MSI GeForce GTX 770 Lightning Video Card Review @ [H]

Devistater

Gawd
Joined
Mar 29, 2001
Messages
651
- Right now Nvidia does have better multi-GPU drivers, although we'll see whether in the future that holds
Hasn't it been that way... for forever? And not just software but hardware too? Nvidia has obviously invested heavily in optimizing SLI and reducing stutter.
 
Joined
Jul 6, 2005
Messages
23
For 570 owners like meself, there was nowhere to go.
I don't think there's anyone who can justify spending 100+ bucks on a 670 over a 7950.

Are you guys shopping around and waiting for deals? I've seen reference 670s going for like $260-$280 brand new a few months back. I know a few people who bought some and the only complaint was that they couldn't get a stable turbo boost above 1,215. Most hit around 1,200 gpu and over 1,700 on mem. That's not bad for the price. Nice bang for the buck.....
 

CAxVIPER

Limp Gawd
Joined
Sep 23, 2012
Messages
339
I really didn't see that big of a difference in performance to warrant another $50
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,942
I think you will be able to watch the "MSRP" price of the 770 stay within a few dollars of the 7970. It will be interesting to see if the 7970 drops further, as surely NVIDIA is watching.
 

raghu78

[H]ard|Gawd
Joined
Apr 7, 2012
Messages
1,401
No overclocking test for a video card reviewed by [H] just doesn't seem right!
ETA for an article comparing a max OC'd 770 to a max OC'd 7970 and 680?
I have a feeling that a custom cooled and OC'd 7970 is faster but i'd be willing to be proved wrong!

yeah take a card like MSI HD 7970 Lightning BE which runs at 1070 mhz core and 1150 mhz boost and it would beat the GTX 770 Lightning even at stock even though it costs 40 bucks more. but the extra 1GB VRAM and the game bundle makes up the cost difference.

http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=8078202&CatId=7387

I noticed the number of games tested is down to 4. Hardocp used to do upto 6 games. games like Hitman absolution, Max payne 3, Sleeping dogs did extremely well on HD 7970 Ghz in previous hardocp reviews.

it would be interesting to see a all out OC review but with haswell, gtx 760 ti and other launches around mostly difficult. but nothing people didn't know. a HD 7970 (1200 - 1250 mhz) would edge out a GTX 680/770 (1300 - 1350 mhz)
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,942
I noticed the number of games tested is down to 4. Hardocp used to do upto 6 games. games like Hitman absolution, Max payne 3, Sleeping dogs did extremely well on HD 7970 Ghz in previous hardocp reviews.


We stick with games that either have huge user bases or are new and graphically demanding, or both. We were kicking around adding Grid 2 today, but the impact on the PC community seems to be minimal so we will likely not be adding it.

So there is no "number" we look to hit as far as how many games, but rather what we see as being important to a broad user base either because of popularity or because of how impactful they are to GPU performance.
 

Brent_Justice

Moderator
Joined
Apr 17, 2000
Messages
17,755
Time is also a factor, these last two launches have not had much lead time, we even had to drop things like OC'ing because of it, OC'ing will be done in a follow-up for the MSI 770 Lightning. Our normal gaming suite for retail games is 5, I like to keep it around 5, 6 being the max if there are games worth it to use. A review with 5-6 games can take 2-3 weeks. Just to clear that up with everyone reading and discussing the amount of games used. Our evaluation method takes time, and is very heavy on workload, there is a reason other sites that have tried to do what we do end up quitting this type of evaluation method.
 

Michaelius

Supreme [H]ardness
Joined
Sep 8, 2003
Messages
4,684
Shogun 2 is kinda boring to test. Engine greatly favours nvidia cards in it.

Also lack of refence cards on sale pisses me off greatly - blowers are great when you have Raven/Fortress 2 type case
 

xoleras

2[H]4U
Joined
Oct 11, 2011
Messages
3,551
Ya, ya thats the reason :rolleyes:

Not sure what your problem is, but I understand their reasoning behind which games are tested and why - it makes complete sense. Not many folks are purchasing 650$ GPUs for Shogun 2 - active hardware buyers have a tendency to be active software/game buyers. So with that being the case, you can assume anyone that is plunking down 650$ on a video card will also be purchasing the most current and most demanding games; current games such as crysis 3, metro last light, tomb raider, among others. THOSE are the games being benchmarked, because those games are very relevant to hardware buyers. Again, active hardware buyers are active software buyers.

Don't get me wrong, Shogun 2 *is* a fantastic game - it just doesn't drive hardware sales as a game such as Metro: Last Light.
 

EdKiefer

Limp Gawd
Joined
Sep 4, 2012
Messages
226
So it looks like MSI no longer supports triple voltage now with GPU-boost2 .
At least you can go to 1.212 but PE/lightening used to go much further with AB .
 

