I don't think that comparison can be taken at face value.
Every source I can find gives the number of transistors for 780 and 780ti as 7.1 billion, with the same area. (Not totally sure, just doing a quick google search, correct me if I'm wrong.)
However some of the transistors for the 780 are disabled. I don't know where to find the "actual" number of transistors for a 780, that is, how many are actually in use. That unused area would be useful for heat dissipation.
Did you really miss this very clear point? This was stock vs stock R9 290X Tri-X vs GTX780Ti, not an OC Tri-X at 1225 core clock.
"the Sapphire card performed better than the GTX 780 Ti in Bioshock Infinite and Crysis 3 while the GTX 780 Ti's only definitive victory came in Battlefield 3. In the three other games tested, the cards were so close that I'll call it a performance tie."
Keep moving those goal posts and missing the obvious.
Then consider Titan, same results.
That still has same die area, with some disabled transistors.
I'm sure there's other design factors involved, but I'm saying transistor density does have an effect, probably a significant one, as does unused die area. They most likely aren't linear effects, either.
There's so many non-linear and inter-dependent effects from different design parameters, that it's amazing we have two companies with different design constraints consistently make products so close to each other in performance.
There can be n number of factors but the main point is 780/Ti clocks much better than it's less denser GK104 counterparts.
Speaking of shifting goalposts I remember not too long ago the idea of comparing aftermarket 780s against reference 290x was considered poor form.. Yet here we are.
Is that really true though?
I see plenty of 1200/1300 even 1400 GK104s, they just don't go anywhere because of their memory bandwidth.
No surprise here, the 780ti also has a $200 higher MSRP, because it is the halo part nvidia has always held for a long while now.
Fully unlocked 2880 shader GK110 is the best flagship nvidia has ever released. It's a better improvement than the prior best flagship improvement they had with the 8800GTX. It also is relatively cool running and low on power consumption if you leave it running in the stock voltage level.
Overclocked 780ti is faster than 680SLI, it's very impressive. Now if only my EVGA stepups would get to the front of the queue
That's like....a free step up , isn't it? 780 classy to 780ti? Not a bad deal.
I really wish aftermarket cards were an option to upgrade to with the step up program...
Speaking of shifting goalposts I remember not too long ago the idea of comparing aftermarket 780s against reference 290x was considered poor form.. Yet here we are.
Which Card would be best for 1440p Gaming?
Which Card would be best for 1440p Gaming?
Which Card would be best for 1440p Gaming?
Which Card would be best for 1440p Gaming?
Obviously the 780Ti.
So for $185 less then the GTX 780Ti and $460 less then a Titan the R9 290X performs equal when gaming at 1440p.780/ti, titan, 290/x, choose your poison.
Not much when it comes to OC'g but you can't dismiss the advantage of the 512bit bus on the 290x along with it's 4GB's of vRam has when it comes to 1440p res gaming.What does going to 4k do to this comparison? Does the 512bit bus on the 290x help at that res?
Which Card would be best for 1440p Gaming?
Which Card would be best for 1440p Gaming?