Gigabyte GTX 980 G1 Gaming 4 GB

[Thuban]

We will not include synthetic benchmarks in VGA reviews. Only stuff you actually play.

The idea is to compare results with your current rig. With synthetics it's easy to replicate (since no default benches exist in most games). Simple is best.

I'll be starting a full rebench soon, with latest drivers and AC Unity + Far Cry 4. Gonna take me like 2 weeks non-stop.

That is great news.

 

[the54thvoid]

The idea is to compare results with your current rig. With synthetics it's easy to replicate (since no default benches exist in most games). Simple is best.

Synthetics mean nothing. It's known that certain GPU's are tailored for benchmark results - they don't give a great true to life idea. The only meaningful bench is an in built game bench (few and far between).

I can also tailor drivers to a specific bench that make the card look awesome yet it has fail sauce in game. Benches aren't useful except for playing overclocking games with.

Also, I don't trust other web sites for 'sound' reviews. A lot of them talk absolute shite. I've had cards that some sites (yes, Guru3D is one) say are silent and the cards are anything but. Fact is - you need to do a meta analysis of reviews to get a sound idea about a card.

 

[Cursed]

Hey W1zzard.
I've been reading and loving your reviews for years now, and It helped me to choose my 6850 Cyclone years ago actually. ;]
Now I'm looking for better performance without sacrificing my silence.

I've seen a lot of reviews of GTX 970/980 lately and it occurs to me that dB/temp testing metodology is quite imperfect.
It seams to me that way better test for this card would be to check if it can run at idle fan speeds (locked with afterburner or some other utility) while under stress and keep temperatures under what is considered too high (whatever that is for those new cards).
I don't think any fan of silent gaming lets his gpu fan run however it wants to....

What do you think?
Thanks for your consideration.

 

[W1zzard]

The temperature limit (when Boost starts to not Boost any higher), is 82°C, so all of those cards are perfectly fine. Go buy the ASUS Strix or MSI Gaming, if you don't like it I'll pay your restocking fee.

 

[Jordi]

For me it is pretty simple. A review should be correct in every way and ppl should decide for themselfs if they only read the conclusion or the whole review. Either way it should represent something. When the reviewer doesn't give any value to his "number" just don't mention it. In the time I went to school and I got a 9,7 for something it meant I was close to perfect except for a very small flaw. I read stuff in this review that can't be named a very small flaw... concidering the whole picture (price, top-end hardware etc...).

Enough said from my part. Not my website and when not open for any comments, I don't mind.

 

[W1zzard]

How do teachers grade essays? How come that different teachers would give different grades for one and the same essay?

 

[1nf3rn0x]

There seems to be lots of negativity in this thread. A number is a number and it is all subjective. The information required and needed for a consumer to evaluate whether they want the card or not is there. The rating is all in compliance with other items that have been reviewed. Ok so the numbers are high but it shouldn't matter since you as a reader should be able to 'decipher' and factor the things you do and do not want such as sound or temperatures.

You guys are being too harsh! W1zz does some of the best if not the best gpu reviews out there

 

[Cursed]

The temperature limit (when Boost starts to not Boost any higher), is 82°C, so all of those cards are perfectly fine. Go buy the ASUS Strix or MSI Gaming, if you don't like it I'll pay your restocking fee.

Umm, thx that won't be necesary
I guess I'll ask someone who owns those cards how they deal with stress temps on lowest fan speed.

Just though you could make your reviews bit better....

 

[Thuban]

Synthetics mean nothing. It's known that certain GPU's are tailored for benchmark results - they don't give a great true to life idea. The only meaningful bench is an in built game bench (few and far between).

Games can just be as tailored to a specific vendor. Synthetics is just another benchmark, but unlike games it can be easily replicated by anybody outside of your focus group.

But that's okay, with a bit of googling I can find those synthetic benchmarks elsewhere. Thanks.

 

[erixx]

I own a 980 so i am not a target for this article. Just want to notice there is a verb missing in the very last sentence: "which would also [BE] the case with another GTX 980, or a GTX 970."
Great work as usual Wiz!!!!!

 

[THE_EGG]

wow wtf Gigabyte haha. That is really disappointing performance from that cooler. Makes me wonder if I should have bought MSI or an Asus Strix for my 970s.

Great review as always W1zzard Although for such a loud card I'm amazed at the fact it received such a high score.

 

[bubbleawsome]

Something I've realized about the gigabyte coolers is they can get a lot quieter with minimal impact. I dropped my 280x from the normal fancurve that maxed at 65% fan to one that maxed at 31% and temps only went from 68c to 75c. I'll admit that is 7c, but it went from audible to totally silent from the rest of my case. The only other fans I have are the kind that come with the T40 cooler from NZXT and you can see their reviews here on TPU! I think this is a case of unoptimised fancurves.