Since nobody have posted this yet and Im not around much, I thought I should update you on Nvidia`s plan for Q2. I think this might mean we are looking at June ish for Pascal.
Geforce 940MX (No GT or GTX)
Specs unknown but listed in a MSI driver
http://forums.laptopvideo2go.com/topic/31280-msi-mobile-geforce-driver-35930/
GTX 970MX
1408 CUDA cores
941 MHz
192 bit
3 and 6GB GDDR5
Replacing GTX 970M with 1280 cores
GTX 980MX
1664 CUDA cores
1048 MHz
256 bit
4 and 8 GDDR5
Replacing GTX 980M with 1536 cores. Not sure where GTX 980 mobile fits in all of this. Its Nvidia and they confuse everyone with their plans
https://translate.google.no/translate?sl=zh-CN&tl=en&js=y&prev=_t&hl=no&ie=UTF-8&u=http://tech.hexun.com/2016-01-01/181547450.html&edit-text=
-
-
Makes sense. ~4 TFLOPS hopefully for the 980MX.
Thanks for posting.
How come you still have "RIP mobile gaming" in your sig? -
Yeah go figure. 980mx is really not worth to upgrade from decent 980m then.
And the english translated from mandarin is super weird in terms of the phrasing. Just ask me if anyone dont understand part of the phrasing.Last edited: Jan 16, 2016 -
So other than the increased cores, the 980MX isn't that much different than the 980M? What would you all estimate performance difference will be? Less than 10%?
-
Clock per clock, I say at max 6%-7%. Probably even less then that at around 3%-4%. If these specs are true, I am seriously unimpressed by it. Unless it higher clocked VRAM, its not really worth the upgrade from 980M to 980MX.
-
These are the specs of the 970, and considering they still make the 970, let me adjust this for you:
GTX 970 (mobile)
1664 CUDA cores
throttleMHz
224-bit + 32-bit
3.5GB + 0.5GB and 7GB + 1GB GDDR5
Assuming it serious anyway.jaybee83 likes this. -
^Nah, I doubt they are doing that fishy business with ROP and bus since it wasnt done on 980m. Though if they did, it would be absolutely hilarious and sad considering how bad a 980M can wreck it at the same clock.
-
This statement makes no sense though.
980M = 1536 cores, 256-bit mem bus, 4GB/8GB vRAM
970 = 1664 cores, 224-bit + 32-bit mem bus, 4GB (segmented) vRAM
980M is a weaker card, even if 980MX is real and has segmented vRAM.
If 980MX has no segmented vRAM, why is the 970 still sold with its broken design and lies on its marketing? And why would nVidrosoft release the same specs card with the vRAM/mem bus segmentation fixed for mobile, but keep the broken design for desktop? They BLATANTLY lie about things to deal with mobile cards because nobody who does the interviews or writes the articles knows anything, and they can pull the wool over their eyes. Like openly stating that unlocked mobile CPUs never happened before, or that their mobile driver API doesn't support voltage control (because their stock vBIOSes don't allow voltage changes).jaybee83 and Spartan@HIDevolution like this. -
I can see those cards being worse than 980Ms.... lol.jaybee83 likes this.
-
I'm gonna literally lol if the geniuses at nvidro$oft did actually decide it was a good idea to recycle those broken 224+32 bit desktop 970 silicon into the 980MX. Goddamn.
Or how overclocking was a "bug" that was present in their drivers for over 10 years until they finally got around to "fixing it".
jaybee83, deadsmiley, TBoneSan and 2 others like this. -
King of Interns Simply a laptop enthusiast
I really hope AMD can smash Nvidia to pieces this year. Such poor rebrands 2 years after the 970M,980M released.
No sign of Pascal and now this. Come on AMD now is your chance, FINAL chance...TBoneSan likes this. -
Its not just consumers AMD needs to win back, its the OEMs as well. They have an uphill battle.
-
So no Pascal until 2017? I can't see them releasing rebrands and then pushing another round of cards out the door a few months later.
-
If 980mx had that f*cked spec, it will not beat 980m with the same 5ghz vram.
Why does nvidia keep using the BS 56 ROP and 224bit bus on 970? In order to differentate the 980 to 970. When 680/670 first launched, everyone was jumping for the 670s since there is almost no performance difference with only 192 CUDA cores cut. Ngreedia have to cut more out of the chip to encourage sales for 980.
Its absolutely f*cking BS, but its ngreedia. -
It could easily have 7GHz vRAM like a mobile 980 could. In that case, even the maximum OC you could get on a 980M's memory wouldn't win, since they top out near 6GHz. The reduction of so many extra cores should mean that the card wouldn't need special cooling systems or designs to fit in notebooks. You don't know what they might do; hence why I am skeptical.
-
Idk, the way they phrased in mandarin make it sounds like the only difference is the core count. Might be I am slowly getting bad at reading mandarin though but I immediately thought the only difference was the shader count by the way they talked about it.
-
They can say what they want, but until it comes out we won't know. I however think if it comes out and it's not broken like the desktop cards, the resulting backlash that laptops get a superior product will be... well... immense.
-
Desktop gamers pay no mind to laptop graphics so it wouldn't really matter.jaybee83, UnknownCode, Cakefish and 4 others like this.
