Monday, July 2nd 2018

NVIDIA "GT104" Based GeForce GTX 1180 Surfaces on Vietnamese Stores

A Vietnamese online store put up the first listing of a GeForce GTX 1180 based ASUS ROG Strix graphics card. The store even put out some specifications of the card, beginning with it being based on the "GT104" silicon, based on the "Turing" series. With "Turing" NVIDIA appears to be forking its GPU architectures on the basis of chips that feature DPFP (double-precision floating point) cores and Tensor cores, and those that lack both (and only feature SPFP cores). "Turing" is probably a fork of "Volta" that lacks both DPFP CUDA cores and Tensor cores; and sticks to the cheaper GDDR6 memory architecture, while "Volta" based GPUs, such as the TITAN V, implement pricier HBM2 memory.

Among the specifications of the GeForce GTX 1180 are 3,584 CUDA cores, and 16 GB of GDDR6 memory across a 256-bit wide memory interface. The memory is clocked at 14 GHz (GDDR6-effective), which works out to 409.6 GB/s of memory bandwidth. Pre-launch prices, just like most specifications, tend to be bovine excrement, which in this case converts to a little over USD $1,500, and isn't really relevant. What is, however, interesting is the availability date of September 28.
Source: samcuu (Reddit)
Add your own comment

122 Comments on NVIDIA "GT104" Based GeForce GTX 1180 Surfaces on Vietnamese Stores

#1
bug
RejZoR said:
No it doesn't. New actual models always jump major numbers. Which is GTX 1xxx and GTX 2xxx. With this logic, it should be called GTX 1090 and still make sense somehow as brand new GPU... Which is why I'm questioning this numbering decision if real.
Man you seem to have a lot of of bones to pick with Nvidia lately...
Posted on Reply
#2
cucker tarlson
bug said:
Man you seem to have of bones to pick with Nvidia lately...
Even about naming scheme.
This guy is a baby :laugh:
Posted on Reply
#5
SDR82
Well, the current GPU generation is called the "Geforce 10 Series" (read "ten" and not "one thousand", OK). Before Pascal, Nvidia incremented the number in the series by 100, starting with the Geforce 100 Series all the way up to the 900 series, where that naming scheme ended. Going forward the series numbers will increase by +1 instead of +100. Hence the next generation will be the "Geforce 11 Series".
Posted on Reply
#6
theoneandonlymrk
RejZoR said:
You literally have to ask? 0_o ? When in history has a whole new model only had a second digit increased? Like, ever in the last 2 decades of graphic card models?
480/580fermi to kepler ,680/780 8series got jumped so 780 to 980 kepler to pascal 980/1080 refresh then 1180 there are four hardware iterations ,i get you though but 1000 jump to 2080 is probably overstating the performance uplift for what it is imho.
7nm or not is actually a more interesting question ,since Nvidias naming schemes are total balls these days anyway and are now a unreliable means of identification given their propensity to name two completely different skus the same all the time.
Posted on Reply
#7
Vya Domus
Having nothing to discuss other than naming schemes shows just how "exciting" Nvidia's next generation products are.
Posted on Reply
#8
Live OR Die
Sounds fake when ever has a aftermarket version has been posted up before the reference version, if you look at the 11 in the 1180 the number spacing you can see it's been photoshopped, there to far apart compared to all of the other font.
Posted on Reply
#9
bug
Vya Domus said:
Having nothing to discuss other than naming schemes shows just how "exciting" Nvidia's next generation products are.
Well, the name is all we know. Plus, now we know it's Turing, a relative of Volta. What else would you want us to talk about at this point?
Oh and I almost forgot: this is a xx80 card, so it's not an architecture "built for the mid-range, because that's where most sales happen" :D
Posted on Reply
#10
ppn
Well 1080 Ti has 11GB at 11Ghz and 11Gflops. 1 missed naming opportunity there.
Posted on Reply
#11
Legacy-ZA
All I can say is; the next graphics card I buy. needs 16GB of VRAM and be on the next nm process, be it 7 or 5.
Posted on Reply
#12
Vya Domus
Legacy-ZA said:
All I can say is; the next graphics card I buy. needs 16GB of VRAM and be on the next nm process, be it 7 or 5.
I think "need" is used too liberally in this case. By "need" I understand that it cannot be viable without and nothing really points out to that.
Posted on Reply
#13
cucker tarlson
theoneandonlymrk said:
480/580fermi to kepler ,680/780 8series got jumped so 780 to 980 kepler to pascal 980/1080 refresh then 1180 there are four hardware iterations ,i get you though but 1000 jump to 2080 is probably overstating the performance uplift for what it is imho.
7nm or not is actually a more interesting question ,since Nvidias naming schemes are total balls these days anyway and are now a unreliable means of identification given their propensity to name two completely different skus the same all the time.
Still not as misleading as rebranding 480 as 580 and 290 as 390.
btw AMD did the same thing with rx560, but cut the cards after the initial reviews, and did it secretly. Tell me that wasn't intentional :rolleyes: Just like their fake launch msrp for vega to get better initial reviews. You know, the real shady stuff that adoretv doesn't mention on his channel.
Posted on Reply
#14
ssdpro
RejZoR said:
Yeah, it has gone up by 100 because the numbers were in hundreds, not thousands... Comparing apples to peanuts...
You have a cognitive problem or a math problem, the other users are correct.
580>680>780>OEM>980>1080. +100 equals 1180. Going +100>+100>+100>OEM>+100+100+1000 wouldn't make sense other than pleasing a noob by sounding like an epic jump.

