Official AMD Radeon 6000 Series Discussion Thread

[cadaveca]

Intel Core i7 isn't disappointing despite having one of the largest CPU dies ever made, and hot. As long as it lives up to performance:Watt for its category (something Core i7 does, while GTX 480 doesn't (HD 5970 having better thermals, better power consumption and better performance)), people won't give a rat's arse.

No offense, man, but that's exactly what I said in the following paragraph...high-end is about performance, #1. High-end users will go that extra mile and spend yet more on cooling, if needed...

I then went on to say it's only disappointing becuase the process is holding the design back. I don't get what YOUR point is, other than reinterating what I already said.

However, you've revealed your own disappointment in Fermi...but the only thing wrong with Fermi was the management behind it. Is that your point?

 

[largon]

http://i403.photobucket.com/albums/pp112/erocker414/26439011101.jpg

Big GPU. Not Fermi big, but big.

Big GPU? How come?
Chip carrier on that PCB is exactly the same size as Cypress'. No way to deduce die size from that pic.

 

[20mmrain]

The 69xx series will be fail to me if it doesn't meet the following specs....

6970 = 5870 in crossfire.
6950 = 5850 in crossfire.

Anything less would be uncivilized.

http://img.techpowerup.org/101101/barkley.jpg

I don't think that will happen considering that my 6870 Xfire matches and in some case beats the 5870's in xfire as well as the GTX 470's.

Another thought to others.... even if this chip is big and hot.... who cares. I mean for ever now we have heard nothing but complaining about How that doesn't matter on Fermi. So if it didn't matter on Fermi it shouldn't matter on Caymen.

Also even if it is hotter then we are used to for ATI cards.... I still don't think it will be as hot as Fermi was. We have been spoiled with ATI cards as of late. Anything past 75c w/ fan on auto would be hot in terms for ATI. So even if it runs 85c under load.... I still think we are in really great shape!

 

[cadaveca]

Anything past 75c w/ fan on auto would be hot in terms for ATI. So even if it runs 85c under load.... I still think we are in really great shape!

Unfortunately, this is not true. I'd say anything over 90c would be hot. My reference 5-series cards ran @ 90c++, and I had many of them.


Only my newest non-reference runs lower than 90C, and not by much. Claiming anything over 75C is hot, is spreading FUD.

 

[TheMailMan78]

Unfortunately, this is not true. I'd say anything over 90c would be hot. My reference 5-series cards ran @ 90c++, and I had many of them.


Only my newest non-reference runs lower than 90C, and not by much. Claiming anything over 75C is hot, is spreading FUD.

Dude I don't know WTF kinda luck you have been having BUT DAMN! My 5850 under load and OC never hits 75c. I mean I idle at 53c and I'm OC with the fan at 20%.

 

[cadaveca]

I idle @ 45c or something. I do not manually adjust fanspeed, or OC my cards(they were all pre-overclocked).

I actually had a good chat with a tech @ XFX the last time, and although my cards may seem hot to you, that doesn't mean mine were really all that hot either.

XFX quotes thier own design as 15c cooler than reference, and should load @ 75C. I agree. It's also alot quieter. that means my cards were a bit hot, but only by a few C, to be honest, and that 90c is the norm.

 

[TheMailMan78]

I idle @ 45c or something. I do not manually adjust fanspeed, or OC my cards(they were all pre-overclocked).

I actually had a good chat with a tech @ XFX the last time, and although my cards may seem hot to you, that doesn't mean mine were really all that hot either.

XFX quotes thier own design as 15c cooler than reference, and should load @ 75C. I agree. It's also alot quieter. that means my cards were a bit hot, but only by a few C, to be honest.

Well I just OC my cards. I never mess with the fan speed. Buts heres a shot of mine. FYI mine is a reference cooler.

 

[the54thvoid]

Well.

My checklist for buying gfx card:

Performance: 10% more than my current (2 x 5850) is okay.
Noise: Not louder than my current.
Power: 10% more than my current.
Temp: Not bothered, as long as it works.
Price: Up to £400 if it meets the above. Will pay more if it beats the above.

If 6970 offers this at launch I'll probably dive in. If I have to wait for the GTX 580 I would only do so if it looks like it'll thrash it.

 

[yogurt_21]

damn you mail man! I kep clicking the post reply in your image, wondering wtf was wrong.

but as always, ambient temp, case airflow, other component cooling and etc. can drastically change results. at stock my 470 never breaks 80, oced it can hit 95+

 

[cadaveca]

Well I just OC my cards. I never mess with the fan speed. Buts heres a shot of mine. FYI mine is a reference cooler.

