt

http://www.sapphiretech.com/pressrelease/Hack-Bios-Warning.pdf



Pretty common sense - if the board was truly an x800xtpe, it would have been sold as an x800xtpe, not an x800pro.

bah,,,

they're just covering their butts which is normal for any business to do.
 
ozziegn said:
bah,,,

they're just covering their butts which is normal for any business to do.

heh I don't see nvidia ihvs issuing detailed press release warnings about not overclocking GT to Ultra speeds and board failures because of doing so :p
 
tranCendenZ said:
heh I don't see nvidia ihvs issuing detailed press release warnings about not overclocking GT to Ultra speeds and board failures because of doing so :p

no but I'll betcha $100 that if they could somehow prove that a damaged card that was sent to them for replacement from O/Cing issues that they'd be doing the same thing.
 
The card is physically the same before and after you flash to XT bios. I don't quite understand how a bios change would physically alter your hardware besides allowing the extra pipelines to be used. This is all changed in software, hence the term softmod. Whether it's activating pipelines, or overclocking, softmods done right are generally recreatable and reversible. If you have a chip that can run stable at a certain speed, and then put it back to stock, would you expect it to be able to run at that speed again if you wanted? Heck yeah, because the physical attributes of that chip were not changed.

Unless of course you push it above what it can physically handle and then something gets physically damaged. Other than that, rumors of modded VIVOs not working anymore are just rumors. If we listened to manufacturers, no one would overclock because if it voids our warranties then it must be bad. However, if someone would come forward with a messed up VIVO, I will kindly put my foot in my mouth. :D
 
Also with the Ultra and GT issue, no one is saying you will mess up your board because they are NOT the same. If you forget, your GT has cheaper ram. Difference might be slight, but my point stands. Manufacturers are kicking themselves for selling these things as Pros with 1.6 ns ram, when they could have made another $100 if they were XT's. So they are bitter and don't want you to get performance at their expense.
 
newster said:
However, if someone would come forward with a messed up VIVO, I will kindly put my foot in my mouth. :D

read this.

just be sure to wash your foot before you stick it in your mouth. :D
 
ok, that guys card was kinda gettin screwed.

BUT - He was voltmodding, Oc'ing reasonably high and whatnot on stock cooling. He disregarded very bad signs such as crashing game and VPU recover. Obviously something is wrong here. But the rumors I've been reading speak of X800's that flash fine, run fine, and then die w/o warning.

Plus, if the card is having major problems like that, just flash back to the 12 pipes, and RMA that bad boy.
 
Doesn't sound like his card died for flashing it to 16 pipes. When it does, then the foot goes in the mouth :p
 
tranCendenZ said:
Sapphire Technology has issued a press release warning exactly that - board death might happen if you mod vivo -> xt

http://www.sapphiretech.com/pressrelease/Hack-Bios-Warning.pdf

Pretty common sense - if the board was truly an x800xtpe, it would have been sold as an x800xtpe, not an x800pro - the demand for x800xtpe is far larger than that of x800pro.

Or Sapphire would say this to get people to buy their XTs, because they do cost more. Using your logic, if a 6800GT was truly a 6800 Ultra, it would have been sold as a 6800 Ultra, and not a 6800 GT. And yet you continue to push them on people with the promise that they overclock to or past Ultra levels...do you wish for others to ruin their $400 purchase? Overclocking is dangerous, you know. :rolleyes:
 
I'm sorry guys, I'm still ROFLing at the standard tranCendenZ 'Rain on your parade' post ni this thread.... :D
 
truffle00 said:
Or Sapphire would say this to get people to buy their XTs, because they do cost more. Using your logic, if a 6800GT was truly a 6800 Ultra, it would have been sold as a 6800 Ultra, and not a 6800 GT. And yet you continue to push them on people with the promise that they overclock to or past Ultra levels...do you wish for others to ruin their $400 purchase? Overclocking is dangerous, you know. :rolleyes:

I'm not trying to pretend the 6800GT is a 6800U and no way would I ever recommend bios flashing a 6800GT to a 6800U. As I've stated many times before, 6800GT cards usually overclock to Ultra or near-Ultra speeds. However, when people post that they got a less-than-Ultra overclock, I also respond that they should not consider that they got a bum card, because a 6800GT is not a 6800U - anything over 350/1000 is bonus. Just like I'd never consider an x800pro vivo an x800xtpe. Also, enabling 4 pipeplines purposefully disabled *and* bios flashing *and* overclocking on top of that is a hell of a lot more risky than just using a simple overclock utility which is all you need with the 6800GT, especially when an ATI IHV has released a press release that doing all those things can cause board failure. Sapphire doesn't need that press release to sell their XTs, because they have no XTs to sell, and if they did they would be sold without the need of that press release. That's common sense, obviously they released the PR because they don't want to be forced into replacing dead x800pro vivo -> x800xtpe flashed cards - by releasing this PR they "cover their ass" a bit as ozzie pointed out.

