Metro: Last Light
Metro: Last Light (DirectX 11)
Beneath the ruins of post-apocalyptic Moscow, in the tunnels of the Metro, the remnants of mankind are besieged by deadly threats from outside – and within.
Mutants stalk the catacombs beneath the desolate surface, and hunt amidst the poisoned skies above. But rather than stand united, the station-cities of the Metro are locked in a struggle for the ultimate power, a doomsday device from the military vaults of D6. A civil war is stirring that could wipe humanity from the face of the earth forever.
As Artyom, burdened by guilt but driven by hope, you hold the key to our survival – the last light in our darkest hour…
Our Settings for Metro: Last Light
At 2560×1440, the Radeon R9 295X2 and the pair of GeForce GTX 780 Ti cards in SLI are running nearly neck and neck with each other, leaving the GTX Titan Z 14-15% behind in average frame rate. Frame time variances are low across the board which results in smooth and consistent multi-GPU frame rates.
When we run at 3840×2160, we see the results from the Titan Z and the 780 Ti cards in SLI merge to a single score, pulling in 45 average FPS. The R9 295X2 is running a bit faster at 50 FPS, a difference of 11%. Frame times are pretty good for all three tested platforms though the NVIDIA solutions had just a bit more variance than the R9 295X2.
more accurately result when
more accurately result when putting 780 ti from 1020 palit
or ghz cause 295×2 is already factory overclocked
Two OC’d Titan Blacks if you
Two OC’d Titan Blacks if you need compute. Two OC’d 780 ti s for gaming. Check out the inno ichill 780 ti which runs OC in the high 50s on AIR!!! AMD can only dream with their power hungry space heaters. Titan Z is for the I don’t care what it costs crowd and Nvidia has been pretty good holding that segment.
2 titan blacks stock is fast
2 titan blacks stock is fast then 1 titan z stock. And the titan z is basically 2 titan blacks in one package.
Any info on usable ram? its
Any info on usable ram? its advertised as 12GB, but is all of that usable or is it merely 6GB mirrored?
I don’t like AMD going
I don’t like AMD going outside the power specifications, but that is the only thing that really makes the product make any sense. Nvidia stayed within the 375 W limit, so they have lower performance than the 780Ti in SLI. Given the high price, there isn’t much of a reason for this card to exist.
These dual gpu cards seem to really only be for marketing, rather than a real product. AMD made dual gpus closer to a real product by actually enabling full speed by using water cooling. It still doesn’t make much sense (higher price for same performance) unless you are building a space constrained system. The Titan Z fails at being a marketing/publicity stunt since it obviously can not compete with the 295×2 in performance. AMD and Nvidia want to stay in the public consciousness, even when they do not have a (real) new product to release. These cards will be super low production volume, but they obviously generate a lot of publicity. Since AMD beat them to market with the 295×2, they do not want any publicity, since this has turned to negative publicity.
These dual gpu cards will not really make sense until they actually connect the gpus together to allow them to share memory and act more like a single gpu. To share memory, they need really high bandwidth interconnect (~100 GB/s), which is probably doable between chips so close together. I have been wondering if they could just shrink the memory controller and use the pins for interconnect instead, or even make them programmable to allow the same chip so be used in a single gpu card with full width memory interface or in a multi-gpu card with narrower memory interface. This would also allow use of less memory. Designing and producing such a design may not be feasible right now due to the low volume nature of the super high-end gpu market though. I don’t think it is that useful for the compute market due the way compute gpus are being used (parallel independent task).
So i am glad i built my dream
So i am glad i built my dream machine with 3 gtx blacks that cost me as 1 z. i know it is overpriced, compared to 780ti, which is also overpriced, but running them underwater gave me huge potential for oc and smooth 4k. it was just a dream machine that i don’t plan to swap for the forthcoming years. To my opinion, graphic cards should not be priced more than 400$ i don’t believe there is no room for profit even for the z if this was the price target.. anyway, just an opinion.
Ahaa, its nice conversation
Ahaa, its nice conversation on the topic of this piece of writing at this place at
this web site, I have read all that, so now me also commenting here.
My site … karnisze nowoczesne
Actually after Watchdogs came
Actually after Watchdogs came out it has shown 3gb at Ultra Settings just doesn’t cut it anymore when developers push the limits.
I’m ready to build my Rig soon but tired of waiting for GTX880 with 8gb. If the 880s are not due out until Christmas I may just go ahead with ASUS Strix GTX780 6gb in SLI.
good test!
good test!
I don’t know who made the
I don’t know who made the decision to price this card at $3000. It can’t even outperform AMD’s $1500 card. This card should be sold for $1500 if not less.
As much as I hate to admit it, I think Nvidia would be unstoppable if it capitalizes on the Maxwell GPUs. The 750Ti is an impressive GPU and that was just the tip of the iceberg
They are just about equal
They are just about equal while the 780Ti’s are running at 876Mhz. Too bad you can easily get cards that run 1300 or even 1400 on water. Just imagine the performance gap they would have with +50% clockspeed on the nvidia cards 😀 What a joke
Note: FLOPS are primarily
Note: FLOPS are primarily important to people mining for Bitcoins or for protein folding. It means very little in the way of gaming.
With that being said, I think the only way that the TITAN-Z would pull ahead of other SLI/Crossfire setups is at 8K resolutions because of its immense 12GB of VRAM and 768-bit bus. Even then, I’m not sure that the ROPs, shaders and texture procs on the TITAN-Z would be able to keep up with that resolution.
I honestly think these cards’ target audience is those planning on mining, and that’s why they justify the exorbitant price tag despite the lackluster gaming performance but with a notable increase in teraFLOPS performance.
I’ve seen a bunch of vendors
I’ve seen a bunch of vendors dropping the price significantly on the TitanZ, Alienware had a TitanZ half off promotion and now CyberPower is giving 500.00 off the price. Has this made this card a better value? Trying to price a new system, have about 4g to spend. I do not want to put it together myself or support it, I am laaaaaazy.