ASUS ROG GTX 780 Ti MATRIX Platinum Video Card Review @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,634
ASUS ROG GTX 780 Ti MATRIX Platinum Video Card Review - In today's evaluation we are breaking down the ASUS ROG GTX 780 Ti MATRIX Platinum video card. We put this head-to-head with the ASUS ROG R9 290X MATRIX Platinum. Which provides a better gaming experience, best overclocking performance, and power and temperature? Which one provides the best value at these high-end price points?
 
Maybe I missed it but I didn't see a comparison between the 780ti and the 980 until the closing comments. I'm guessing the review was too far done to add it once the NDA dropped? Good stuff all around tho
 
Or better yet, as a company, why put this card out at this point in the cycle? Seems like a bad move, the folks who would spend that much are the folks that read reviews like this. I guess thats why I don't sell GPUs tho
 
Why was it not compared to the evga 780ti classified.......is the real question

And why does Asus always release their best cards toward the end of the generation!
 
Why was it not compared to the evga 780ti classified.......is the real question
Don't think evga sends them stuff for reviews. Shame since the classified and kpe editions were arguably the best 780ti's for overclockers.

Excellent review nonetheless, though the 980 makes this card a hard sell.
 
Hmm?:confused:
what is this 780ti:D
I've heard about 980/970(which no review on H btw:eek:)
 
I had to read this article with another tab open to do the comparisons with the 980 review (done just before this one on H). With the 980 out at a lower price point, these ROG GTX cards seem pretty irrelevant.

With that being said, I managed to get a great deal on a ASUS 680 Direct CU just as they were going out of stock. At one time (not all that long ago) that was one of the best 680s around. I'll keep my eyes open :cool:
 
  • Worse performance than GTX 980
  • Worse overclocking than GTX 980
  • Higher power draw than GTX 980
  • 25% less VRAM than GTX 980
  • $630.00

nicolas-cage-laughing.gif
 
Was Asus taken by surprise with the release of nvidias's new cards, or did they just take too long making them? I'm curious as to the story behind these matrix cards...
 
I think the most surprising thing in this review to me was that 3GB of vram seems to be good enough now in Watch Dogs for Ultra textures. I guess that means they've "fixed something"? If I hadn't already played it on my PS4 because I got my hands on a cheap used copy...

Anyway, the card itself definitely leaves one wondering why now? Seems a little late in the game for this one Asus hehe... but I'd still be curious to see how it stacks up against an OC'd 980.
 
I think the most surprising thing in this review to me was that 3GB of vram seems to be good enough now in Watch Dogs for Ultra textures. I guess that means they've "fixed something"? If I hadn't already played it on my PS4 because I got my hands on a cheap used copy...

This. [H] was pushing their conclusion of VRAM being the main factor in Watch Dogs very aggressively until now. I actually tried to point out earlier that the stuttering while driving was there for Nvidia users even when the game wasn't hitting the VRAM limit. And now, after the game and the debate over it is pretty much gone, a 3GB card is quoted as giving the superior gaming experience?

Quote from the 290X Matrix article (Sep 8th 2014):
The ASUS ROG R9 290X MATRIX-P averaged 43.5 FPS, which has the ASUS GTX 780 Ti DC II OC running 13.3% faster. Despite this clear framerate advantage, the overall gameplay experience was better when playing on the ASUS ROG R9 290X MATRIX-P. Even with 3GB of memory, the ASUS GTX 780 Ti DC II OC experienced consistent stuttering as a result of "Ultra" textures. The chopping was the most notable when driving at high speeds into new areas of the open world.

Quote from this article (Sep 22nd 2014):
At these settings the ASUS ROG GTX 780 Ti MATRIX-P averaged 50.5 FPS out-of-box. This is 16.1% faster than the ASUS ROG R9 290X MATRIX-P that averaged 43.5 FPS at its stock operating speed. During our gameplay, we did experience a few cases of "stuttering" that occurs due to "Ultra" textures loading. These are represented on the graph by downward spiking framerates that drop under 40 FPS for a split second, before shooting back up a steady framerate.

This stutter occurred a few more times on the 780 Ti than it did on the R9 290X. The frequency of the stutter was low, becoming most noticeable when traveling at near top-speed through the middle of the city. It did not detract from the overall quality of gameplay.