Blue Falcon said:
I'm sorry guys, I'm still ROFLing at the standard tranCendenZ 'Rain on your parade' post ni this thread....

Rain on your parade meaning reality? You don't think an official press release from ATI's biggest board partner stating board failure could result from x800pro vivo -> x800xtpe flashes is relevent in a thread titled "X800 VIVO to XTPE kills the card?" which poses this exact question? If you personally want to ignore Sapphire's warnings of board failure, fine, but that doesn't mean others might find their comments useful in determining whether they want to flash or not. The only reason I posted in this thread is because I noticed the thread title and was aware of the press release by Sapphire which addressed this exact issue quite clearly.
 
tranCendenZ said:
Sapphire Technology has issued a press release warning exactly that - board death might happen if you mod vivo -> xt

http://www.sapphiretech.com/pressrelease/Hack-Bios-Warning.pdf



Pretty common sense - if the board was truly an x800xtpe, it would have been sold as an x800xtpe, not an x800pro - the demand for x800xtpe is far larger than that of x800pro.

I agree...people have absolutely no right at all to complain. There have been reports of 6800 GTs that have been OC'd to Ultra speeds crapping out too. ATi is having so much trouble getting XTPE's out the door, there is a reason for that (other than the GDDR3 shortages). Why would they put XTPE capable cores in VIVOs and Pros if they are truely capable of going at XTPE speeds? It makes no sense.

Pros are Pros for a reason
GTs are GTs for a reason

Neither of the above make the grade to run at XTPE and U levels. Ask yourself why if they cant meet demand for higher end cards would they sling em in lower models?
 
p0tempkin said:
While I won't say absolute hardware failure is non-existant, I will say it's uncommon.

In my history of building computers and testing hardware, I have never had a component completely fail out of the blue.

In all likelihood, enabling 4 extra pipelines should only cause artifacting or hardlocks in a worst case scenario, but it should be reversible through a BIOS flash.

exactly what i would say
 
tranCendenZ said:
Rain on your parade meaning reality? You don't think an official press release from ATI's biggest board partner stating board failure could result from x800pro vivo -> x800xtpe flashes is relevent in a thread titled "X800 VIVO to XTPE kills the card?" which poses this exact question? If you personally want to ignore Sapphire's warnings of board failure, fine, but that doesn't mean others might find their comments useful in determining whether they want to flash or not. The only reason I posted in this thread is because I noticed the thread title and was aware of the press release by Sapphire which addressed this exact issue quite clearly.

The press release makes very specific references to the VIVO and the non-VIVO Pro model and I'm pretty sure they were referring to the latter. So that still doesn’t really answer the question of whether people who are modding the VIVO model of the X800 Pro are actually having any issues with the softmod process. I personally haven’t seen anyone post with an issue on the softmod other than low overclocks and I’ve only seen one person with an actual VIVO fail the mod so far, but then again that doesn’t mean people with those issues don’t exist.
 
Some people seem to forget that companies do this all the time -- they underclock a higher-end product so they can sell it cheaper and make more money. There is more demand for the cheaper x800pro. Believe it or not, that is fact -- most people cannot afford an xt-pe. Standard business practice here, to supply cards where they are going to sell. Also, sapphire's cards have not had a good success rate from the outset. If powercolor came out and said it (with their 100% success rate) then I would start to worry (especially since I have a powercolor x800pro vivo). Also, that Sapphire release did not specifically say that vivo to xt-pe mods were failing, they were saying that modding an x800pro to an xt-pe thru bios flashing is dangerous, which it is, and they are protecting themselves from having to replace cards. Also, if you notice they are advertising, with reviews, their Toxic card. What would they rather have? To potentially replace cards that people have failed with, or to sell a card that's $60 more. Also note that the card they are touting is the vivo edition... so as an alternative to the x800pro modded to xt-pe they would suggest their vivo. Obviously they're not talking about a vivo to xt-pe mod.
 
