GeForce GTX 1180, 1170 and 1160 coming in August. Prices inside

[sze5003]

Boost for a gtx1080ti is 1580 going by its specs. So the boost for both the 1080ti and 2080ti are about the same.

Yea the boost on my current card is 1708 in OC mode granted I never run it in that mode. Doesn't look like it's that much more powerful just by looking at the numbers. Will be curious to see performance results, maybe waiting is not a bad idea after all.

 

[Konan]

Judging from the PNY 1080 Ti prices on their site I'm thinking that RTX 2080 Ti price is a placeholder (hoping)

 

[IEC]

My worst 1080 Ti boosts to 1860+.

Best one boosts to around 2000.

I'll wait for 7nm.

 

[sze5003]

There will probably be other AIB models that boost more but as of those numbers I'm not really that much more excited. It's just the way they are marketing it since the 2080 should be around $700 it makes sense to get those people that usually spend big a small bump up otherwise if the Ti is too much performance, no one would buy the 2080 when they can spend a bit more for 40-50% more performance. And usually pny cards are the more expensive ones so I expect other models to be a little less but 1k seems like a good ballpark.

 

[happy medium]

If the 2080ti is only 21% faster than a gtx1080ti, id be shocked.

Its not bad ,but not what I was expecting. I was thinking 30% or more.
I guess it all comes down to price.
$1000 for 21% more performance? That don't even sound right.

 

[happy medium]

Nice little article over at hardocp comparing the gtx780 with the 780ti, the ti was on average 22% faster. Gtx980 vs the 980ti was on average 30% faster, and the 1080 vs the 1080ti was 32% faster on average.
https://m.hardocp.com/article/2018/08/16/nvidia_gpu_generational_performance_part_3/17

Lets hope the gtx2080ti is 30% faster than the gtx2080 also.

 

[moonbogg]

If the leaked price is correct, it looks like Nvidia will simply try to turn their Ti buyers into Titan buyers. I expected it due to the lack of competition. They keep doing this thing since 600 series. We'll see how it goes, but if it's a grand, then they've priced me out. Does it really make sense for them to force their high end buyers down an entire performance bracket? Forcing Ti buyers down to the X80 class has got to be a little irritating for a lot of people, myself included. I just won't spend that much on a GPU. It has to be around $6-700 and then add an extra $150 for the water block.

If they release the Ti at the same time as the X80 and simply increase the price of the Ti to $1000 (lol) then something feels very messed up there, haha. I get the lack of competition and this doesn't surprise me, but still, it seems crazy. I guess they will all sell out anyway.

I am honestly willing to bet that if Nvidia increased ALL of their prices across the board, including the X60 and below, by 50%, they would all sell out instantly anyway. Charge people like $400 for the GTX 2060 and I bet they all sell just the same. Crazy. I am starting to find these price hikes making the hobby simply not worth it for me.

 

[sze5003]

Yea I'm hoping it's at least more than 20% increase in performance as I'm not inclined to spend 1k for that small of a gain. If the 2080ti is 1k then how much is the titan?

At what point do people say alright I dont care about having the highest tier card? I guess they can push it until they see what the market will bear.

 

[Hail The Brain Slug]

It doesn't matter if the 2080ti is $700, if the performance improvement is only ~25% as I suspect based on the core count increase, the clockspeed decrease, and the IPC improvement (barring RTX performance), I'm not going to consider upgrading anyway.

Give me +50%-70% like the last two generations and I'd buy it in a heartbeat. I don't think that's even in the realm of possibility at this point.

 

[SteveGrabowski]

If the 2080ti is only 21% faster than a gtx1080ti, id be shocked.

Its not bad ,but not what I was expecting. I was thinking 30% or more.
I guess it all comes down to price.
$1000 for 21% more performance? That don't even sound right.

That would be horrible considering the 1080 was 37% faster than the 980 Ti in techpowerup's testsuite at 1440p

 

[sze5003]

It doesn't matter if the 2080ti is $700, if the performance improvement is only ~25% as I suspect based on the core count increase, the clockspeed decrease, and the IPC improvement (barring RTX performance), I'm not going to consider upgrading anyway.

Give me +50%-70% like the last two generations and I'd buy it in a heartbeat. I don't think that's even in the realm of possibility at this point.

It's more likely to be closer to $1k for close to 25% if all this is true. $700 is more doable for only that small gain and I'd maybe be ok with that seeing how a 1080ti right now is around $650, at least one of the evga models.

But sure seems odd that past years gains in the one tier up offered more value for your money.

 

[tamz_msc]

It's pretty simple really- the Titan V had 42% more CUDA cores than the 1080 Ti, yet HardOCP testing showed that typically it was only around 30-35% faster than the 1080 Ti FE. There's a utilisation issue with having all those cores, so Turing drops some cores and perhaps has some underlying tweaks at the SM level to make up some of the deficit. This probably results in ~30% gains over the 1080 Ti that we're talking about.

 

[moonbogg]

It doesn't matter if the 2080ti is $700, if the performance improvement is only ~25% as I suspect based on the core count increase, the clockspeed decrease, and the IPC improvement (barring RTX performance), I'm not going to consider upgrading anyway.

Give me +50%-70% like the last two generations and I'd buy it in a heartbeat. I don't think that's even in the realm of possibility at this point.

I also won't consider ray tracing performance in my upgrade considerations. I don't care if it's 25x faster in ray tracing if it's only 20% faster in everything else. My 1080Ti is about as fast, and often faster, than my dual 980Ti's I used to have. That's a good upgrade.

