Official AMD Radeon 6000 Series Discussion Thread

[TheMailMan78]

cadaveca I think we are arguing the same thing in different ways. I agree with you 100% that unless the 69xx series doesn't one up the 480 by at least 30% Ill call BS on AMD.

In all honesty I think we will see up to 35% or maybe even 40% in some cases with the 69xx series over the 480. Otherwise they wouldn't be screwing around with the names.

 

[bear jesus]

I've been buying every card on launch day for like 6-7 years now. This launch is the first that I am ignoring. I'm gonna spend my money playing with Intel chips...I won't buy another card until I can get 3 monitors working together in a single card, with 120+FPS(for 3D), and 4xAA.

I assume you mean every high end card on launch thus you would not have cared about the 6870 no matter what, but how can you say you will be ignoring the 6970 launch when you don't even know how powerful it will be?

 

[cadaveca]

Because I can make an "educated" guess, and even the most optimistic of those guesses will not be enough for me.

There's only so much technology in existence, and anything new, gets patented. So it's quite easy to look at the market, what's available, find out what's possible, and go from there. I didn't know what nV's price drop was going to be, pretty much to the dollar, for any other reason than that's what made the most sense.


Sometimes, something comes that is really surprising, but really...that doesn't happen all that often.

And no, not just high-end cards...but mostly, yes.

 

[TheMailMan78]

Because I can make an "educated" guess, and even the most optimistic of those guesses will not be enough for me.

There's only so much technology in existence, and anything new, gets patented. So it's quite easy to look at the market, what's available, find out what's possible, and go from there. I didn't know what nV's price drop was going to be, pretty much to the dollar, for any other reason than that's what made the most sense.


Sometimes, something comes that is really surprising, but really...that doesn't happen all that often.

If I'm right can I rub it in?

 

[cadaveca]

LoL...only if you realize you rubbing it in isn't even gonna give me pause...

Just be sure that you aren't swallowing some troll bait.

 

[pantherx12]

If shader architecture scales well from 6870 the new card should be 20% better than gtx 480, how ever that does not take into account rop counts and memory band width and crap like that.

That is if stuff like this even scales like that. heh.

 

[TheMailMan78]

LoL...only if you realize you rubbing it in isn't even gonna give me pause...

Just be sure that you aren't swallowing some troll bait.

I like living in my bubble. Its warm in here.

 

[cadaveca]

I'm gonna leave this thread now, till maybe the new cards come out, but I gots one last thing to say...


GF104 & Barts = not square chips(ie rectangular). Find out why, and you'll know why Cayman is most likely gonna be skipped by me.


Also, noone has mentioned that the new displayport 1.2 does 3 monitors with one onboard mini-dp connector, but only using a device that doesn't exist on the market at the moment(MST Hub, available 2011), or monitors that do not exist either(DP1.2 monitors with daisychain ports, ETA unknown)?

And how does this affect Eyefinity?

Also, only one DVI is dual-link...not both...so using 2xDVI and one DP, max Eyefinity res is 5760x1200...

Oh, and a link:

http://www.amd.com/us/Documents/AMD_Radeon_Display_Technologies_WP.pdf

Why would I skip Cayman? Should be kinda obvious, now....

 

[HalfAHertz]

I'm gonna leave this thread now, till maybe the new cards come out, but I gots one last thing to say...


GF104 & Barts = not square chips(ie rectangular). Find out why, and you'll know why Cayman is most likely gonna be skipped by me.


Also, noone has mentioned that the new displayport 1.2 does 3 monitors with one onboard mini-dp connector, but only using a device that doesn't exist on the market at the moment(MST Hub, available 2011), or monitors that do not exist either(DP1.2 monitors with daisychain ports, ETA unknown)?

And how does this affect Eyefinity?

Also, only one DVI is dual-link...not both...so using 2xDVI and one DP, max Eyefinity res is 5760x1200...

Oh, and a link:

http://www.amd.com/us/Documents/AMD_Radeon_Display_Technologies_WP.pdf

Why would I skip Cayman? Should be kinda obvious, now....

Eyefinity is not enough? Did you want a res higher than 3xFull HD?