fastcoke11 said:
Some people seem to forget that companies do this all the time -- they underclock a higher-end product so they can sell it cheaper and make more money. There is more demand for the cheaper x800pro. Believe it or not, that is fact -- most people cannot afford an xt-pe. Standard business practice here, to supply cards where they are going to sell. Also, sapphire's cards have not had a good success rate from the outset. If powercolor came out and said it (with their 100% success rate) then I would start to worry (especially since I have a powercolor x800pro vivo). Also, that Sapphire release did not specifically say that vivo to xt-pe mods were failing, they were saying that modding an x800pro to an xt-pe thru bios flashing is dangerous, which it is, and they are protecting themselves from having to replace cards. Also, if you notice they are advertising, with reviews, their Toxic card. What would they rather have? To potentially replace cards that people have failed with, or to sell a card that's $60 more. Also note that the card they are touting is the vivo edition... so as an alternative to the x800pro modded to xt-pe they would suggest their vivo. Obviously they're not talking about a vivo to xt-pe mod.

Sorry but that is wrong.

Think about it. The core of a X800 Pro cost the same as an X800 XT-PE to make. The cost is the same for the materials and manufacturing....I am talking about the GPU not the board which obviously the XT has more expensive ram. With people screaming to buy an XT card why would it make any sense to put XT capable GPUs in Pro cards. Companies with good quanities of high end parts have been known to do this. (Like 2500+ bartons running like 3200+ a stock core volt.) BUT like with the AMD example there was availablity of the high end part and it was not selling as well as the cheaper part.

I have X800 Pro cards in stock.....Made by ATI brand. They are slow sellers....in that price range most people are wanting the 6800 GT which sell as fast as I can get them. On the other hand I still have people Pre-ordering the X800 XT and I could sell them for huge price over MSRP. (Even though I am selling them for MSRP.)

Go look on E-bay.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=40158&item=5120238862&rd=1
 
As an eBay Associate, HardForum may earn from qualifying purchases.
gordon151 said:
The press release makes very specific references to the VIVO and the non-VIVO Pro model and I'm pretty sure they were referring to the latter. So that still doesn’t really answer the question of whether people who are modding the VIVO model of the X800 Pro are actually having any issues with the softmod process. I personally haven’t seen anyone post with an issue on the softmod other than low overclocks and I’ve only seen one person with an actual VIVO fail the mod so far, but then again that doesn’t mean people with those issues don’t exist.

Nah, its pretty clear sapphire is talking about the bios softmod. Read the PR, that is what they refer to throughout the whole thing, using the bios flash to enable 16 pipes. At the end they mention their own product as free advertising.
 
I haven't heard of any board failures because of this ( I am referring to the VIVO's I have modded). I don't see boards just failing for no reason, unless it was modded or seriously over clocked. Even then the one part that I would think would fail would be the memory. I noticed that they wont run much faster than 560 and even at those speeds they are scortching hot. Which is why I always throw some ram sinks on them..
 
I don't see any VIVOs in your auctions. I guarantee if you sold VIVOs at $400 versus the XTs at your ridiculously butt raping price, you would be sold out in a day.
 
AACDIRECT said:
Sorry but that is wrong.

Think about it. The core of a X800 Pro cost the same as an X800 XT-PE to make. The cost is the same for the materials and manufacturing....I am talking about the GPU not the board which obviously the XT has more expensive ram. With people screaming to buy an XT card why would it make any sense to put XT capable GPUs in Pro cards. Companies with good quanities of high end parts have been known to do this. (Like 2500+ bartons running like 3200+ a stock core volt.) BUT like with the AMD example there was availablity of the high end part and it was not selling as well as the cheaper part.

I have X800 Pro cards in stock.....Made by ATI brand. They are slow sellers....in that price range most people are wanting the 6800 GT which sell as fast as I can get them. On the other hand I still have people Pre-ordering the X800 XT and I could sell them for huge price over MSRP. (Even though I am selling them for MSRP.)

Go look on E-bay.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=40158&item=5120238862&rd=1


No, it's not wrong. Manufacturers have been known to do this for many years. Read up on it. I don't have time to link you right now.
 