I don't like price increases, but I do love the social/market experiment Nvidia is doing. It's actually very interesting. I love seeing what people's limits are for a product of this nature. $1000 clearly wasn't enough for Titan buyers, so now they get to spend $3000. That's still not enough, so I do expect them to push the Titan pricing pretty soon. Now previous $1000 Titan buyers get less for their $1000 with the Ti card, and the Titan buyers who can hang in there will get the good stuff for $3000.

They have a price for every buyer, from sub $200, to $300, $4-500ish, $6-700ish, $1000ish (if new Ti rumors are true), $3000ish for Titan buyers, and likely with the advent of maybe a new SLI Nvlink thing, Titan people would easily buy two of those cards or two 2080Ti's if Nvlink worked well.

My limit is $700ish for the card, and it won't be a mid range X80. That's for the Ti or I just sit the round out no problem. I can't trust the naming scheme either though. Nvidia is likely to brand the mid range product as the Ti now. They'd do it easily.

 

[Hail The Brain Slug]

It's more likely to be closer to $1k for close to 25% if all this is true. $700 is more doable for only that small gain and I'd maybe be ok with that seeing how a 1080ti right now is around $650, at least one of the evga models.

But sure seems odd that past years gains in the one tier up offered more value for your money.

That's what I'm saying: Even if they didn't increase the price, the expected performance gain is not worth it to me. Only if the price stayed in the same neighborhood and they provided the same gain we saw the previous two generations would I upgrade without hesitation.

At this rate, no.

 

[happy medium]

It's pretty simple really- the Titan V had 42% more CUDA cores than the 1080 Ti, yet HardOCP testing showed that typically it was only around 30-35% faster than the 1080 Ti FE. There's a utilisation issue with having all those cores, so Turing drops some cores and perhaps has some underlying tweaks at the SM level to make up some of the deficit. This probably results in ~30% gains over the 1080 Ti that we're talking about.

The Titan V was not optimized for gaming, thats a bad example.

 

[beginner99]

So basically we went from x80 being full die to x80 being cut-down second tier die all while hiking up the price. Add to that the RAM cartel and it's save to say this is the worst time for pc gaming in decades. At this point I rather get a console. NV sure isn't getting my money.

 

[tamz_msc]

The Titan V was not optimized for gaming, thats a bad example.

It is a perfectly good example - for 42% more cuda cores you got around 35% more performance than a 1080Ti. In Kingdom Come Deliverance you got up to 40% more performance fully maxed out at 4K. Titan V is fairly representative of what a linear increase of cuda cores would translate into real world gaming performance. That's a starting point in analyzing what a 2080ti could do with reduced cuda cores but tweaked architecture to improve utiliziation.

 

[tamz_msc]

So basically we went from x80 being full die to x80 being cut-down second tier die all while hiking up the price. Add to that the RAM cartel and it's save to say this is the worst time for pc gaming in decades. At this point I rather get a console. NV sure isn't getting my money.

It isn't cut down by that much - 2944/3072 for the 2080. 4354/4608 for the Ti.

 

[Head1985]

My worst 1080 Ti boosts to 1860+.

Best one boosts to around 2000.

I'll wait for 7nm.

reference 1080TI runs around 1700Mhz in reviews.Reference 1080 runs around 1850Mhz in reviews.

 

[Head1985]

So basically we went from x80 being full die to x80 being cut-down second tier die all while hiking up the price. Add to that the RAM cartel and it's save to say this is the worst time for pc gaming in decades. At this point I rather get a console. NV sure isn't getting my money.

Yes fermi GTX580 was full BIG die priced at 500USD.Today we get basically GTX560 which was cutdown GTX560TI and that GTX560 is priced 700+ and sell like RTX2080.

Btw GTX560 MSRP was 199USD
https://www.techpowerup.com/gpudb/287/geforce-gtx-560

And its only 7years since fermi.Nvidia is far worse than intel when intel have monopoly....

 

[alcoholbob]

It is a perfectly good example - for 42% more cuda cores you got around 35% more performance than a 1080Ti. In Kingdom Come Deliverance you got up to 40% more performance fully maxed out at 4K. Titan V is fairly representative of what a linear increase of cuda cores would translate into real world gaming performance. That's a starting point in analyzing what a 2080ti could do with reduced cuda cores but tweaked architecture to improve utiliziation.

I think the problem with the Titan V was memory bandwidth. It has 42.8% more CUDA cores than 1080 Ti but only 35% more memory bandwidth, which is a 6% reduction on a per-core basis.

If you look at GamerNexus Titan V Hybrid overclocking numbers, memory overclocking increased framerate at almost the same linear rate as core clock overclocking, which never was the case with Pascal. Remember the V100 has 900GB/s of memory bandwidth compared to the Titan V which is gimped at 652GB/s.

 

[Ottonomous]

EVGA 2080 'XC Ultra' smile for the camera

https://wccftech.com/beautiful-nvidia-geforce-rtx-2080-evga-xc-ultra-graphics-card-pictured/

 

[alcoholbob]

Hopefully 2080 Ti will have a dual fan ACX style solution that's the size of a regular reference card (i.e., 10.5 inches or less). I can't fit these monstrous triple fan cards into my Silverstone SG13...

 

[moonbogg]

2080Ti, oh thank god it's only $900. Whew.

 

[RichUK]

I am running a 1060 GTX 6GB - how much faster will the RTX 2080 Ti be?

I recently bought an Asus 144hz 1440p G-Sync monitor and get about 50fps at max setting in PUBG at 1440p (OW is typically 110 FPS).