It doesn't make sense to buy if it is going to be more expensive than the 690 and not faster than it.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
GTX Titan (final specs and bench)?
- Thread starter GoldenTiger
- Start date
- Status
It doesn't make sense to buy if it is going to be more expensive than the 690 and not faster than it.
If I was going to spend that kind of coin, I'd take the Titan over a 690 every time.
gamingarena23
Limp Gawd
- Joined
- Dec 7, 2006
- Messages
- 247
Ditto
Actually cant wait to sell my 680SLi to get one Titan...
People dont realize that even though the 680SLi or 690 might be faster, but when you average all the games that dont scale well with SLi and the games that dont use SLi at all, there is just no contest the Titan is the card to get even if the rumors are 80% true...
I rather have 90% of performance all the time, then 100% half the time, do the math
Last edited:
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
2 Titans in sli would murder 690 sli. quad sli does not scale for shit in many cases and at the settings and resolution that it does scale fine, the 2gb of vram would kill it. so really Titan would be able to run much higher settings too since its not limited to 2gb of vram.
Well, I'm not gonna flame or start sum shitty war/debate...But....I know sli is iffy in certain games...But usually with driver updates, or updating sli profiles with inspector...the issues can be fixed....Looking from this benchmark from Metro 2033...690 sli scaling seems to be working on this game...I don't know...I want the best I can get...new tech is always nice, but why pay more for less performance?
http://www.overclock3d.net/reviews/gpu_displays/evga_gtx690_review_including_quad_sli/11
Ditto
Actually cant wait to sell my 680SLi to get one Titan...
People dont realize that even though the 680SLi or 690 might be faster, but when you average all the games that dont scale well with SLi and the games that dont use SLi at all, there is just no contest the Titan is the card to get even if the rumors are 80% true...
I rather have 90% of performance all the time, then 100% half the time, do the math
Than why did you get 2 680?
This is what I don't understand about people who go Sli. People who go sli always end up complaining about it. A single 680 would give you enough performance. So giving that there is always a newer single gpu in the works you telling me you can't get enough performance out of it until a new single gpu comes out? Especially considering there haven't really been any super demanding games to really take advantage of it fully. It I were 690 or 680 Sli users right now I'd be pretty annoyed. As I don't think in this case the money has been worth it if you are considering "upgrading" to the Titan.
Last edited:
So, I am confused. a GTX 690 is close to two GTX 680 in SLI. Let us just call them equal for now. Let us do some simple math. Titan is said to be 90% AS powerful as a 690 so 90% AS powerful as BOTH 680 cards COMBINED. How much more powerful would that be over ONE card.
I take 90% of 2, which is 1.8, than say 1.8 - 1 = .8, so the card is .8 more powerful than one card so .8/1 * 100 = 80% more powerful than ONE card. Why are people saying it is only 60% more powerful? Now if you say the 690 is only 95% of 680 SLI the math is
95% of SLI, 90% Titan of 690, times 2 680 is 1.7, then -1 gives .7 or 70% more powerful. I guess that is close to the 60%
I take 90% of 2, which is 1.8, than say 1.8 - 1 = .8, so the card is .8 more powerful than one card so .8/1 * 100 = 80% more powerful than ONE card. Why are people saying it is only 60% more powerful? Now if you say the 690 is only 95% of 680 SLI the math is
95% of SLI, 90% Titan of 690, times 2 680 is 1.7, then -1 gives .7 or 70% more powerful. I guess that is close to the 60%
So, I am confused. a GTX 690 is close to two GTX 680 in SLI. Let us just call them equal for now. Let us do some simple math. Titan is said to be 90% AS powerful as a 690 so 90% AS powerful as BOTH 680 cards COMBINED. How much more powerful would that be over ONE card.
I take 90% of 2, which is 1.8, than say 1.8 - 1 = .8, so the card is .8 more powerful than one card so .8/1 * 100 = 80% more powerful than ONE card. Why are people saying it is only 60% more powerful? Now if you say the 690 is only 95% of 680 SLI the math is
95% of SLI, 90% Titan of 690, times 2 680 is 1.7, then -1 gives .7 or 70% more powerful. I guess that is close to the 60%
why would you take 90% of 2 when your comparing them with 2 GPUs in sli that makes no sense... ofcourse it will be 80% cause you chose a variable of TWO which stands for two GPUs... factor in 90% x 2 then you get 180% vs the 200% base of TWO 680s or basically ONE 690... the principle still stands that the titan is 90% faster than ONE STOCK 680
Well lets see here.
1) There are games that can take advantage of it
2) There are folks running 1440/1600p displays
3) Surround gaming
4) Not everyone is willing to wait until the next card comes out
5) $$$ someone who already has a 680 can improve their performance dramatically by getting a second 680 for a fraction of the cost, particularly if they pick up their 2nd 680 used, which a lot of people do.