As an eBay Associate, HardForum may earn from qualifying purchases.
newster said:
I don't see any VIVOs in your auctions. I guarantee if you sold VIVOs at $400 versus the XTs at your ridiculously butt raping price, you would be sold out in a day.

I think you miss read my post. First that is not my auction. I have more pre-orders than I have X800 XT-PE cards which I am selling for MSRP 499.99. The point was that why would ATI put good X800 XT-PE chips on a X800 Pro when they could sell every X800 XT-PE that they could make with them?
 
fastcoke11 said:
No, it's not wrong. Manufacturers have been known to do this for many years. Read up on it. I don't have time to link you right now.


Re-read what I posted I know manufactures do this....they don't do it when there is a demand for the high end part and there is not a surplus stock. Why sell a part that they can get 500.00 for at 400.00. Not logical. When you have a surplus of good parts like when there was tons of 3200+ ships and demand is lower it is common for them to be sold as a lower level part like the 2500+. This was also the case with the R360 core cards. Most would run at 9800 XT speeds because yields were good on the chip. X800 XT-PE are not yielding in large quanities or you would not see people everywhere looking for the cards. And people would not have 6 month old backorders. Companies would not be canceling peoples backorders. In a high demand market ...why would you take a good XT-PE core and put it in a Pro card. That is not how to make money and I am sure ATI does everything they can to make to most profit off every part they make.
 
Re-read what I posted I know manufactures do this....they don't do it when there is a demand for the high end part and there is not a surplus stock. Why sell a part that they can get 500.00 for at 400.00. Not logical. When you have a surplus of good parts like when there was tons of 3200+ ships and demand is lower it is common for them to be sold as a lower level part like the 2500+. This was also the case with the R360 core cards. Most would run at 9800 XT speeds because yields were good on the chip. X800 XT-PE are not yielding in large quanities or you would not see people everywhere looking for the cards. And people would not have 6 month old backorders. Companies would not be canceling peoples backorders. In a high demand market ...why would you take a good XT-PE core and put it in a Pro card. That is not how to make money and I am sure ATI does everything they can to make to most profit off every part they make.


It does make sense, your R360 core is a perfect example. The old classis Athlon is another. There is another possibility too. What if you have a GPU that has the 4 extra pipes functional but will not operate at the higher speeds of the XT with stock cooling? You don't toss it, you put it in a X800. I firmly believe that the ATi process is up to speed and is getting good yields. I also think the shortage of GDDR3 is what is forcing them to sell XT cores as plain X800's. They would rather make less money for a x800 than none at all for not selling the GPU. My2c..


Now to respond to the original poster. No, I have not heard of a single occurance of a card flat out dying from the flash upgrade alone. There have been bad flashes and failures not related to either. Mostly extreme OC'ing and mods gone bad. I not a single incident of, I flashed fine and ran for a bit, then it dieed of a sudden.
 
Evil Scooter said:
It does make sense, your R360 core is a perfect example. The old classis Athlon is another. There is another possibility too. What if you have a GPU that has the 4 extra pipes functional but will not operate at the higher speeds of the XT with stock cooling? You don't toss it, you put it in a X800. I firmly believe that the ATi process is up to speed and is getting good yields. I also think the shortage of GDDR3 is what is forcing them to sell XT cores as plain X800's. They would rather make less money for a x800 than none at all for not selling the GPU. My2c..


Now to respond to the original poster. No, I have not heard of a single occurance of a card flat out dying from the flash upgrade alone. There have been bad flashes and failures not related to either. Mostly extreme OC'ing and mods gone bad. I not a single incident of, I flashed fine and ran for a bit, then it dieed of a sudden.
Possible but highly improbable. Plus nVidia is using GDDR3 on their GTs, correct? They don't seem to be having major issues with yields.

It's a core issue, you know it. The plus side is, I doubt it has anything to do with defective pipes and more a problem with MHz. What I don't understand is what led people to believe that the pipes are defective? How could the 4 pipes that are disabled always be the ones causing problems? Am I missing something here?
 
I call bullshit on this entire fucking thread.

Here's some facts of life boys and girls:

1) Sapphire is full of shit, softmodding your card isn't going to grim reaper the damned thing. They're just saying that to a) cover their asses b) not have thier XTPE product line still-born.