 

[erocker]

If I'm right can I rub it in?

LoL...only if you realize you rubbing it in isn't even gonna give me pause...

Just be sure that you aren't swallowing some troll bait.

I like living in my bubble. Its warm in here.

I'm gonna leave this thread now, till maybe the new cards come out, but I gots one last thing to say...


GF104 & Barts = not square chips(ie rectangular). Find out why, and you'll know why Cayman is most likely gonna be skipped by me.


Also, noone has mentioned that the new displayport 1.2 does 3 monitors with one onboard mini-dp connector, but only using a device that doesn't exist on the market at the moment(MST Hub, available 2011), or monitors that do not exist either(DP1.2 monitors with daisychain ports, ETA unknown)?

And how does this affect Eyefinity?

Also, only one DVI is dual-link...not both...so using 2xDVI and one DP, max Eyefinity res is 5760x1200...

Oh, and a link:

http://www.amd.com/us/Documents/AMD_Radeon_Display_Technologies_WP.pdf

Why would I skip Cayman? Should be kinda obvious, now....

So, I take from this that TheMailMan enjoys giving backrubs and bubble baths. Cadaveca will allow TheMailMan to give the backrub, but then is going to leave.

I had an ex girlfriend who left me for pretty much the same reason. Be gentle you two.

 

[DanishDevil]

 

[bear jesus]

http://gpuz.techpowerup.com/10/10/25/f77.png

i'm like a little kid the day before christmas, i want my toys

Do you plan on doing any overclocking?

 

[JrRacinFan]

@DD

How much of a difference over that 460 1GB?

i'm like a little kid the day before christmas, i want my toys

Do you plan on doing any overclocking?

Who are you kidding?? DD, an overclocker? I havent seen him overclock in a LOOONNNGGG time.

 

[DanishDevil]

Not at first. Going to be testing it out at stock with some two's. Starcraft 2, Bad Company 2, and Mafia 2.

Now ladies and gentlemen, what you've all been waiting for!

 

[trt740]

Very nice and can Asus package and item or what

 

[bear jesus]

Awesome pic's, but once again there is another reason for me to be jelous of you... i want your torture rack

 

[dir_d]

That turtle sure does look happy

 

[CDdude55]

That looks so hawt!!!

 

[mastrdrver]

I think one thing to remember about the 68XX cards is that these cards have the biggest redesign since the HD2900XT, ATI's first unified card, every card since then has basically had the same exact shaders rehashed and updated until AMD fixed them in the 4XXX series.

The shaders are not changed. The dispatch processor that feeds the shaders has changed as well as cache organization, etc.

ahhhh so the design they took from the now 7000 series is just the dispatch processor etc set up and things like that.

so this is the same 5D architecture as the 5870 with and updated tesselator, . . . . . . . damn

All wrong. (5D yes but no updated tessellator)

All the reviews I briefed through expecting to say what some of the technical changes that were made to Barts (over Cypress/Juniper) were non existent (have not got a chance to read W1zz's yet). Today TR posted their review of the Barts cards and finally got some detail on what changed.

Result:

The image above maps out the major components of the Barts chip in a familiar fashion. For the most part, this is the same core GPU architecture we know from the Cypress chip behind the Radeon HD 5800 series, only scaled down slightly and tweaked in several ways

....and about the dispatchers and rasters?:

By the way, you may have noticed the presence of two "ultra-threaded dispatch processor" blocks in the diagram above, and if you're into these things, you may have recalled that the diagrams of Cypress only showed one of these blocks. Truth is, though, that this diagram of Barts is simply more detailed than the earlier one of Cypress. AMD's David Nalasco tells us both chips have dual "macro sequencers," as AMD calls them internally, to "dispatch instructions to the SIMDs." (There's also a "micro sequencer" in each SIMD.) As the diagram shows, each macro sequencer has instruction and constant caches. One bit of detail missing above is a crossbar between the two "rasterizer" blocks and the macro sequencers, so either sequencer can be fed by either rasterizer.

