Power Consumption and Sound Testing
Well that's interesting, the power consumption on the GTX 770 reference card is actually lower than the original GTX 680 reference card by a few watts. While that might be counter intuitive you should assume that NVIDIA has done some PCB improvements along with process technology improvements on the GK104 GPU over the course of the last year. Compared to AMD's Radeon HD 7970 GHz Edition though, the power consumption gap is very apparent – the AMD card pulls 73 more watts in our testing!
Our multi-GPU results showcase this even more dramatically with the HD 7970s in CrossFire pulling an additional 156 watts compared to the pair of GeForce GTX 770 cards in SLI.
Because the reference GTX 770 is using the same cooler as the GTX Titan and GTX 780, the sound level results are very similar; much better than both the GTX 680 and the awful HD 7970 GHz Edition reference sound levels.
Ryan nice review have you
Ryan nice review have you tried this with a 680? http://www.reddit.com/r/hardware/comments/1e8u6i/flash_your_geforce_gtx_680_to_the_unreleased_gtx/ it seams very interesting
I just wanted to point out
I just wanted to point out that the link “If you don’t need the example graphs and explanations below, you can jump straight to the benchmark results now!!” still links back to the 7990 review where it was first used. This was also true in the 780 review from last week. I’d think it should jump to the results for the current review.
Great write up as always Ryan. Makes me even more eager to see Haswell numbers next week so I can decide whether, after 5 years, my i7 920 finally should be retired…
I’m like you. I’m still
I’m like you. I’m still wondering when I’ll benefit from upgrading from my i7 920, and it looks like my pair of GTX 580s will still do ok for another year or more.
Thanks for the heads up,
Thanks for the heads up, fixed!
Me, too! My i7 920
Me, too! My i7 920 overclocked to 3.8ghz still doesn’t choke (CPU wise) at whatever game I throw at it.
The power consumption data is
The power consumption data is for the whole system, right? not graphic card-only.
I see EVGA have announced their lineup as well ( http://eu.evga.com/articles/00748/#3773 ), and most of their models specification say 254mm in length, which would fit perfectly in my SilverStone SG05… as opposed to other models. I hope their specifications are accurate.
Maybe we will see some reviews on the non-reference models, as well as EVGA’s ACX cooler, later in the week.
Yes, the power results are
Yes, the power results are full system power draw, at the wall.
Ill be honest 680s never
Ill be honest 680s never coming down in price, even now and 770s coming in at a lower price and faster leaves a sour taste in my mouth.
Eager to see how 770 in sli
Eager to see how 770 in sli compare to Titan and 7990. Considering it would be a cheaper alternative to them.
Great review as always. If I
Great review as always. If I could say one thing, it would be to change your FPS graphs to Bar graphs. So much easier to read.
Any concern with having this
Any concern with having this high end of a card only have 2GB of memory? I wonder if that will become a limitation in the next year or two, especially with 1440p becoming more and more common.
Awesome, they basically
Awesome, they basically lowered the price of a non reference 680 and called it a 770. A $50 price drop constitutes a Gold award? Maybe I just fail to see the innovation here. I just think the next generation of cards should be more than a copy and paste, seems some what bland for such a high award.
Again, I know your reviews
Again, I know your reviews are great, but I can’t help but just skip to the conclusion every time due to the lines instead of bar charts (useful to show how long or how many times a card goes below 30fps etc but harder to read). Is it really so difficult to put up a bar chart also like every other site for those of us who hate trying to figure out all the lines? I’m not saying remove them, just add to them with charts. Not knocking the review though ryan, they are the most informative out there IMHO (along with hardocp who shows max you can expect to run at in nice details). Just dislike the missing bar charts with mins+avg showing.
For the whiners this is a REFRESH, not a next gen. It’s faster, quiet, less heat and cheaper. For the 2GB whiners, you’ll be spending all day running at under 30fps min anyway no matter how much memory you have without 2 cards or a Titan/dualgpu card.
Look at all the reviews that show ABOVE 1920×1200 (1440p etc) WITH minimum fps showing. Tombraider, hitman absolution etc all hit BELOW 30fps so who cares about 2GB? You need to spend over $700 to run there. Heck, hardocp had to turn stuff down on GTX 780 even at 1920×1080! 7970, 680, 770 etc are all 1080P cards at best and many games need details turned down to run there. Discussing memory is pointless if you end up below 30fps to prove it’s useful.
It seems to me that they are
It seems to me that they are attempting to approach benchmarking from a different mind set than you seem to. They have said this many times since the start of developing this FCAT system that FPS is a terrible way to measure a GPU’s performance. They probably do not show you an average FPS, as a way to make you look at the graph instead. It doesn’t matter much if the average FPS is better on one card or another, if it is not consistent and smooth.
There are lots of other sites who will show you FPS as the main indicator of how good a card is. This site attempts to measure how smooth the experience is. The thing that actually matters.
I find this system pretty interesting. Of course I also look at other sites as well.
Ryan, how about the new
Ryan, how about the new display overclocking, can you test it with observed fps that you already developed ?
see if it can resolve our problem with tearing and stuttering
It’s bad when the GeForce GTX
It’s bad when the GeForce GTX 770 can out preform the Tahiti-based HD 7970 GHz Edition in crossfire because of the Micro Shuttering, of course at resolution 1920×1080 but come on AMD get that Shutter problem under control.
On the flip side I would never use FRAPS not alone pay for it to benchmarking my system.
Thanks 4 the Review PCPer I’ll be sticking with my EVGA GeForce GTX 670 and dropping another one in for SLI, knowing that Crossfire has Micro Shuttering problem in software and Nvidia uses hardware that’s build right in for such problems.
P.S I only use Nvidia Products!
Anthony. although Troy`s
Anthony. although Troy`s story is incredible… I just received a gorgeous Honda NSX since I been making $9717 this last 5 weeks and-just over, $10k lass-month. it’s realy my favourite work I’ve ever done. I started this three months/ago and immediately startad bringin in minimum $80.. per hour. I work through this link, Bow6.com
what I want to know is can I
what I want to know is can I drive 4 different desktops, extended desktops, off the one card, I’m not gaming, more like video on one, and browsers and so on on the others…
WHat would be the best card for 4x1080p monitors?