680M = 1344 Stream Processors @ 720MHz with 256-bit memory bus
Titan = 2688 Stream Processors @ 837MHz (876MHz boost) with 384-bit memory bus
Add on top of that SLI scaling and a single Titan is faster.
It's not estimation but real world testing. Read again what I wrote until it sinks in. I had both the titan and sli 680 and at 1440p the sli 680m is faster when both are overclocked on air.
Feel free to add in more bs while completely ignoring real world results.
The 680m has a 256-bit memory bus per chip, sure... but there are two chips, doubling the effective bandwidth of the system. The SLI system as a whole effectively has a 512-bit memory bus.
Lol look at this fool going on and on. I don't need to pull desktop graphs and speculate. I tested both and 680m sli with the unlocked vbios is faster than an oc titan with both on air. Keep posting nonsense until you get tired.
READ what I'm writing about the unlocked vbios until it sinks into your thick skull.
So you have a 2560x1440 laptop with 680M SLI now? I clearly show Titan being 2x faster than GTX670 in Company of Heroes 2. That alone disproves that 680M SLI can beat Titan in "every game" because you'd need to overclock the memory on 680M 65% to get to 670's memory bandwidth and the GPU clock almost 40%. You expect people on this forum to believe you had oveclocked 2x 680M 40% on the core and 65% on the memory on air in a laptop, and the CPU was also overclocked? Was your laptop dipped in liquid nitrogen? And that's only enough to match a stock Titan in CoH2. In some games where Titan excels, the 680M loses by 15-30% to a 7970M for crying out loud. HD7970M is slower than a 7870. You expect people to ignore those tests too?
My guess is you are comparing 680M SLi on a 1080P laptop against Titan on a 2560x1440 monitor. In that case sure the 680M SLI OC on your previous laptop is faster than Titan on a 1440P desktop monitor. Such a comparison is completely misleading however and is not how Adam was comparing the 2 setups.
You seem to also be ignoring that 680M has a voodoo power rating from games in reviews. 670 is 50% faster than 680M and in some games it takes 670SLI to match a stock Titan. Please explain to everyone here how your 680M could overcome a 50% performance deficit against a desktop 670 in the first place? Or you are just going to keep telling us you tested games and we are expected to believe that?
I just showed you some basic evidence (3dmark run) that shows the 680M SLI beating the Titan. Not only that, I have consistently mentioned to you that the 680M SLI has an unlocked vbios (you can find it on my site) which enables massive overclocks with no throttling or overheating. Repeat that last part in your head about 10 times and you might understand. You are talking out of your rear and no amount of graphs will convince me because I have the Titan and had the 680M SLI and know exactly what they are capable of achieving. You have neither and aren't worth talking to.
P.S. I'll have a buddy of mine who still has an M18x-R2 jump into this conversation and have him run some benchmarks against my titan in some popular DX 11 titles. That should shut you up once and for all..though I think you enjoy posting meaningless graphs so you'll keep at it.
Why didn't you address the question as to what resolution you are running the 680m SLI vs. the Titan? I have no dog in this fight, but in your entire reply you do not respond to what is the central question regarding your comparison.
So instead, I'll bring in an M18x-R2 owner who will help confirm what I've been saying all along
Confirmed by posting various game benches right? Sounds good.
I did already, 1440p but he chose to ignore it. It beat the Titan in BOTH resolutions when I last tested it. He keeps comparing apples to oranges and tries to present it as some kind of proof. So instead, I'll bring in an M18x-R2 owner who will help confirm what I've been saying all along. Look, the Titan is an impressive card, if it weren't I wouldn't have built a system around it. However, that doesn't negate the fact that the M18x-R2 with an unlocked vbios and 680M SLI is equally impressive if not more so given its TDP and power constraints.
Just for fun, here's another M18x-R2 score from a guy who also has a Titan like me:
I'm sure he'll be told of this thread as well..he loves benchmarking so RS may be biting off more than he can chew.
You are the one comparing apples to oranges. Hes comparing stock vs stock. You seem to be comparing modded/oced vs stock.
People are now running modded titans and running substantial overclocks, with no throttling etc.
Can't see the first one without registering.
Look up a couple posts, I already posted it.
What's the post number? I don't see anything.
Big difference in CPU speeds there...
GTX 680M SLI OC: 18030 Graphic score http://forum.notebookreview.com/att...-benchmark-thread-part-2-3d11680m2960xm55.jpg
GTX Titan OC: 15827 Graphic score http://legitreviews.com/images/reviews/2144/titan-oc-93.jpg
Game set match. Joker is correct
Don't be daft. We need to see multiple game benches to properly compare, which means CPU speeds about the same to make a better comparison. Cherry picking a synthetic bench is not enough data. RS posted multiple benches, you need to do the same seeing you called him out (and resorted to insults in the process).How is that relevant for GRAPHIC score? Unless the CPU used with Titan is bottlenecking it, which it doesn`t since its 3960X, the comparison is highly valid
You don`t get higher graphic score with a faster CPU unless the GPU is bottlenecked
I commend RS for putting up with some of the idiocy he deals with so professionally.Don't be daft. We need to see multiple game benches to properly compare, which means CPU speeds about the same to make a better comparison. Cherry picking a synthetic bench is not enough data. RS posted multiple benches, you need to do the same seeing you called him out (and resorted to insults in the process).
One result doesn't make an argument. Post some more benchmarks and until then work on your decorum.Yeah the desktop one that cloudfire posted had much higher physics and combined (as expected from a 3960X) but lower graphics score. So that just helps the 680M SLI's case even more. Remember the M18x-R2 uses a mobile cpu (4 cores), not an extreme desktop one with 6 cores.