To take you further down the rabbit hole, the presence of two rasterizers in the diagram above may be a little bit misleading. As with Cypress, Barts has dual scan converters, but it lacks the setup and primitive interpolation rates to process more than one triangle per clock cycle. That's in contrast to the GF104, which can process two polygons per clock tick, or the GF100, whose max is four.

So wtf changed?!:

Although the setup rate hasn't changed in Barts, the chip's internal geometry processing throughput should be higher thanks to some selective tweaks..............As essentially a very effective form of compression, tessellation involves a relatively small amount of input data and a much larger, sometimes daunting amount of output data. To better deal with this data flow in Barts, AMD "re-sized some queues and buffers," according to Nalasco, "to achieve significantly higher peak throughput" in certain cases. At the same time, thread management for domain shaders, which handle post-expansion geometry processing, has been improved.

AMD claims these changes had "negligible impact" on Barts' transistor budget and power draw, yet the firm has measured tessellation throughput for Barts at up to twice that of Cypress in directed tests. The biggest gains come at lower tessellation levels, as show in the image below. At higher levels, the chips' common setup rate likely becomes a limiting factor, and the two are separated only by Barts' slightly higher clock speed.

In other words we did some minor tweaks to the shaders and cut FP64 to save on transistor count. Anand gets it right we he states that "Barts is a very direct descendant of Cypress". Reason being is that nothing really changed! Most of the changes were to the display output and some other minor things.

The shader changes made to Barts helps throughput and thread management of tessellation. TR shows ~40% increase in performance in the Heaven bench using heavy tessellation at 1280x720. Sounds awesome until you realize its a 10fps improvement over ~20 and the exact same percent increase is seen from a GTX 460 768MB to a GTX 470. Barts is still triangle setup limited which is why there is no real improvement when it comes to heavy tessellated games.

Barts appears to really just be a 5850 with less shaders but a higher core clock that makes up the difference. Since it has less shaders and kicked FP64 to the curb it also has a smaller die. Thanks to TSMC screwing 32nm we all get meh from video cards this time around.

 

[MxPhenom 216]

Leaked HD6970 Results

What's being said, is the HD 6970 is scoring X12000 in 3D Mark Vantage, with a P score of P24499.

Unigine Heaven results (1920 x 1200, 4xAA, 16xAF) giving a result of 36.6 fps.

Core speeds seem to be at around 800 - 900MHz, 6 + 8pin power requirements and an estimated TDP of 255W or more.

Source

I dont think this is reliable, but if its true then ati messed up look at the TDP thats what the GTX480 is isnt it? And my GTX470 gets 32 FPS with the same settings with Tesselation on Extreme. So their GTX480 score isnt accurate

If nvidia can drop the price on the 480 to be below the 6970 launch price then ati may be screwed

 

[aCid888*]

Lets leave the speculation to BTA for the front page news.



At the moment, this thread (along with any others like it) are just troll magnets and as a result, anyone with any sense wont be replying to save themselves an argument*.






*I guess I have no sense lol



Edit: This post was moved from a thread created by the user above, sorry if it offends any one or looks out of place.

 

[MxPhenom 216]

Lets leave the speculation to BTA for the front page news.



At the moment, this thread (along with any others like it) are just troll magnets and as a result, anyone with any sense wont be replying to save themselves an argument*.






*I guess I have no sense lol

way to just completely bombard my thread

 

[Athlon2K15]

Chiphell usually has a good source on ATI cards,if these are true in the slightest bit it's fail for ati only 10% faster than 480 is weaksauce.

 

[MxPhenom 216]

Chiphell usually has a good source on ATI cards,if these are true in the slightest bit,its fail for ati only 10% faster than 480 is weaksauce.

Yeah the the GTX580 could be 20% or 30% faster then the 480!

I already promised myself im going to skip out on the next gen cards from either nvidia or ati and just get another GTX470 since it seems more and more console ports are going to be coming out so i could jsut live with sli for like 2 or 3 more years. im pretty sure even BF3 or Crysis 2 wont be able to put 470 sli to its kneeze

 

[AltecV1]

Chiphell usually has a good source on ATI cards,if these are true in the slightest bit,its fail for ati only 10% faster than 480 is weaksauce.

calling these cards weaksauce is not quite accurete BUT i get what you are saying and i agree !