On a more topical note, 409.6 GB/s memory bandwidth from 256 is awesome.
Posted on Reply
#15
qubit
Overclocked quantum bit
qubit said:
Gtxx sounds really cool. :cool:


It makes perfect sense. It went from 9xx to 10xx and now to 11xx. 20xx would be a quantum leap.
RejZoR said:
No it doesn't. New actual models always jump major numbers. Which is GTX 1xxx and GTX 2xxx. With this logic, it should be called GTX 1090 and still make sense somehow as brand new GPU... Which is why I'm questioning this numbering decision if real.
Yeah, it does make sense. It's gone Maxwell, Pascal, Turing. Three different architectures. Why the hell should it jump the numbering system? I think you just like arguing again.
Posted on Reply
#16
cucker tarlson
Vya Domus said:
I think "need" is used too liberally in this case. By "need" I understand that it cannot be viable without and nothing really points out to that.
I think the stuff that comes out in 2019 (new metro etc.) will verify that. I think the jump from 4GB to 6-8 was very significant, but I don't think we'll see midrange cards require 10-12GB too soon.
Posted on Reply
#17
oxidized
This is fake most likely, but i still think it'll be series 11, makes much more sense than 20, although it sounds a bit worse.
Posted on Reply
#18
AsRock
TPU addict
RejZoR said:
No it doesn't. New actual models always jump major numbers. Which is GTX 1xxx and GTX 2xxx. With this logic, it should be called GTX 1090 and still make sense somehow as brand new GPU... Which is why I'm questioning this numbering decision if real.
Well they have nothing to worry about with numbering for many many more years this one, plenty of time in case they think of some thing else up in mean time :P. And will have people understanding the numbering system for a much longer time.

Although how cards are coming out these day even jumping to 1k on each range would take a long ass time.
Posted on Reply
#19
R0H1T
AsRock said:
Well they have nothing to worry about with numbering for many many more years this one, plenty of time in case they think of some thing else up in mean time :p. And will have people understanding the numbering system for a much longer time.

Although how cards are coming out these day even jumping to 1k on each range would take a long ass time.
There's always the XTGTXX route they could take, I mean AMD have pretty much abandoned the ATI love for confusing nomenclature, maybe Nvidia picks it up o_O
Surely Roman numerals are better to understand than the decimal system :laugh:
Posted on Reply
#20
oxidized
I still like better the old naming scheme, the one with GT GTS and GTX but whatever...
Posted on Reply
#21
Octopuss
Rejzor, are you having a shitty day? Why are you caught in endless loop of nonsense arguments about model numbers? You can do better than that :rolleyes:
Posted on Reply
#22
cucker tarlson
R0H1T said:
There's always the XTGTXX route they could take,
Or they could name themselves the name of that rapper.
Posted on Reply
#23
RejZoR
Octopuss said:
Rejzor, are you having a shitty day? Why are you caught in endless loop of nonsense arguments about model numbers? You can do better than that :rolleyes:
No, I just like looking at people going absolutely ape just because it's me typing whatever I type. It's amusing watching people meltdown over absurdities while they think I'm having a meltdown. It's absolutely beautiful.
Posted on Reply
#24
R0H1T
cucker tarlson said:
Or they could name themselves the name of that rapper.
Tbh he has a point about better nomenclature that clearly defines what you're buying, we badly need this in the computing industry when model numbers are recycled like anything! Hey does anyone remember what gen of Core we're on?
Posted on Reply
#25
cucker tarlson
RejZoR said:
No, I just like looking at people going absolutely ape just because it's me typing whatever I type. It's amusing watching people meltdown over absurdities while they think I'm having a meltdown. It's absolutely beautiful.
Except it's always you who sounds like he/she is having a meltdown. :laugh:
Posted on Reply
Add your own comment