A lot of people who go SLI do so because they need/want more performance and either aren't any better options available, or the better options are not cost effective. Most people who start off with a clean slate will opt to get something like a Titan vs a 690 or SLI 680s for the better consistency and at that point, cost difference between the three options isn't much.
I guess what I'm getting at is for a 690 or 680 Sli user I don't see how it makes sense to get a Titan. The Titan is more positioned in my mind for someone who either owns a 680 or earlier who wants the best and wants to avoid Sli issues completely. Yes I get there will always be people who want the absolute best , but even here that is questionable and subjective as we know a Titan will probably not be faster than a 690 or 680 Sli.
A 690 or 680 Sli will handle games today and even tomorrow still as great.
Its not about getting a titan to replace a sli setup its more about getting two titans for sli can you imagine how much e-peen that would mean for some people here!? theyd be bragging about their titans (pun intended) for days! lol
For someone who already has a 690 or 680 SLi, I agree. Unless they are having some major SLI related issues and they can sell of their current hardware to offset the cost of Titan by a substantial amount, it probably doesn't make sense, assuming the performance figures end up being what we think they'll be.
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
if you only look at fps then that is the problem as there is more to smooth gameplay than that. and while you are sitting there waiting on updates to fix sli issues then someone with a Titan will already be getting much more consistent performance. there is no need to put up with less vram, more overhead and less consistent performance for just a slightly faster overall card all for around the same price.
Its not about getting a titan to replace a sli setup its more about getting two titans for sli can you imagine how much e-peen that would mean for some people here!? theyd be bragging about their titans (pun intended) for days! lol
But then they are just getting into more potential Sli issues.
Metro 2033 is the RARE exception for good quad SLI scaling. 99% of other games will either show like a 3% improvement or negative scaling for quad sli. 99% of the games we purchase are console ports and it will never change unless tablets and smartphones completely cannibalize consoles. I use two GTX 690's for quad SLI and I already know that two Titans will completely obliterate the performance of my cards as long as they are 50%+ faster than GTX 680's (some rumors say 80%, 90% faster which just justifies the purchase even more) Id rather have two Titans working at 100% and 75% on 2 GPU scaling rather than GTX 690's working at like 100%, 75%, 25%, and 5% scaling for the 4 GPU's. Plus I run out of VRAM on a single 2560x1440 with my 690’s and I usually have to lower my AA down to 4x or 2x on some more demanding games out these days.
Whach
[H]ard|Gawd
- Joined
- Dec 22, 2011
- Messages
- 1,244
I have not had any problems. I actually enjoy my setup, and thus wont be going to the Titan any time soon. I don't think SLI owners ALWAYS complain about it.
that graph makes no sense its saying that a 590 is as fast as a 690 sli, really now? mhmm
cause if thats the case then no 690 scaling is not 100%... lol
180% vs 200% correct, so one card is 200/2=100 so 180-100=80% more than ONE 680 like I said
gamingarena23
Limp Gawd
- Joined
- Dec 7, 2006
- Messages
- 247
Well i never complained about the SLi, it works perfectly and i had SLi since it came out yeas ago, just that Titans is that much better at this point in time, they both overkill for single 2560 and under but i rather have one Titan then 2x680 and open option for 2nd Titan,plus the Resale value is there for 680 now, so might not cost me nothing to get Titan.
If you want to be on the bleeding edge of the hardware, that's how you stay.
And im sure the actual people that buys Titan are the same people that already owns 680Sli or 690.
You will not see 650-670 owners buying them, those are the same people that complain about the price.
And i know you dont understand like you said but im sure lots of people that think of buying the actual Titan understands perfectly.
No one is saying that anymore.
addictedto60fps
[H]ard|Gawd
- Joined
- Mar 31, 2010
- Messages
- 1,329
Have any sites performed any kind of price/performance ratio against different cards? Something like each fps will cost you so and so amount on this card on this game and you then aggregate the results and take the average across multiple games. I just made some random numbers up and came up with this. I used these as the baseline for my calculations. Yes, I do realize these numbers/assumptions may not be entirely realistic or accurate, but I'm just trying to use this as a means to compare the perceived value of these setups:
1. Titan costs $1100
2. Two GTX 680's cost $920
3. Titan is 80% faster than a single gtx 680 and this translates to 80% increase in performance in all games over a 680
4. Baseline of 60 fps for x game
Now, let's take a game where SLI doesn't work and Titan yields an 80% increase over the baseline of 60 fps.