Starrbuck

2[H]4U
Joined
Jun 12, 2005
Messages
2,840
Not sure what your problem is, but I understand their reasoning behind which games are tested and why - it makes complete sense. Not many folks are purchasing 650$ GPUs for Shogun 2 - active hardware buyers have a tendency to be active software/game buyers. So with that being the case, you can assume anyone that is plunking down 650$ on a video card will also be purchasing the most current and most demanding games; current games such as crysis 3, metro last light, tomb raider, among others. THOSE are the games being benchmarked, because those games are very relevant to hardware buyers. Again, active hardware buyers are active software buyers.

Don't get me wrong, Shogun 2 *is* a fantastic game - it just doesn't drive hardware sales as a game such as Metro: Last Light.

I just want to see benchmarks for BF3 and WoW and I'm happy. ;)
 

Burticus

Supreme [H]ardness
Joined
Nov 7, 2005
Messages
4,649
Just checked Ebay and my GIGABYTE GV-N66TOC-2GD 660 TI that I paid $300 and change for last year is selling used for $240ish. Wondering if I need to sell and pick up a 770 for $150 more...
 

Stoly

Supreme [H]ardness
Joined
Jul 26, 2005
Messages
6,714
@Stoly, the original GTX 680 was memory bottlenecked, so to that end, the extra memory is not unwelcome.

Kyle, out of curiosity, will there be an article on overclocking this chip through Afterburner? Hopefully unlike the 680 Afterburner is working out of the box - after all this card is very similar to the 680 Lightning.

Otherwise, pretty nice review. I'm thinking about getting this 770 now.

No it wasn't, it benefited more from gpu OC than memory.
 

MaZa

2[H]4U
Joined
Sep 21, 2008
Messages
3,442
My GTX 570 is starting to show its age and I have been wondering wether I should get another one and go SLI since my mobo supports it, or get a new card. It seems like GTX770 is the answer once the prices stabilize a bit.
 

Derangel

Fully [H]
Joined
Jan 31, 2008
Messages
19,871
Fair enough, but when you reduce the number of games you test, you introduce the possibility of card and game-specific issues impacting your recommendation. It's called idiosyncratic (specific) risk.

The more games you test, the less the ending aggregate can be impacted by issues specific to the cards and games which are tested. This makes sense statistically, as the larger the number of trials in a test, the more the average of the trials has relevance. For instance, you can't predict what any specific person will do, but you CAN predict precisely what a statistically average person will do based on sampling a large number of people.

You can flame me or even ban for me for saying this, but I don't think I'm the only person who has noticed a qualitative bias towards nVidia in the reviews written by [H] as of late. That's completely within your right as reviewers and journalists. After all, what is journalism and writing if not the expression of the writer's opinion. However, QUANTITATIVE bias - which can be introduced by selecting specific games to achieve a desired result - is much more egregious as it's implicit in the testing methodology and not explicit.

I'm not saying that the staff at [H] is guilty of this, all I'm pointing out is that by limiting the number of games you test, you open yourself up for that accusation. And, frankly, that accusation is a logical one which does have merit based on statistics and logic, not opinions. So, if what you say is true and GPU performance is in fact what you're concerned about - you should test a wider array of games. I understand not testing games which are already over 70-80 FPS, but there's plenty of games not on your list which tax GPU's adequately.

Derangel: Games with AMD logos on them have not been known to favor AMD cards specifically. Whereas, nVidia "the way it's meant to be played" games have, historically, been gimped on competing hardware. That's my understanding based on years of researching GPU's.

You've been here six years and accuse them of being biased towards Nvidia? Do you not remember when AMD was the king of performance and how much people bitched about the site being AMD biased? From my experience they care about one thing: Performance. Nvidia is the king of performance. Right now, they have better multi--GPU support (would be nice if that changes soon). Right now, they have the top three best single-GPU cards. And so on. If things were the opposite do you really think they would still be praising Nvidia?

AMD sponsored games lately haven't given AMD a huge advantage, correct. Wouldn't having three fourths of the games used be potentially more balanced or slightly to the Red side point more to them being fair? If they wanted a land slide they could stack the deck with Nvidia titles. You can argue that a larger sample size would be nice, but you have to look at the time commitment and what they need to test. Within a span of a week two new video cards came out. If H ran timed demos and canned benches they could add tons of games, but I'd rather have real world results and an analysis of how well these cards play games beyond FPS. I have tried to test video cards emulating H's style, it is incredibly time consuming even on one card. They have to test every card in every game for every single GPU review. They also have to get reviews done on time for the NDA because the longer it takes the less likely they are to get a huge influx of hits.
 

rainzor

n00b
Joined
May 4, 2012
Messages
18
Well lets take a look at lets say 7850/7870 DC2 reviews from couple of weeks back. Games tested were Tomb Raider+tFX, Max Payne 3, Hitman, Sleeping Dogs....all the games that run better on AMD hardware. Along with FC3 and C3 that are "neutral". Where is nvidia bias here coz all i see are AMD "sponsored" Never Settle titles? If you are bringing up certain conspiracy theories at least back em up with proper facts.
 