After applying our max overclock to the ASUS ROG GTX 780 Ti MATRIX-P, we experienced a 5.1% performance boost. This improved our framerate to 53.1 FPS. The stuttering still occurred at a low frequency that did not vastly disturb our overall gameplay experience. The ASUS ROG R9 290X MATRIX-P received a 5.7% performance boost from its highest stable overclock, bringing its framerate up to 46 FPS. The stuttering was slightly less severe on the R9 290X, compared to the GTX 780 Ti. That being said, the ASUS ROG GTX 780 Ti MATRIX-P with its max overclock performed 15.4% faster than the R9 290X running its highest overclock. We conclude in Watch Dogs the ASUS ROG GTX 780 Ti MATRIX-P provided the better overall gameplay experience.

So what exactly happened in between there? The VRAM capacity of the 780Ti didn't change, that much is clear. So... it wasn't a VRAM thing?
 
This. [H] was pushing their conclusion of VRAM being the main factor in Watch Dogs very aggressively until now. I actually tried to point out earlier that the stuttering while driving was there for Nvidia users even when the game wasn't hitting the VRAM limit. And now, after the game and the debate over it is pretty much gone, a 3GB card is quoted as giving the superior gaming experience?

Quote from the 290X Matrix article (Sep 8th 2014):


Quote from this article (Sep 22nd 2014):


So what exactly happened in between there? The VRAM capacity of the 780Ti didn't change, that much is clear. So... it wasn't a VRAM thing?

Basically, we are trying to deduct from actual gameplay which of the video cards feels the smoothest and most consistent. In the 290X review the ASUS 780 Ti (non Matrix version) did perform faster, but seemed to stutter much more than the R9 290X. The ASUS 780 Ti Matrix video card stuttered less than the ASUS 780 Ti (non Matrix), despite having the same amount of VRAM.

Our experience with the texture stuttering was about the same on both the ASUS 780 Ti MATRIX and the R9 290X MATRIX video cards. Since the stuttering was identical between these two specific video cards, performance obviously favored the ASUS 780 Ti MATRIX. The shifting of the sand in this scenario allowed the 780 Ti MATRIX the more impressive performance.
 
Basically, we are trying to deduct from actual gameplay which of the video cards feels the smoothest and most consistent. In the 290X review the ASUS 780 Ti (non Matrix version) did perform faster, but seemed to stutter much more than the R9 290X. The ASUS 780 Ti Matrix video card stuttered less than the ASUS 780 Ti (non Matrix), despite having the same amount of VRAM.

Our experience with the texture stuttering was about the same on both the ASUS 780 Ti MATRIX and the R9 290X MATRIX video cards. Since the stuttering was identical between these two specific video cards, performance obviously favored the ASUS 780 Ti MATRIX. The shifting of the sand in this scenario allowed the 780 Ti MATRIX the more impressive performance.

I appreciate the answer and your work, this is a nice review no matter if it's a little late for the contemporary competitive landscape.

I know the [H] way of analyzing things starts with the actual gameplay which is IMHO actually the very best way of grading video cards. I was just curious about inconsistencies in this particular issue. My personal view on the Watch Dogs stuttering issue is based on actual gameplay experience too and so far, I have not found myself in agreement with [H]. Now this article actually draws a very similar picture to what I concluded in August:

http://hardforum.com/showpost.php?p=1041032848&postcount=17
 
After every released patch I fired up Watch Dogs on my PC and still found the stuttering unbearable and gamebreaking (for me) on my 780Ti on Ultra textures.

As for this article, I think readers would prefer reading a review of 970 - a quote from Digital Foundry's review:
The GeForce GTX 970 is that rarest of things in the graphics card market - a genuine game-changer. In fact, it's actually more like a cooked hand grenade strategically lobbed into the high-end GPU market, designed to cause maximum damage to the competition - but in the process impacting just as many Nvidia cards too. It costs £260, which is a lot of money, but its performance per pound ratio is so strong that some might say there's little point considering any other high-end GPU currnetly available - and that includes Nvidia's own flagship GTX 980.
The headlines are this: AMD's R9 290 and top-line R9 290X - £300 and £400 GPUs respectively - are effectively obsolete: hot, power-hungry, inefficient products that are effortlessly bested by Nvidia's £260 upstart. Nvidia is caught in the blast radius itself too - the GTX 780 is history, while the 780 Ti - a card that sold for over £500 just a few months - is also bordering on the irrelevant. Overclocked, the GTX 970 also beats the stock performance of the £430 GTX 980 in most applications too - a phenomenal showing bearing in mind that the latter is 65 per cent more expensive
 
Wow, judging by the number of comments this article is both fascinating and relevant ;)
 
Back
Top