• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Geforce Pascal Review thread

As a starter, the 1080 looks very good indeed and as D.P pointed out, a boost like 40%+ is very nice indeed. If I did upgrade, not that I need to, I would be waiting for the Ti model! Impressive non-the less.
 
Going by the 6.75 TFLOP/s on the 1070 against the 8.2TFLOPS/s(from non boost clock) on the 1080 - that's 15%(rounded up) slower than the 1080.

Looking at that link I would have thought it would have been slower.
TFLOPS aren't everything though, as a glance at the Fury X's spec sheet and then its actual gaming performance will attest. It'll be interesting to see how the cutbacks affect real world numbers.
 
I am seriously looking forward to running some side by side bench results now. Pitting it against my TX with max OC on both will be interesting.
 
I'd hardly call it a mid range card. The gtx980 still isn't mid range to me.

I didn't say the 980gtx was mid range. The 980Ti was the previous high end, the Titanx was a halo card.

So effectively two die shrinks later, from 28nm->20(skipped)->14/16 you think that NVidia's high end GPU is going to be just 315mm2? NVidia, the company renound for pushing lithography to it's limits.. OK the node is new, I still expect a few bigger graphics chips from nvidia to come in 16nm. But to say that the 1080 is their new high end card is to have a very short memory as they did a similar thing with 28nm.
 
One has to realize that the 1080 is a mid-range chip being compared to a previous gen high-end chip, and its it beating the fastest GPU of the last gen by 30-42% in most benchmarks. That is in general a higher performance increase than we see in almost every single generation.


When comparing to the actual predecessor, the 980. the 1080 is widely 65->85% faster. The bigger jump in the newer games like hitman or ashes. That is larger jump than many generation changes.
 
I didn't say the 980gtx was mid range. The 980Ti was the previous high end, the Titanx was a halo card.

So effectively two die shrinks later, from 28nm->20(skipped)->14/16 you think that NVidia's high end GPU is going to be just 315mm2? NVidia, the company renound for pushing lithography to it's limits.. OK the node is new, I still expect a few bigger graphics chips from nvidia to come in 16nm. But to say that the 1080 is their new high end card is to have a very short memory as they did a similar thing with 28nm.

16nm is not 2 die shrinks smaller than 28nm, it is 1 generation plus a little extra. 16nm is mostly a marketing term, it is mostly based on 20nm planar. Same with 14nm.
 
Em...different vendors would not have a unified design for the PCB, so how is EK going to be able to design a block that will fit (and work well) on all of them?

In terms of compatibility of block with non-reference card, just thinking back about people asking about watercooling their non-reference card...nobody could really answer for sure if the waterblock would be compatible for cards with custom PCB. Somehow I got a feeling that Nvidia will dictate that AIB partners are not allow to use reference design for their PCB for the custom cards...so the only option is probably custom "hydro" cards with pre-fitted block or go for the more expensive reference card.

Just because the cooler isn't reference does not mean the PCB wont be. I have bought many a GPU from both AMD and Nvidia that are custom cooled and also have reference design PCB. As far as I am aware, that will still be the case (I, of course stand to be corrected) until the AIB's start releasing their premium stuff with different power delivery etc (EVGA Classy etc)
 
I didn't say the 980gtx was mid range. The 980Ti was the previous high end, the Titanx was a halo card.

So effectively two die shrinks later, from 28nm->20(skipped)->14/16 you think that NVidia's high end GPU is going to be just 315mm2? NVidia, the company renound for pushing lithography to it's limits.. OK the node is new, I still expect a few bigger graphics chips from nvidia to come in 16nm. But to say that the 1080 is their new high end card is to have a very short memory as they did a similar thing with 28nm.

16nm is just 20nm with finfets. It still has a 20nm Back end of line. just the transistors are smaller.

But even with the possible 2x density the node can bring, nvidia have not even gone near it with gp106. i think this is more for safety than anything else.
 
I like the performance, but disappointed on the price. AMD need to hurry up with their attempt and hopefully bring the pricing down.
 
I'm not going to bother with the 1080, but I will be considering the 1080-ti.

The performance gain against a very high clocking 980-ti under water doesn't appear to be that significant - once the cost has dropped to sane levels it appears to be a nice cool, low power usage card mind.

I'm going to wait for a single card which can do 4k @ 60fps before I upgrade again, but still - impressive so far.

I like the performance, but disappointed on the price. AMD need to hurry up with their attempt and hopefully bring the pricing down.
That's kinda the main problem, the price per % increase is negligible/negative - if it was significantly cheaper it would be more attractive for an upgrade.
 
One has to realize that the 1080 is a mid-range chip being compared to a previous gen high-end chip, and its it beating the fastest GPU of the last gen by 30-42% in most benchmarks. That is in general a higher performance increase than we see in almost every single generation.
Only on this site can you find a £500+ GPU being called mid-range! It's absurd.
 
Last edited:
u might see the tx prices go back up, just crazy time right now
so you not waiting for custom cards? :o

I would have done but I picked up early on about the temperature clocks and how when the card runs cooler, the 3.0 boost gives better performance off the bat and if the card is hot, it will start to choke. It just makes sense for me to add it into my custom loop that has a 480 rad cooling a CPU only at the mo. I just couldn't be bothered to put my TX under water, even though I had a block (sold it now).

https://www.youtube.com/watch?v=-cLYn937ZFE

HardwareCanucks review Greg, they hit 2135 on the core :)

Cheers and added to the OP :)
 
Back
Top Bottom