why does 660ti beat 7950 so often?

[Fastx]

What exactly are you waiting to hear?

The OP picked the 3 games that he thought NV was faster at, but coincidentally was shown that 1 of them happened to have outdated drivers and the 7950B was indeed considerably faster. The other 2 run better on NV period.

Thread title could/should say (yet lose so often and overall) as it's completely misleading.

According to thread title, everything has been relevant. If OP still wants to wonder why NV wins in 2 NV leaning games, at least put it in the title.

What's next a thread on why the 7950B wins the 660 ti so often, or why not 670 or even 690 if there is a game or two it wins (but not overall)?

Considering that the 7950B is the faster card, has amazing OC potential, and is clearly not slower than the 660 TI, why not discuss the 7950B and show the facts in the thread with a misleading title as it is?

If you check my link for Crysis 3 the His 7950B is also beating the 660 TI (per his post 2560x1600) so its 2 of them or 2 out of 3.

 

[amenx]

I see some people focusing on 2560x1600. If the OP was on that res, then yes, the 7950 would be the way to go. But the OP is interested in 1920x1080 (post 36), which is mostly a toss-up with most recent games.

I went thru the same 7950 vs 660ti dilemma when I chose my card, but seeing that I'm on the 'majority res' (1920x1080) for the next year or 2 and dont OC heavily, I just went with the card I believed would be more comfortable with, which was the 660ti.

Also contrary to what some may think or suggest, Nvidia does not stop future driver improvements once games run well when first released. I frequently see improvements in game performance long after games are out.

 

[wuseche]

Because Nvidia has better drivers.

 

[Black Octagon]

Of course you cannot compare the specs directly. That's the interesting question: Why?

RS already said why. They are different GPU architectures...

P.S., Cards' clock speeds ARE among their 'specs'

P.P.S;, It may indeed be 'interesting' to answer the OP literally by only comparing the 660Ti to the original 800Mhz 7950. No doubt. But the OP asked about a card that literally no longer exists on the (first-hand) market, so it's equally, if not more, 'interesting' to talk about those versions of the 7950 that actually exist. As it turns out, from Page 2 of this thread, the OP was actually referring to the Ghz edition of the 7950 and not the 800Mhz original. So those like RS who chose to look beyond the original 7950 were justified after all.

 

[MrK6]

Why are people flipping out over this? The 660Ti is faster in nvidia sponsored games and at lower resolution because because that's the target of its design. It's not unfeasible that such a card could beat a higher tier card, especially in situations where the higher tier card isn't being pushed. And yes, the thread title is misleading, but like many people on this board the OP is probably trolling/doesn't understand high school statistics.

 

[Black Octagon]

Why are people flipping out over this? The 660Ti is faster in nvidia sponsored games and at lower resolution because because that's the target of its design. It's not unfeasible that such a card could beat a higher tier card, especially in situations where the higher tier card isn't being pushed. And yes, the thread title is misleading, but like many people on this board the OP is probably trolling/doesn't understand high school statistics.

Yes, it's on-par with the 7950(s) in my opinion. As always, some games tend to perform better on NVIDIA and others on AMD hardware. The OP was thinking, as it turns out, mainly about 3 games. Not a majority of games but nonetheless 3 quite popular games. It wasn't, AFAIK I detailed survey of all the games commonly benchmarked by review sites. He/she seems to have interpreted this as the 660Ti having a lead overall, and this has been demonstrated to not be true

 

[Fastx]

I see some people focusing on 2560x1600. If the OP was on that res, then yes, the 7950 would be the way to go. But the OP is interested in 1920x1080 (post 36), which is mostly a toss-up with most recent games.

I went thru the same 7950 vs 660ti dilemma when I chose my card, but seeing that I'm on the 'majority res' (1920x1080) for the next year or 2 and dont OC heavily, I just went with the card I believed would be more comfortable with, which was the 660ti.

Also contrary to what some may think or suggest, Nvidia does not stop future driver improvements once games run well when first released. I frequently see improvements in game performance long after games are out.

If your taking about me focusing on the 2560x1600 it was a response to the below post by the op and also with the 13.1 the 7950 also beat the 6600TI at 1920x1200 (C3 & FC3) in my link just for the record with the 13.1's. I was just responding to the OP post's nothing more.

660ti beating 7950 GHZ at Crysis 3 even at 2560x1600 Very High Settings:

http://www.guru3d.com/articles_pages/crysis_3_graphics_performance_review_benchmark,7.html

Looks like either Nvidia simply has the better card, or Guru3d has some seriously flawed tests.

 

[videogames101]

Driver version matters.

This is a troll thread.