cost per fps in sli = $15.33
cost per fps for titan = $10.19
Let's now take a game where SLI yields 50% increase over single GTX 680 and Titan still yields 80% increase over single GTX 680:
cost per fps in sli = $10.22
cost per fps for titan = $10.19
And lastly, let's take a game where we get 100% performance over a single 680 using SLI versus 80% for Titan:
cost per fps in sli = $7.67
cost per fps for titan = $10.19
When you aggregate everything and take the average for all 3 games under the above assumptions, Titan comes out on top, on a $ amount per fps average:
SLI = $10.22 per fps
Titan = $10.10 per fps
1. Titan costs $1100
2. Two GTX 680's cost $920
3. Titan is 80% faster than a single gtx 680 and this translates to 80% increase in performance in all games over a 680
4. Baseline of 60 fps for x game
Now, let's take a game where SLI doesn't work and Titan yields an 80% increase over the baseline of 60 fps.
cost per fps in sli = $15.33
cost per fps for titan = $10.19
Let's now take a game where SLI yields 50% increase over single GTX 680 and Titan still yields 80% increase over single GTX 680:
cost per fps in sli = $10.22
cost per fps for titan = $10.19
And lastly, let's take a game where we get 100% performance over a single 680 using SLI versus 80% for Titan:
cost per fps in sli = $7.67
cost per fps for titan = $10.19
When you aggregate everything and take the average for all 3 games under the above assumptions, Titan comes out on top, on a $ amount per fps average:
SLI = $10.22 per fps
Titan = $10.10 per fps
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
yeah there is no way that Titan will be 90% faster than a 680. it will be around 50-60% faster based on the rumored specs.
what review sites are telling us right now is that a titan is 90% faster or so they claim.
so lets say 0% baseline is one GTX 680. and reviewers say titan is +90% since their claim says 90% faster.
in reality the math would look like this: +90% (titan) : 0% (GTX 680) in sli it would be +180% (titan) : 100% (GTX 680) i get why your getting 80% its cause your comparing it in sli... lol
yes your argument is true in SLI that TWO titans is 80% faster than TWO GTX 680s in SLI
.
but if you go with a single card config the titan is 90% faster
note that when i say the titan is 80% faster in SLI mode it is almost equivalent to 4 GTX 680s in QUAD SLI
On what games? I've never heard of VRAM issuses at the res except for heavilly modded Skyrim. What games tax quad SLI at that res? I run 2560x1600 with 2 680s fine.
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
please link me to one halfway reputable site claiming Titan is 90% faster than a 680.
values are skewed you should base it on msrp all these sites selling them for 1100-1200 right now dont even have inventory so they dont count
your math is all wrong and so are the rumors that the Titan is 90% faster than a 680. 2 in SLI vs 2 in SLI = 1 vs 1, ie 80% faster
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
still waiting for a link to one of these review sites claiming 90%...
im sticking with my math... take it for what its worth lol but i do agree that 90% rumors are mythical hell even 80% might be an overstatement i quoted cause i thought something was off in your calculations but to each his own.
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
so you just made up the 90% claim all by yourself then cause no site is claiming that anyway.
if i was to say that the titan was 80% faster in a single card configuration (according to my calculations) then i would have stated that in sli it would only be 60% faster in sli.
i used 90% for numbers sake
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
THIS is what you said and it was total BS as NO review site claimed that.
addictedto60fps
[H]ard|Gawd
- Joined
- Mar 31, 2010
- Messages
- 1,329
Yes, values may very well be skewed, hence my disclaimer in post:
"Yes, I do realize these numbers/assumptions may not be entirely realistic or accurate, but I'm just trying to use this as a means to compare the perceived value of these setups:"
I'm just trying to give an example of a way to measure the perceived performance of these cards, that's all. And 80% is wishful thinking on my part. I doubt that it will be 80% increase in performance over 680, but if it is, that would be something.
Oh, nice. Thank you. This is the kind of analysis that I was looking for.
last time i checked these forums yall were raving that the titan is going to be twice as fast as the 690 and close to the 690 im basing my statement on what information i have and what the general consensus is.
and your right i cant give you a link to a site claiming 90% thats my bad
http://www.overclock.net/t/1359269/obr-hardware-new-gtx-titan-benchmarks-have-arrived
check that out
i dont think youd count that legit but fwiw its something
check that out
i dont think youd count that legit but fwiw its something
- Joined
- May 18, 1997
- Messages
- 55,629
Some of you are adding NOTHING to this conversation. Stop it.
cannondale06
[H]F Junkie
- Joined
- Nov 27, 2007
- Messages
- 16,180
what? YOU said review sites were claiming 90%. I asked for links because I knew that was not true and now you are just making yourself look silly.
im sticking with my math...
well I hope you don't use math much since I went to a math help forum just to check my numbers and they said I was correct. I'll try one last time.
Let us say a 680 is 50 units strong, so 2 in SLI is 100 units. If Titan is 90% of 2 in SLI it is 90 units. 90 (one titan) / 50 (one 680)=1.8, it is 1.8 times more powerful than 1 680, so 1.8 - 1 is a .8 increase, or 80%
edit - MY FINAL OFF TOPIC POST
What happens to the review samples after the review? I assume they are returned.
- Status