
Article Title: NVIDIA GeForce GTX 780 Ti Video Card Roundup – Tested at 4K
Article URL: http://www.legitreviews.com/nvidia-gefo ... dup_137646
The numbers are the numbers! Trust me when I saw that I don't have time to fake stuff or make stuff up. The sound meter I used has a tolerance of 1.5dB and it is not moved during testing. The cards change, but that is left alone. That means the positions of the fans move and since I'm testing from the edge of the board that matters. Also the types of bearings matter as you can have several fans with different bearings running at x RPM and they will sound different. The fan blade pitch, heatsink, fan shroud and choke whine (if any) also play a role when it comes to idle noise levels. There are a ton of variance possible, so I hope that helps.jodiuh wrote:No mention of the lower RPM's on the MSI.
Something is wrong w/ the noise levels as well. The ACX is spinning 500+ RPM faster than the MSI. The MSI fans are larger as well. Yet they have the SAME db @ load? Also, a 36 db noise floor is quite high. My old apt in the city is @ 25 db. How do I know something is wrong? Simple, I've owned both. The ACX is whiny, loud, and generally a nuisance compared to the sheer awesome that is the TF4. But it makes sense, doesn't it? It is the ACX 1 V the TF...FOUR. Hopefully, EVGA will consider all the complaints from Amazon, Newegg, OCN, etc into their next revision.
Finally, a 12,700 GPU score in Firestrike for a 1300 Mhz 780 Ti is very low. I can hit that @ 1200. Actually, I just ran my MSI GTX 780 Ti Gaming OC @ stock, unmodified by the MSI program speeds and got 12,158.
Simple answer... Space is an issue right now as we have five monitors taking up two 8-foot tables for testing and to move the sound meter off the table would mean we need another tripod and so on. I'm also thinking about doing videos for all sound testing in the future as I think that better shows sound differences than bar charts - http://www.youtube.com/watch?v=cYCSe5MM ... tube_gdatajodiuh wrote:I love how polite you are about the ACX fans. :D
I guess my issue w/ the noise readings is twofold.
Firstly, a 36 dbA noise floor is simply too high. Secondly, why measure so close to the chip? Wouldn't it make sense to test from a few feet away to imitate what it would be like to own this card? I mean...in a case w/ the door closed, etc. Sure, some of us use test benches, shoe boxes, or just lay it out on the desk. But I would bet dollar to dollar that the vast majority of users have their cards in a case.
See Techpowerup's reviews on these cards. They mirror my findings and I tested the same way they did.