• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

6770 full specs

Hehe yea. It's not like people (well us anyway) would be would dashing out to buy the new card on day one without even reading reviews on performance and price :p

Well, if past releases are anything to go by the Americans will get them first, so they'll be plenty of reviews out by the time we see anything. If they're as good as people as saying, the staff at ocuk will probably take the first batch home for further "testing".;)
 
Well, if past releases are anything to go by the Americans will get them first, so they'll be plenty of reviews out by the time we see anything. If they're as good as people as saying, the staff at ocuk will probably take the first batch home for further "testing".;)

Actually the Chinese get them first as the bulk of the worlds cargo gets shipped through China and Hong Kong hence why you see threads with links to some Chinese site that’s actually managed to bench the card before anyone else.
 
Last edited:
Actually the Chinese get them first as the bulk of the worlds cargo gets shipped through China and Hong Kong hence why you threads with links to some Chinese site that’s actually managed to bench the card before anyone else.

I thought that was down to the fact that they're manufactured in China, rather than they "pass through" China.
 
That was down to the exchange rate dipping a lot, however I do think they should have maybe adjusted the pricing for such a situation, then again they didn't exactly need to considering how well the 5 series has sold.

Yes but also it was because amd wanted market share and they made nvidia drop their prices a lot, if amd never wanted market share they would have priced their card just below nvidia's.

The 4870 was about 1/2 price in the end,if i recall right
 
Actually the Chinese get them first as the bulk of the worlds cargo gets shipped through China and Hong Kong hence why you see threads with links to some Chinese site that’s actually managed to bench the card before anyone else.

Agreed, you do some some cards out in hong kong first
 
I think they're made in Taiwan - if that's made in China then I guess that depends if you believe if Taiwan is Chinese or its own independent state. :p

Taiwan indeed, whether it's classed as China though, I suppose that's dependant on who you ask. :p
 
Yes but also it was because amd wanted market share and they made nvidia drop their prices a lot, if amd never wanted market share they would have priced their card just below nvidia's.

The 4870 was about 1/2 price in the end,if i recall right

Well not just that, it's also about selling the most cards for the best ratio of profit per card to overall profit really. No point in them pricing a 4870 at £300 if they'd sell 10x more at £200, for example. That example would be better suited in US Dollars though.
 
Kyle yeah I know I'm just saying china is china and Taiwan is Taiwan,
Us Brits could give a hoot if china thinks Taiwan belongs to china, as long as one of them makes our graphics cards they can do what they want
 
what exactly are those rpes ? maybe they put 2 cores in single card ? that would explain the renaming and performance.
Nah. I would guess that the RPEs are either their name for an indivisible bank of shaders (160 shaders = 1 RPE) or more likely (IMO) the processing unit responsible for geometry shading/tesselation etc.
 
Last edited:
RPE = "render and processing engine". That is all we know for sure...

However, I doubt it is specifically to do with tessellation, as tessellation is a geometry-based operation (i.e. the sub-division of polygons), and has little to do directly with rendering. The above suggests a broader application. My guess is that it is similar to what Baboonanza suggests; an indivisible bank of shaders. Increased modularity is a broad requirement of maintaining / improving scalability to larger numbers of SPs and greater data throughput (something nvidia moved towards with Fermi). Whether a tessellation unit is associated with each RPE, or whether a global tessellation unit exists, will depend on the nature of the design (I suspect that a tessellation unit for each RPE would be logical, given the modular design paradigm, but we will have to wait and see).

Also, if you take a look at this picture (posted in another thread):

confidential237549.jpg


It SEEMS, from the shape of the blurs, that the Cayman has 3 RPEs, and 480(x4) stream processors. It's hard to be sure, but the number of texture units is definitely a two-figure number (i.e. not 128), so this adds some weight to the assumption. Also, the number of RPEs looks to be the wrong shape for a "4".

It will be interesting to see how independent these RPEs are, how many transistors (and so much die area) is given over to improved command / control logic, and how they interface with the memory. IF they interface with the memory independently and in parallel, we could be looking at a 384bit bus on Cayman (though this is not something we've heard much about so perhaps not).
 
Last edited:
96 TMUs, 48 ROPs is my guess :)

Edit:I shamelessly robbed this from another board:

"RPEs refer to the chunks of SIMD Cores with TMU (and ROPs this time?).

So Barts has 2, resulting in 320 (x4) SPUs plus 64 TMUs and 32 ROPs;
Cayman will have 3, thus 480 (x4) SPUs plus 96 TMUs and 48 ROPs... "
 
Last edited:
Compared to 80 TMUs and 32 ROPs for the 5870, sounds reasonable.

So 5870 had 1600 shaders but if we are to believe the rumours they are only 4/5 efficient compared to 6970 so in theory that means 1280 vs 1920
 
Last edited:
Back
Top Bottom