Why don't NVIDIA make 512 and 448 bit cards today like in the Fermi era?

[cst1992]

So, the 970 and 980 have launched. Great cards, though I'm a little disappointed that they are having a 256 bit memory interface. Why don't NVIDIA produce wider memory interfaces now if they used to do it for the 5xx series cards? 224GB/s seems a little low compared to the 320GB/s of the 290x and the 336 GB/s of the 780Ti, even though the 980 is faster than the 780Ti, and has insane memory clocks - more than 7GHz!
It just doesn't make sense to me that they're going for the small improvement in frequency rather than the large one in bit width. I saw the overclocking profile of the 290x Lightning and it was 1600 MHz memory for a total of 432 GB/s! I was like what the heck!
What do others think? Is there a logic behind this that I'm missing?

 

[RobCrezz]

The bus width is irrelevant. Its the total bandwidth that is relevant to the performance - This is affected by not just bus width, but also vram speed and also any compression processes.

But that's exactly what they state, total bandwidth.

GTX 680 - 192GBs.
GTX 770 / 970 / 980 - 224GBs.

If the GTX 680 was already hitting the ceiling, I am not so sure about the others, even with compression technology and all. And the GTX 670 (with the same bandwidth of the GTX 680) was having trouble to follow the HD 7870 in 8XAA. They are just too close to say it is safe. Hence why I think a review could shed some light on it.

as i can understand different GPU will have different efficiency in using their bandwidth. despite having much less bandwidth 680 still perform much better than 7970 in general (which means kepler is more efficient than GCN in term of utilizing available bandwidth) . so in the test you link above 7870 is beating 670 when 8x MSAA being use. so did you think bandwidth limitation causing such problem. then again look again at 7870 spec.

http://www.techpowerup.com/reviews/AMD/HD_7850_HD_7870/

it has the same bus with 670 which is 256bit. 670 have total bandwidth of 192GB/s while 7870 only have around 154 GB/s. it has the same memory width and much lower bandwidth than 670 then why 7870 slightly beat 670 when using 8x MSAA? the way i see it AMD GCN architecture are much better in handling MSAA than Kepler.

but there are also talks about 680/670 being bottlenecked by bandwidth. but to simply attribute that 680/670 being weak with higher level of AA because lack of bandwidth is not entirely accurate. it is more like 680/670 in general being bottleneck by it's total bandwidth.

+1

You need to see the whole picture. Its easy to just point the finger at "memory bus width" without understanding how it all works.

Nvidia wouldnt intentionally go with something that would make its product perform worse - There is obviously other benefits to using a smaller bus in some cases.

 

[cst1992]

It's not that they are intending that their product perform worse - they're just fine with that much bandwidth.
Have any of you guys heard of Linear Programming? There we have a set of constraints and a function we have to maximize or minimize; for example, given raw material and labour constraints and to maximize profit/reduce cost.
Maybe they're using a similar approach using various test cases - past data shows that there is x % improvement in general using more memory bandwidth, but it increases cost and power consumption by y %, so strike the right balance of x and y to get out the maximum real world performance out of the card.

 

[Contiusa]

I see that the 7870 has lower bandwidth, but the review still points out a bottleneck and I remember AnandTech mentioning that the GTX 770 was an improvement over the GTX 680 bottleneck at higher resolutions, so my take is that Nvidia is skimming on the quality, the same way they kept selling high end cards with only 2GB. When I bought the 770 I thought about it all and it was really a hard decision to make. I won't go back to AMD because the Nvidia drivers are not giving me the trouble I had with AMD, but I don't like to pay that much for a card with outdated features and that might well be bottlenecking my game.

 

[Ironsounds]

Hello... Good point CST... Untill the day they start making their Own memory Chips you are at the mercy of your venders on price.

I became a Nvidia Guy, back with Geforce 3, Using, installing, and configering the ATI Software is still, to this Day, something I never look forward too. I Never fear the NVida driver installs for a Build, Upgrade, or repair.