PinchedNerve

Limp Gawd
Joined
Jul 30, 2008
Messages
354
Got twin boys who have done very well in school this year. 1 is getting an entire new comp (i5-3570K, Asus P8Z77 & so on) & the other is getting a new Asus 3D monitor with the nVidia glasses. Both will be getting 1 of my EVGA FTW GTX 670's & I just ordered the MSI GTX 770 Lightning for my replacement.

Edit: Correction. I just ordered 2 MSI GTX 770 Lightnings for my replacement.
 
Last edited:

hodor.

n00b
Joined
Sep 10, 2011
Messages
15
Brent/Kyle,

Thank you for focusing your launch-day coverage on a non-reference card. It was surprising to see how many other websites did the full review on the never-to-appear reference 770.
 

xoleras

2[H]4U
Joined
Oct 11, 2011
Messages
3,551
Brent/Kyle,

Thank you for focusing your launch-day coverage on a non-reference card. It was surprising to see how many other websites did the full review on the never-to-appear reference 770.

Agreed. The fact that most reviews among other websites were based on a Titan reference cooler probably creates false impressions - customers likely now want to purchase a 770 with a Titan cooler. And they just can't because no AIB manufacturer is creating such a card. So, kudos to H for not creating false impressions in that respect, totally agreed with you.
 

MacLeod

Supreme [H]ardness
Joined
Jul 28, 2009
Messages
8,054
LOL @ Nvidia bias. Go back and read the Fermi reviews. Only time I've ever seen a reviewer recommend not getting a video card. If they're biased towards Nvidia, they're doing it wrong.
 

w35t

[H]ard|Gawd
Joined
May 29, 2013
Messages
1,349
Man this is tempting at $450. I sure would have bought a 680 lightning for $450 a month ago lol.
 

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
52,942
Brent/Kyle,

Thank you for focusing your launch-day coverage on a non-reference card. It was surprising to see how many other websites did the full review on the never-to-appear reference 770.

We TRY to always do this, but the fact is that many times we have the choice of reviewing the reference card or not having a review up on launch day.

LOL @ Nvidia bias. Go back and read the Fermi reviews. Only time I've ever seen a reviewer recommend not getting a video card. If they're biased towards Nvidia, they're doing it wrong.

LOL @ Shitty post. You can do better than that.

Man this is tempting at $450. I sure would have bought a 680 lightning for $450 a month ago lol.

I think this is what a lot of folks are losing scope of. I can live with it being a "re-badged" and slightly re-specified 680 as long as the price increase shakes out.
 

w35t

[H]ard|Gawd
Joined
May 29, 2013
Messages
1,349
^ Definitely. Plus the stock clock is 1150 and 1201 with boost! I would be shocked if this thing wouldn't do 1300mhz core eeeeasily.
 

Tyler-Durden

2[H]4U
Joined
Nov 18, 2012
Messages
2,302
I've had two GTX 680 Lightning cards fail over the past four months. MSI has been great about RMA replacements, but I'm going to sell my 680 and get a 770. Not sure I'm ready to take a chance on another MSI Lightning despite all the great reviews it's received... but, damn this 770 Lightning looks sweet.
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
I've had two GTX 680 Lightning cards fail over the past four months. MSI has been great about RMA replacements, but I'm going to sell my 680 and get a 770. Not sure I'm ready to take a chance on another MSI Lightning despite all the great reviews it's received... but, damn this 770 Lightning looks sweet.
sell the 680 to get what essentially is nothing more than another 680 with slightly faster memory?
 

Tyler-Durden

2[H]4U
Joined
Nov 18, 2012
Messages
2,302
and perhaps you did not realize that you are posting in a public forum where not every reply will be what you want to hear.
In my brief few months on the forum, I learned to expect the occasional pointless, absurd replies from people like you. I usually just ignore them, but in this case, I’ll play along.

Let’s see… a 770 runs a little faster, a little cooler, a little quieter and costs a little less than a 680. So, why on Earth would anyone choose a 770 over a 680, especially someone like me who has seen two 680s die in his PC over the last four months? Good thing for me that you stopped by to show me the light.

I am in your debt.
 

GoldenTiger

Fully [H]
Joined
Dec 2, 2004
Messages
24,034
In my brief few months on the forum, I learned to expect the occasional pointless, absurd replies from people like you. I usually just ignore them, but in this case, I’ll play along.

I've been down this road before.... it always leads to hilarious responses that eventually just get annoying, by these "people" :eek: .
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
I've been down this road before.... it always leads to hilarious responses that eventually just get annoying, by these "people" :eek: .
says the resident Nvidia troll who defends all things green no matter what. :rolleyes:

anyway my only point was that a 770 is not an improvement over the 680 core clock for clock and only the memory is better. its not worth selling a 680 to get a 770 and even you know that.
 

Tyler-Durden

2[H]4U
Joined
Nov 18, 2012
Messages
2,302
Interesting how people who don't care what the hell I do with my cash just can't resist commenting on what the hell I do with my cash.

I guess everyone needs a hobby.
 
Top