2) To Nvidia fans, this is the SAME FUCKING THING as BFG coming out and saying running an Ultra [at stock] on anything less than a 480W PSU will kill your card. Again, they're just a) covering their asses b) hoping you'll buy their vid card/power supply combo.

3) Pro VIVO cores are typically capable of XTPE speeds [if just barely] for US [H]ard guys. You have to keep in mind ATI doesn't give a fuck about us, they care about Joe Blow. And his Dell box ISN'T capable of running a Pro Vivo at XTPE speeds without melting something. I took out my GT from my current comp, stuck in my old dell box, and watched the speeds shoot up about 10C and my OC get cut in half. This is what ATI means when they say these cores didn't pass their speed bin, they didn't pass in a dell box, not a modded Lian Li with a 126CFM side-panel fan blowing right down on it.

4) The hardline Nvidiots [I can use this word because I am one myself :p] are going to do everything in their power to shoot down the VIVO mod because as of this moment it pretty much is the only thing that can challenge the GT/Ultra for supremecy [until the XTPE comes out in full force]. Personally, I'd still take an OCed GT over a flashed VIVO simply because the Ultra/XTPE seem to be pretty neck and neck, most VIVos don't go much over XTPE speeds [if they hit them at all], while most GTs do go over Ultra speeds [420-430 isn't all that uncommon, at least not here]. And that's without a bios flash, you flash your GT bios to an UE bios [hey fair is fair, if we're comparing flashed ATI products we should do the same to Nvidia products wouldn't you agree?], get some more voltage out of it with that [1.4 to 1.6], and who knows where you'll hit your OC ceiling [though I'd never try it myself, 1-2FPS is not worth the risk of a $400 doorstop]. Besides, it does seem that a GT could be had for less then a VIVO [you could get a waterblock or NV5 with the money left over too]. Or you could get the ASUS GT with 2 molexs and not have to worry about voltage. But like I said that's my preference. Your mileage may vary.
 
^^ i have to agree with the man, It's more that Sapphire just doesn't want you to do it, not that it's going to kill your card - just like any other company that sells VIVO's wouldn't want you to do it.
 
i have had mine running for a big while without any troubles, you are bound to have some failures many could be user related if they volt mod them, etc.
 
sleepeeg3 said:
How could the 4 pipes that are disabled always be the ones causing problems? Am I missing something here?

I think its just easier saying that. Its more of an issue of the core handling data from 16 pipes ( pure speculation). I've only seen it once where the card would artifact with 16 pipes at pro speeds but run fine with 12 pipes at xt speeds....
 
gsboriqua said:
I think its just easier saying that. Its more of an issue of the core handling data from 16 pipes ( pure speculation). I've only seen it once where the card would artifact with 16 pipes at pro speeds but run fine with 12 pipes at xt speeds....

Who said that the same 4 pipes are always disabled? On the 9500 non pros I know for a fact that the disabled pipes varied from card to card. From your experience it would look like the big issue is that many of the chips won't clock to XT speeds with 16 pipes running. If that is true I wish ATI would release a lower speed 16 pipe card....more like a GT level card and drop the price of the 12 pipe card to the 6800 level.

The Batman said:
3) Pro VIVO cores are typically capable of XTPE speeds [if just barely] for US [H]ard guys. You have to keep in mind ATI doesn't give a fuck about us, they care about Joe Blow. And his Dell box ISN'T capable of running a Pro Vivo at XTPE speeds without melting something. I took out my GT from my current comp, stuck in my old dell box, and watched the speeds shoot up about 10C and my OC get cut in half. This is what ATI means when they say these cores didn't pass their speed bin, they didn't pass in a dell box, not a modded Lian Li with a 126CFM side-panel fan blowing right down on it.

Very much right on....every company has a set of specs for a part. The part runs this speed at this temp and air flow with this size psu...etc...etc.

I had a 9800 XT that ran @ 500 core/410 mem all day long on stock cooling. It was a hand selected card and was far above the average spec. All X800 chips are made to be X800 XT-PE GPU's most fail to reach the spec for one reason or the other. Be it clock speed or physical errors or errors at a specific temp. The great thing is that smart overclockers figure out how to make the card run at a higher spec. I took the same card that ran at 500/410 on air .....volt modded it.....added a pelt and water cooling and ran it at 575/430. Point is the gpu was bin for 412/365 which it ran easy. Every 9800 XT I saw ran stock fine....but to run above that speed varied form card to card. I have seen tons of 9800 Pros that would run @ XT speeds ut I saw some that would not. But everyone ran stock fine.

