Choosing the right 980 ti

Scroll down on either newegg link you posted, it shows you a chart that shows the difference, that being a 8+2 vs 14+3 power phase. Better power phases usually result in being able to overclock more, hence why it has a higher PSU requirement as well.

As SR-71 mentioned, the EVGA cards are very good.

I'm also partial to the Gigabyte G1 Gaming cards.
https://www.youtube.com/watch?v=B2-gCXscH04
Note: Jay also recommends EVGA, I just couldn't find a good video for an EVGA 980ti...

They will both OC well and have quiet coolers that work keep heat under control and prevent throttling.

Here's how they stacked up in TPUs reviews:

Brand fps OC % Core Memory link

Gigabyte G1 134.8 131.4% 1512 2100 http://www.techpowerup.com/reviews/Gigabyte/GTX_980_Ti_G1_Gaming/33.html
Palit Jetstream 133.1 129.7% 1515 2100 http://www.techpowerup.com/reviews/Palit/GeForce_GTX_980_Ti_Super_JetStream/33.html
Asus Strix 131.7 128.4% 1472 2070 https://www.techpowerup.com/reviews/ASUS/GTX_980_Ti_STRIX_Gaming/33.html
MSI Gaming 130.5 127.2% 1507 2040 http://www.techpowerup.com/reviews/MSI/GTX_980_Ti_Gaming/33.html
Zotac Amp 130.4 127.1% 1465 1990 http://www.techpowerup.com/reviews/Zotac/GeForce_GTX_980_Ti_Amp_Edition/33.html
EVGA SC 126.8 123.6% 1491 1900 http://www.techpowerup.com/reviews/EVGA/GTX_980_Ti_SC_Plus/33.html

Three of them break the 1500 Mhz barrier.... Gigabyte, Palit and MSI. But that isn't the be all and end all as memory OCs affect that too (Note EVGAs poor showing @ 1900) likely due to the lack of memory cooling an the SCs reference PCB). Giga and Palit stand out at the top, EVGA at the bottom and the rest are clumped in the middle but ALL of them are in 27-31% OC territory, except for EVGA.

Couple of things I have noticed since the 7xx series:

1. nVidia has stepped up the limitations on what 3rd party vendors can modify, including both legal and physical limits. Unless you physically modify the card or used a hacked BIOS, what you can do voltage wise is strictly controlled. The advantages offered by the "super cards" like the Classified, Lightning and HOF has therefore been greatly diminished because no matter how hefty a VRM you have, if you can't take advantage of it and raise the voltage significantly higher, the advantages of those VRMS and hand picked GPUs are negated. You can look at the threads over on overclock.net to see that the Lightning, Classified and HOF are just not delivering anything close to what we saw in past years or anything to warrant the price premium.... unless of course you getting into LN2 and / or card mossing.

2. Custom BIOSs can be used to deliver higher base core clocks but, after extensive research with the 780s. I found a lot of BIOSs that delivered higher base cores, but none that delivered higher fps. I surmised that this was caused by the way the boost technology works, in that the higher base core isn't making the boost significantly higher, and even if it does, it's not doing so in a way that brings higher fps.

3. Even water cooling doesn't have the impact it had in earlier generations as we're generally hitting the voltage wall well before we hit the heat wall.

4. Warranty return rates for the major players over the past 12 months have been:

- Gainward 1.56%
- Zotac 1.83%
- ASUS 2.54%
- Gigabyte 2.68%
- MSI 2.37
- Sapphire 3.38%

Aside from the performance, noise and other data you can gather in the above links, I also find reading the user reviews / ratings on newegg to see what % of card owners are really unhappy with their purchase giving it jut 1 or 2 egg ratings.

No , I would purchase any of them , all killer cards.

EVGA is quality when it comes to cards and high end PSU's.

I have found that all manufacturer's support has tanked in recent years but my last EVGA experience involved 20 support calls over 18 months and 5 RMAs.... none of which provided a card that could run at advertised factory overclocked speeds. I appreciated that they didn't blow us off but the process, not much unlike anywhere else, seems to be designed to frustrate the customer and force them into a situation where they feel it's not worth the time and effort. But they took it to a new extreme in that on every call we started from scratch ... on the 20th call it as as if the previous 19 calls did not happen.

They blamed MoBo, memory, PSU and the only way I finally got them to acknowledge that it wasn't anything else was to show that the MoBo, PSU, memory etc was able to handle two Asus cards with a 28% OC (borrowed from an about to be completed user build) and yet their card was failing at "outta the box" speeds.....it failed just 10MHz above reference speeds too. It had gone on so long that the next generation cards were out and they replaced the card with a reference next gen model.

I hope nVidia loosens up the reigns again on what they will let vendors do because they really killed the value of these super cards .... investing in large, beefy VRMs and other components just isn't bringing the returns it used to. At $700, I just don't think it brings an adequate return on investment, especially when twin 970s will be cheaper and faster. But if ya set on a single card, that being the 980 Ti, I'd look at the Amp Extreme and HOF before making final decision as these seem to be favorites reporting the higher OCs and lower noise on sites like overclock.net




The EVGA SC is just a reference card with a nice cooler. I haven't seen a complete tear down article for the 980 Ti but here's a good comparison of the 970s.... you can see the difference on the bottom third of each page.

http://www.bit-tech.net/hardware/graphics/2014/09/19/nvidia-geforce-gtx-970-review/3

EVGA sticks with the reference power input design, equipping the GTX 970 SC ACX2 with dual 6-pin power connectors.

Examining the PCB reveals a 4+2 phase power design – four phases near the rear I/O for the GPU, and two in the bottom right corner for the memory. This is a slight upgrade from the 4+1 stock specification but unlike MSI and ASUS, EVGA does not use any specially crafted components.

The GPU MOSFETs are directly cooled by the main heatsink, which has a thermal strip on to draw heat up into the fin stack. On the other side of the GPU is a metal contact plate that partially cools two of the four memory chips on this side, leaving the other two exposed. It also cools the MOSFETs of the power phases serving the memory, but no thermal pads are used, so heat transfer is likely to be limited.

Click to expand...