blazorthon
Glorious
- Sep 24, 2010
- 14,022
- 21
- 43,265
GTX 640 has too much memory. GTX 640 and 650 are overpriced even compared to current cards in that price range. GTX 660 TI, 670, 680, and 690 all don't have enough memory. I can't imagine the 680 being used outside of huge 3D Eyefinity resolutions even in Metro 2033 and BF3, so it's memory capacity is obviously not enough. It should be 4GB by default. The 670 should be 3.5GB, the 660 TI should be 3GB. The 640 should be 1GB.
All the GPU horsepower in the world isn't going to help you if you don't have the memory capacity to play at the resolutions with the quality settings set high enough to make a difference. I'd prefer 4K @ 60FPS over 1080p @ 300FPS, all with quality settings and AA maxed out. The 7970 can be overclocked something like 45%, so that 680 has better be able to be similarly overclocked or it doesn't have enough value even if it had 4GB of memory capacity.
Also, what happened to Nvidia abandoning hot-clocking and gonig for a many core approach like AMD has? Didn't Nvidia come out and tell us that GK104 has 1536 cores?
[citation][nom]letsdothis[/nom]I don't know about you all but I think NVidia's lost the crown and isn't really doing a lot to get it back. The only comparisons aren't much better than before. What nano-scale are they building at? And AMD is going to have a 3/4 year head-start before these new cards come out? At again not much savings to the consumer? They have to get with the program on 28 nanometers, price, bang for the buck, power consumption, noise. These new cards in my opinion are just trying to catch up not go beyond what AMD has already. Absolutely not sold on NVidia for about a year now. I compared the actual thruput to AMD and for the money there was no contest. I ended up with a AMD HD 6870 this time around for less than $180 and got a 2Tbit output that would have costed close to $500 to a comparable output with NVidia.[/citation]
If the 680 is really 45% faster than the 7970, then Nvidia obviously isn't just playing catch up, they want to win. These cards have better price/performance than current video cards so it's not at an added price to the consumer. These are all 28nm cards so that's a pointless demand. Please tell me how 45% faster than AMD's flag-ship and offering much lower MSRPs for comparable performing Kepler cards than anything on AMD's lineup is not going beyond where AMD is right now. The only problem that we know of is the delay.
Nvidia has been pretty much beaten on all important matters for a while, but they were not playing catch-up, Nvidia still had the fastest card at any given time, the problems were other things like cost, memory capacity, power usage, etc, performance itself was not a limiting factor over the last few years. For example, GTX 285 versus Radeon 4890, GTX 480 versus Radeon 5870, GTX 580 versus Radeon 6970,and if this is to be believed, GTX 680 versus Radeon 7970. The Nvidia card wins in absolute performance in most games in all examples. So no, Nvidia probably won't be losing the performance crown any time son unless AMD makes a video card with a GPU die as large as the GF/GK 100/110 dies. Overall, AMD and Nvidia seem to have somewhat similar performance per square mm of die area, with AMD leading on Nvidia a little if the GTX 680 really performas 45% better than the 7970 and has a 550mm2 die, so all it should take AMD to compare or beat Nvidia in the one performance place where they lost (flagship single die performance) is for AMD to use larger dies. However, that would defeat the purpose of the smaller die, cheaper GPUs.
Another problem I see here is the relative performance of the dual GPU flagships. The GTX 580 performs more like a dual GTX 570 than a dual GTX 580 because of it's lower clock frequencies. This is why AMD's 6990 can keep up with the GTX 590, AMD's card doesn't need to lower it's frequencies. However, looking at this, Nvidia seems to be tow tiers ahead of AMD instead of the usual one tier ahead. AMD's 7970 wouldn't be too far ahead of the GTX 660 TI. If the GTX 690 performs like dual 670s, well then a dual 7970 (7990) won't compare to it as AMD has done in the past unless the 7990 has upgraded Tahitis or higher clock frequencies than the 7970 has at stock. Considering the overclocking headroom of Tahiti, AMD could do it, but whether or not AMD can tolerate being second best in both single and dual GPU cards remains to be seen.
All the GPU horsepower in the world isn't going to help you if you don't have the memory capacity to play at the resolutions with the quality settings set high enough to make a difference. I'd prefer 4K @ 60FPS over 1080p @ 300FPS, all with quality settings and AA maxed out. The 7970 can be overclocked something like 45%, so that 680 has better be able to be similarly overclocked or it doesn't have enough value even if it had 4GB of memory capacity.
Also, what happened to Nvidia abandoning hot-clocking and gonig for a many core approach like AMD has? Didn't Nvidia come out and tell us that GK104 has 1536 cores?
[citation][nom]letsdothis[/nom]I don't know about you all but I think NVidia's lost the crown and isn't really doing a lot to get it back. The only comparisons aren't much better than before. What nano-scale are they building at? And AMD is going to have a 3/4 year head-start before these new cards come out? At again not much savings to the consumer? They have to get with the program on 28 nanometers, price, bang for the buck, power consumption, noise. These new cards in my opinion are just trying to catch up not go beyond what AMD has already. Absolutely not sold on NVidia for about a year now. I compared the actual thruput to AMD and for the money there was no contest. I ended up with a AMD HD 6870 this time around for less than $180 and got a 2Tbit output that would have costed close to $500 to a comparable output with NVidia.[/citation]
If the 680 is really 45% faster than the 7970, then Nvidia obviously isn't just playing catch up, they want to win. These cards have better price/performance than current video cards so it's not at an added price to the consumer. These are all 28nm cards so that's a pointless demand. Please tell me how 45% faster than AMD's flag-ship and offering much lower MSRPs for comparable performing Kepler cards than anything on AMD's lineup is not going beyond where AMD is right now. The only problem that we know of is the delay.
Nvidia has been pretty much beaten on all important matters for a while, but they were not playing catch-up, Nvidia still had the fastest card at any given time, the problems were other things like cost, memory capacity, power usage, etc, performance itself was not a limiting factor over the last few years. For example, GTX 285 versus Radeon 4890, GTX 480 versus Radeon 5870, GTX 580 versus Radeon 6970,and if this is to be believed, GTX 680 versus Radeon 7970. The Nvidia card wins in absolute performance in most games in all examples. So no, Nvidia probably won't be losing the performance crown any time son unless AMD makes a video card with a GPU die as large as the GF/GK 100/110 dies. Overall, AMD and Nvidia seem to have somewhat similar performance per square mm of die area, with AMD leading on Nvidia a little if the GTX 680 really performas 45% better than the 7970 and has a 550mm2 die, so all it should take AMD to compare or beat Nvidia in the one performance place where they lost (flagship single die performance) is for AMD to use larger dies. However, that would defeat the purpose of the smaller die, cheaper GPUs.
Another problem I see here is the relative performance of the dual GPU flagships. The GTX 580 performs more like a dual GTX 570 than a dual GTX 580 because of it's lower clock frequencies. This is why AMD's 6990 can keep up with the GTX 590, AMD's card doesn't need to lower it's frequencies. However, looking at this, Nvidia seems to be tow tiers ahead of AMD instead of the usual one tier ahead. AMD's 7970 wouldn't be too far ahead of the GTX 660 TI. If the GTX 690 performs like dual 670s, well then a dual 7970 (7990) won't compare to it as AMD has done in the past unless the 7990 has upgraded Tahitis or higher clock frequencies than the 7970 has at stock. Considering the overclocking headroom of Tahiti, AMD could do it, but whether or not AMD can tolerate being second best in both single and dual GPU cards remains to be seen.
Facebook
Twitter
Instagram