All Sapphire is saying is that if you flash the bios on your VIVO card to a XT-PE bios and the card dies. NO WARRANTY. You modded the card....you broke it....go buy a new one. To be honest it is far....you bought a X800 PRO.....you got a X800 PRO....you tried to make an X800 XT-PE out of it....your attempt failed....go buy a new one.

I volt modded my 9800 XT...will ATI take it back if it burns up when I return it with the volt mod on it....H3LL NO. It was worth it. ;) So fast......
 
AACDIRECT said:
Who said that the same 4 pipes are always disabled? On the 9500 non pros I know for a fact that the disabled pipes varied from card to card. From your experience it would look like the big issue is that many of the chips won't clock to XT speeds with 16 pipes running..

huh? You must have misunderstood, but no matter. I actually have had very good luck with cards running at XT speeds. So far I've only had 3 out of 19 fail to run as an XT.....
 
AACDIRECT said:
Re-read what I posted I know manufactures do this....they don't do it when there is a demand for the high end part and there is not a surplus stock. Why sell a part that they can get 500.00 for at 400.00. Not logical. When you have a surplus of good parts like when there was tons of 3200+ ships and demand is lower it is common for them to be sold as a lower level part like the 2500+. This was also the case with the R360 core cards. Most would run at 9800 XT speeds because yields were good on the chip. X800 XT-PE are not yielding in large quanities or you would not see people everywhere looking for the cards. And people would not have 6 month old backorders. Companies would not be canceling peoples backorders. In a high demand market ...why would you take a good XT-PE core and put it in a Pro card. That is not how to make money and I am sure ATI does everything they can to make to most profit off every part they make.

I think you are one of the few people to have made sense.

There is SO much demand for the XTPE that they would absolutely not put XTPE capable cores in a Pro. Just would not make good business sense right now. GDDR3 was inititally thought to be the problem but I dont think that is the case now. Nvidia is chucking out Ultras. I can get you a BFG 6800 Ultra in the UK today. If Nvidia can get the RAM so can ATi.

ATi is almost certainly having trouble getting high clock speed and 16 fully working pipes out of their chip fab plants. What the reason for these low yields I dont know. And I bet ATi has every chip designer and fab engineer trying to fix it.

This is from an Nvidia GT thread:

....at least, thats what mine had. BFG 6800 GT has been running 410/1110 without any problems at all for the last month. Until a week ago when i started getting artifacts while running star wars galaxies with AA enabled through the driver. Ran the high-dynamic range light demo again to monitor temps and everything was still the same... nothing higher than 81C under load for over 30min.

Bumped back down to stock and scrubbed coolbits, reinstalled the 61.77's and i'm still getting arts at stock!?! and now nothing but garbage out...no post...no boot...nothing but garbage then straight to reboot. Fan on the card works fine (and no, the sticker did not come off and get stuck in the heatsink). My 465watt psu has 35amps on the 12v rail and will occasionally flux between 11.92-11.86v, but my understanding is that this isn't anything out of the ordinary (correct me if wrong). So, the only conclusion i can come to is that the card had a faulty core that progressively got worse over time. The overclock could have sped up the process a bit but its still a fairly tame overclock compared to what others are doing.

Your thoughts, opinions, advice, etc is greatly appreciated.

also this might be interesting to know:

http://www.guru3d.com/newsitem.php?id=1828

The point I am trying to make is that its early in the game this round. And demand for high end parts is not being met. It would be business suicide to sell high end parts at a lower price range if there is nothing wrong with those parts.

X800Pros != X800XTPEs at the moment

and if you read that Guru3D article you will see that GTs != Ultras either.
 
Syphon Filter said:
I think you are one of the few people to have made sense.

Thanks....I have a strong feeling that the issue is the supplier of silicon wafers that ATI has had problems with recently. The wafers were out of tolerance due to the manufacture using old equipment. I would bet that is the reason that the yield of high end parts is lower than expected.

Wish ATI would clear up what the deal is.
 
Thanks....I have a strong feeling that the issue is the supplier of silicon wafers that ATI has had problems with recently.

I thought the work was done at TSMC.?
 
Back
Top