You know I'm gonna bite you on this one, right?


You are comparing temps of a 5850, to a 5870.


:shadedshu





In the end, the only reason the high temps my cards had were of any concern to me was throttle. If they don't throttle, I don't care. ATI cards do not throttle @ 90c.

 

[TheMailMan78]

damn you mail man! I kep clicking the post reply in your image, wondering wtf was wrong.

but as always, ambient temp, case airflow, other component cooling and etc. can drastically change results. at stock my 470 never breaks 80, oced it can hit 95+

lol I did the same thing!

You know I'm gonna bite you on this one, right?


You are comparing temps of a 5850, to a 5870.


:shadedshu

Mailman fail. I just read 5 series card.

 

[the54thvoid]

My 2nd 5850 runs at 80+ degrees when full out. The other stays at 70 ish. (XFX 5850 - no o/c).

I must buy from the same guy as Cadaveca.

 

[cadaveca]

Mailman fail. I just read 5 series card.

Well, by and large, 5830/5850/5870 are all the same gpu, so shoudl ahve the same thermal limits. but only gpus with excessively high leakage in the 5830/5850 models will run the temps of a 5870.

Likewise, the 57-series and lower use the same process, so are stuck with the same thermal limits, too...they are simply a smaller slice of Cypress, anyway.

 

[TheMailMan78]

My 2nd 5850 runs at 80+ degrees when full out. The other stays at 70 ish. (XFX 5850 - no o/c).

I must buy from the same guy as Cadaveca.

The ref. design for the 5xxx series is the best IMO. My temps are much lower then that on my 5850. Of course that depends on your OC.

Well, by and large, 5830/5850/5870 are all the same gpu, so shoudl ahve the same thermal limits. but only gpus with excessively high leakage in the 5830/5850 models will run the temps of a 5870.

Likewise, the 57-series and lower use the same process, so are stuck with the same thermal limits, too...they are simply a smaller slice of Cypress, anyway.

Yeah but transistor count affects temps does it not?

 

[cadaveca]

The ref. design for the 5xxx series is the best IMO. My temps are much lower then that on my 5850.

I dunno. I was kinda along with the same thinking as you, but now that Ive spent some time with non-reference design, I actually prefer it. No software votlage control doesn't bother me...if I wanted more volts, soldering on a VR isn't something that's ever stood in the way of me doing what I want.

In fact, this new card isn't even a XXX model..it's a brand new vanilla model, that they flashed the Black Edition bios onto for me. It's cooler quieter, and clocks better, than all of the others I had...and i could give the others more juice!

I'm actually of hte opinion that now that AMD has taken over ATI, reference boards will no longer be the desired board to have...it will be the custom boards that matter.

Yeah but transistor count affects temps does it not?

Really, that's dependant on cooling. But the per-area leakage should be the same with all of the 5-series, as the transistor density doesn't change, merely the count. 5830 and 5850 run cooler becuase they use less transistors on the same slice, effectively making the unused sections heatsinks. Just take a look at Juniper power consumption vs Cypress...Juniper, the gpu itself, draws about half of Cypress...please don't forget that memory, etc are included in most benchmarks, so actual gpu power usage isn't really evident there.

 

[trt740]

Until I added my Zalman cooler my 5850 non ref would hit in the 90c range when oced , but now it almost never goes over 70c. I asked Asus about this and they said one it is because I have a solid side on my case with no fan and two that 90c is acceptable for a load temp. I jus like mine to run a bit cooler.

 

[Benetanegia]

Really, that's dependant on cooling. But the per-area leakage should be the same with all of the 5-series, as the transistor density doesn't change, merely the count. 5830 and 5850 run cooler becuase they use less transistors on the same slice, effectively making the unused sections heatsinks. Just take a look at Juniper power consumption vs Cypress...Juniper, the gpu itself, draws about half of Cypress...please don't forget that memory, etc are included in most benchmarks, so actual gpu power usage isn't really evident there.

True that. I think that this three reviews can help prove that*:

ASUS HD6850 Direct CU

ASUS GTX460 Direct CU

ASUS GTX450 Direct CU

3 cards using the same cooler. According to what most people believe the card(s) with lower power consumption should be cooler, but this is not the case because die area also plays a big role on heat dissipation. That's how it can happen that GTX460 (13W iddle / 187W full load) and GTS450 (10W/148W) have the same temps on the same cooler with similar fan settings and how the HD6850 (15W/125W) has higher temps despite having higher fan speeds and a lower power consumption. Transistor density is higher on the HD6850 and that affects temps.

