GeForce GTX 780 Ti Round Up

The crew at HotHardware have rounded up a handful of GeForce GTX 780 Ti video cards and pit them against one another. The round-up features cards from MSI, Gigabyte and more.

Do you know what gets our juices flowing even more than high-end hardware? It's having the opportunity to wrangle and evaluate multiple versions of a top-tier product. In this case, we've gone out and collected NVIDIA GeForce GTX 780 Ti graphics cards from three different enthusiast brands: EVGA, Gigabyte, and MSI.

Click to expand...

Yes, because testing cards at their stock (even if OC'ed) speeds show the differences... They should have OC'ed them and then see who came out on top. My guess, MSI gets a really nice OC because of the lower heat and power consumption vs. the other two. But hey, doing an OC test would make sense when choosing cards that people buy to OC, so of course they can't do that.

They didn't even compare cooler performance? What a waste.

Pretty poor job of "rounding up" cards... Just 3? Lots more 780Ti offerings out there these days. Comparing them stock for stock is also sort of a meaningless exercise - the ones with slightly faster factory clocks are obviously going to perform better given that the base architecture/chip/mem is identical in all of them. A more interesting comparison would be to take them all up to the max stable overclock of whatever the weakest card of the lot is and then do a comparison. Would provide much more useful info as to how well each card's cooling system /layout performs at a minimum. Then do a max overclock per card comparison. Comparing them at stock settings just isn't all that interesting - many enthusiasts buying top tier GPU's would overclock a bit anyways.

I'd have to agree. Very poor roundup. The most important consideration as an enthusiast with these cards, are manual overclocks. I don't see that in the roundup, and I find that rather disappointing. Stock speeds have their place but there's a significant portion of the buyer base buying custom cards specifically for overclocking. And that's not there in the review. Additionally, while there's a good selection of games - most of them are old. I feel like they could use an updating to their game selection, I quite like what H does in terms of continually evaluating and updating the test games. It gives the readership a good idea of where cards perform in games that are relevant today, games that people are buying today. Not games from two years ago.

All in all, while I hate to be critical, i'm not a fan of that roundup.

Bleh I need to upgrade...

There is one problem with reviewing overclocks, you never know if the overclock achieved is due to card design or a great chip.

rinaldo00 said:

There is one problem with reviewing overclocks, you never know if the overclock achieved is due to card design or a great chip.

Click to expand...

Having a good chip or a bad chip is chance and something you cannot controller, so should have no factor in deciding whether to do a competent review or not.

TraptPatriot said:

Having a good chip or a bad chip is chance and something you cannot controller, so should have no factor in deciding whether to do a competent review or not.

Click to expand...

OK, let me ask you this. Let us say that in the review above the card with the lowest factory OC happened to have a golden chip and the reviewer got 1300 mhz on his overcloock and the other cards only got 1200 mhz.. You buy the 'fastest' card based on that review. When you get your card all you can get is 1200 mhz.

Now tell me, how did the reviewer OC tests help your purchase?

rinaldo00 said:

OK, let me ask you this. Let us say that in the review above the card with the lowest factory OC happened to have a golden chip and the reviewer got 1300 mhz on his overcloock and the other cards only got 1200 mhz.. You buy the 'fastest' card based on that review. When you get your card all you can get is 1200 mhz.

Now tell me, how did the reviewer OC tests help your purchase?

Click to expand...

More than not doing OC tests. Plus I don't look at reviews and think "Wow, that card got to 1300mhz, that must mean every card can." Yea, I don't think so. Over clocking is never the same, but still helps to show what the limitations of cards could be. Plus I look at multiple reviews, and compare the OC results and FPS results to see which ones average out to be the best. In a field where nothing is absolute, averages are key to finding what suits your needs best.

Again, a more interesting comparison would have been to take them all up to the max stable O/C of whatever the weakest card in the group could muster (such that they are all running the exact same O/C) and then do a comparison. Would provide much more useful info as to how well each card's cooling system /layout performs (as to temps and noise) at a minimum.

Yeah what I would like to see is temps at stock/boost clocks, temps with all of the cards at the highest stock "boost clock" of the group. And finally the highest clock they could get each card with a warning that it's mostly luck if they got a good chip or a bad chip, with temps.

Also comparing temps, cooling, and decibel levels.