 

[Lepton87]

Comparing 660ti to the original 7950 800MHz is like comparing FX8350 to 3770K at 2.5GHz, sure it's faster in some cases at stock, but does it make it a better card? I'd say the best competing card from AMD is actually 7870. 7950 is like an engine that revs to 7K but has a rev limiter at 4.5K.

Sure, 660Ti is faster then 7870 but they are within 10% of each other but both are significantly cut-down and both are clocked properly.

It can even lead to erroneous results when comparing different GPU architectures from the same brand. 925mhz 32 ROP 7970 vs. 880mhz 6970. On paper the former is barely faster in pixel fill-rate. In the real world, 7970 puts down 51% higher pixel fill rate performance despite both cards having 32 ROPs and 7970 being clocked just 5% higher (925mhz vs. 880mhz). I

please, not this again, this test tests memory bandwidth not fill-rate. Don't believe me? Reduce memory clock on 7970 by half and see what happens.

 

[amenx]

Comparing 660ti to the original 7950 800MHz is like comparing FX8350 to 3770K at 2.5GHz, sure it's faster in some cases at stock, but does it make it a better card? I'd say the best competing card from AMD is actually 7870. 7950 is like an engine that revs to 7K but has a rev limiter at 4.5K.

Sure, 660Ti is faster then 7870 but they are within 10% of each other but both are significantly cut-down and both are clocked properly.


please, not this again, this test tests memory bandwidth not fill-rate. Don't believe me? Reduce memory clock on 7970 by half and see what happens.

Again, another 2560x1600 example . As has been abundantly pointed out, if that is your intended res to use the cards with, then yes, 7950 is the logical choice. At 1920 resolutions, then its a toss-up where card features, cost, preferences, power draw, etc, may play the bigger role in choice of cards.

p.s. If you're on 2560x1600, then probably you would not be looking at a 7950, but rather a 7970 or 680 or above.

 

[3DVagabond]

Again, another 2560x1600 example . As has been abundantly pointed out, if that is your intended res to use the cards with, then yes, 7950 is the logical choice. At 1920 resolutions, then its a toss-up where card features, cost, preferences, power draw, etc, may play the bigger role in choice of cards.

p.s. If you're on 2560x1600, then probably you would not be looking at a 7950, but rather a 7970 or 680 or above.

Does this make you feel any better? @ 1920*1200 the 660ti still isn't faster than an 800MHz 7950. Add to that the 7950 boost is faster yet. Plus, if you are an O/C'er, most (all?) 7950's will go past the stock 7970GHz performance.

 

[amenx]

Does this make you feel any better? @ 1920*1200 the 660ti still isn't faster than an 800MHz 7950. Add to that the 7950 boost is faster yet. Plus, if you are an O/C'er, most (all?) 7950's will go past the stock 7970GHz performance.

Depends on the games being tested. Any review could have a selection of games where the net result is more with one card vs the other. I cared more about Crysis 3, FC3 (the 2 games that actually pushed me to upgrade to a 660ti) than I do about games I've long ago played and forgotten about (crysis1, metro 2033). All other games (except Tomb Raider, which still runs nicely), my 660ti plays perfectly fine maxed out, incl Sleeping Dogs and a few others where the 7950 may be ahead, so here a slightly faster card becomes irrelevant to me in those games. Finally, 4% isnt enough of a clincher to make me choose one card over another. There are other attributes I may value in a GPU (mentioned earlier) over a slight net FPS gain another card may have.

 

[3DVagabond]

Depends on the games being tested. Any review could have a selection of games where the net result is more with one card vs the other. I cared more about Crysis 3, FC3 (the 2 games that actually pushed me to upgrade to a 660ti) than I do about games I've long ago played and forgotten about (crysis1, metro 2033). All other games (except Tomb Raider, which still runs nicely), my 660ti plays perfectly fine maxed out, incl Sleeping Dogs and a few others where the 7950 may be ahead, so here a slightly faster card becomes irrelevant to me in those games. Finally, 4% isnt enough of a clincher to make me choose one card over another. There are other attributes I may value in a GPU (mentioned earlier) over a slight net FPS gain another card may have.

That is fine. I'm not trying to say you, or anyone else, should select a 7950 over a 660ti. Just showing that in a wider selection of benches that overall the 660ti is not faster than the 7950. Even at 1200 without heaps of AA, as some are trying to say (TPU uses 0-4xAA).

For the record, the 7950 is only about 2.5% faster overall using TPU as the reference. That's truly an irrelevant amount. Of course, that is the 800MHz model. Boost cards, which are the current reference, are clocked at 925MHz (You have to use the powertune slider to get that full time, but not including that would be semantics.). That's ~16% increase (Conservatively, maybe an additional 10% performance?). That's not counting any O/C'ing the user can get, because I don't want to expand the comparison too much and just complicate things unnecessarily. The point is made without O/C'ing..