I think this is something that people should learn, because far too many people think that size = consumption = temps is a fixed relation and that is not truth.

* Maybe Wizzard reviewd more cards with this same cooler. If so and someone wants to link to them, perfect. They may further prove my point or disprove it and it would just be nice to find it out, either way. I just didn't find more cards, I didn't search too much either, I consider my time precious.

 

[Tatty_Two]

Here is a screenie of my 5850 overclocked, running OCCT GPU tool for a minute and a half, 98% load, temps get to 70C, however you will note fan is inaudible at just 45%

 

[cadaveca]

Nice one. I'd be quite interested to see how the 6-series performs in the same test.

 

[trt740]

Here is a screenie of my 5850 overclocked, running OCCT GPU tool for a minute and a half, 98% load, temps get to 70C, however you will note fan is inaudible at just 45%

Now come on ya old monkey who plays a game for a minute and a half and mine won't get hot in a minute either. What I want to see is a 5850 oced to the same specs as a 6870. Say 900 /500 and I bet the 5850 wins. Plus I bet it runs just as cool in most cases. It appears to me the 6870 doesn't overclock any better than a 5850. Mine will do 1000 core plus no problem. You can also pick up a used 5850 for allot less than a 6870. I'm willing to bet tess. isn't better at the same clocks either. Still crossfire seems better on the 6800 series and power consumption as-well. Don't get me wrong the 6870 is a very nice card and a decent buy for a new build. The point of these cards was to find a way to make ATi's gpu's more affordable and more profitable which they have done. Heat wasn't a issue with the 5800 series.

 

[bear jesus]

one after a little over a min

one once the temps have leveled out

Sorry about the res

*edit*
Damn it i defaulted my CCC settings

 

[trt740]

Now come on ya old monkey who plays a game for a minute and a half and mine won't get hot in a minute either. What I want to see is a 5850 oced to the same specs as a 6870. Say 900 /500 and I bet the 5850 wins. Plus I bet it runs just as cool in most cases. It appears to me the 6870 doesn't overclock any better than a 5850. Mine will do 1000 core plus no problem. You can also pick up a used 5850 for allot less than a 6870. I'm willing to bet tess. isn't better at the same clocks either. Still crossfire seems better on the 6800 series and power consumption as-well. Don't get me wrong the 6870 is a very nice card and a decent buy for a new build. The point of these cards was to find a way to make ATi's gpu's more affordable and more profitable which they have done. Heat wasn't a issue with the 5800 series.

a minute is not near enough time but those are similar temps as the 5800 series.

 

[Tatty_Two]

Now come on ya old monkey who plays a game for a minute and a half and mine won't get hot in a minute either. What I want to see is a 5850 oced to the same specs as a 6870. Say 900 /500 and I bet the 5850 wins. Plus I bet it runs just as cool in most cases. It appears to me the 6870 doesn't overclock any better than a 5850. Mine will do 1000 core plus no problem. You can also pick up a used 5850 for allot less than a 6870. I'm willing to bet tess. isn't better at the same clocks either. Still crossfire seems better on the 6800 series and power consumption as-well. Don't get me wrong the 6870 is a very nice card and a decent buy for a new build. The point of these cards was to find a way to make ATi's gpu's more affordable and more profitable which they have done. Heat wasn't a issue with the 5800 series.

Lol OCCT GPU tool will warm her up more than most games, but for the sake of consistency, I will manulaly set the fan to as high as I can without hearing it and then run the tersts to see what she will do again, watch this space.

 

[bear jesus]

Grrrr damn OCCT, 975/1100 seamed stable with ATI tool, furmark and 3dmark06.... it was not, OCCT caused a fail, i would agree it warms the gpu up pretty fast and hot, after about 1:30 it was almost at the leveled out temp.
My ambient temps are about 28c.

*edit*
It was the memory that was unstable, see below for the 1ghz core at stock volts

I may have just damaged my hearing by setting the fan to 100% though

 

[Tatty_Two]

here you go, just for you Tom, thought I would really go to town, GPU set at 900mhz, fan set manually to 50% only (still cannot hear it), Furmark for 8 minutes, forced 8xMSAA, 97% GPU load, temps don't go above 66C Now who's gonna tell me these DirectCU coolers don't do their job!