-
Exactly this.
The reason nVidia gets away with pretty much anything in the mobile sector is because, well, desktop gamers are (for the most part) arrogant, elitist, and ignorant that care about themselves only.jaybee83, Cakefish, deadsmiley and 6 others like this. -
I mean look at the fact nobody really complained when mobile cards got 6GB and 8GB of VRAM when the desktop cards were on 2GB and 3GB. Weren't the 880Ms the first 8GB cards? And the 870M 6GB? Consumer anyway.jaybee83 likes this.
-
You'd be surprised. The amount of people who opened their eyes at mobile 980 was somewhat large. To hear that a mobile 970 has a properly working product when desktop 970 is still broken... there might be some more bad blood.
Of course with all the absolute, complete, utter idiots who defend the 970 in desktop land, it may be less than I expect will happen. But it shouldn't be small at all.
This was when mobile cards had reduced clockspeeds and memory speeds and stuff. If the 980MX fully copies 1:1 a 970 like the mobile 980 does, that's going to be a different scenario. Because it'll mean that it is a superior card which the 970 will never be able to compete with, due to it having a full 256-bit memory bus as opposed to 224-bit + 32-bit. Assuming it isn't cut-down, of course. But yes, people generally ignore laptop-only cards and the problems of their users, I know. -
thegreatsquare Notebook Deity
...At least AMD is somewhat relevant in the desktop market, ...market share sucks, but they have decent enough offerings on the high-end.
I don't know about a desktop 970 with it's VRAM issue in a mobile card. How does that work with the 8GB cards? Is it going to behave like it is 3.5GB+0.5GB+3.5GB+0.5GB and have issues switching and then switching back again or 7GB+1GB and you'll hardly ever hit the 7GB limit where issues start to show up? Plus lower overall VRAM bandwidth is produced when splitting up the bus and mobile cards have had slower VRAM vs desktop as is.
If the 980m had the desktop's 970's memory bandwidth, it would probably match it pretty well. The 980m has fewer TMUs, but more ROPs.Last edited: Jan 17, 2016 -
If this is still 80-100w TDP card from Nvidia then will this be a godsend for Razer, Aw and OEM's who still swear by making only slim design for their BGA gaming laptops. Everyone knows that 100w TDP cards can be used in slim laptops. 150-200w TDP cards as Gtx980 is unacceptable for such baby laptops. This know Nvidia very well, LOL
-
Ionising_Radiation ?v = ve*ln(m0/m1)
I see that the xx60 series are conveniently missing. I believe (accurate or not, I don't know) that we'll see the successor to the 750 Ti and the GTX 860M/960M in March. GP107 or something.
-
killkenny1 Too weird to live, too rare to die.
Must be the same idiots who itch and cry about how nVidia sucks, while still buying their stuff...Talon, i_pk_pjers_i, triturbo and 2 others like this. -
thegreatsquare Notebook Deity
The way many desktop users were full ready to throw laptop users under the bus where overclocking was concerned, I wouldn't expect much support/sympathy of any kind.
-
I can't see Pascal slipping to 2017. This is just Nvidia giving the notebook OEMs what they want: something compelling to sell in this half of the year.
-
No, but Q4 2016 at the earliest if these cards are coming out Q2. That's still another 9-10 months away.
-
Will the 980MX get Nvidia's VR Ready badge?
Sent from my Micromax AQ4501 using Tapatalk -
I don't see why it wouldn't, pretty sure the regular 980M got it already, and this is theoretically more powerful.
-
Actually it wasn't the 980m that got the Ready for VR badge, it was the new Mobile Desktop 980 - at release Oculus gave it the nod.
The 980MX might be fast enough, but we don't know for sure yet
-
Aha so true! So damn true. 99% of desktop gamers have no clue whatsoever about laptop GPUs.
Sent from my E5823 using TapatalkEthrem likes this. -
http://www.bhphotovideo.com/c/product/1213132-REG/msi_cx62_6qd_047us_15_6_cx62_gaming_notebook.html
940MX. Will be released sooner than Q2.Mr Najsman and Ethrem like this. -
I'm pretty sure Gtx980MX is intended for VR. Deliver most likely the same performance as Gtx970. Nvidia will deliver VR features for slim *** laptops
Last edited: Jan 18, 2016 -
So this is essentially what the 780M was to the 680M?
Add +1 SMM (just as 780M was +1 SMX) and bump up the clocks on core and memory?
At least it beats the [lack of any] improvement that the 880M offered from the 780M.
Sent from my E5823 using Tapatalk -
Yup, those lower-end MX models were leaked already longer ago and were expected for Q1 2016 (920MX, 930MX, 940MX, 965MX / 965M Ti):
http://www.notebookcheck.net/New-Ma...930MX-and-940MX-GPUs-on-the-way.155886.0.html
http://www.notebookcheck.net/Nvidia-Geforce-GTX-965M-Ti-coming-Q1-2016.155408.0.html
GTX 970MX and 980MX leaks are new and kinda more fuzzy. Would be nice to get confirmations from more sources. Though info so far does makes sense and kinda completes the pattern. -
Well, in case we really see those MX brandings will suggest that those won't be just rebrands/replacements, but more expensive alternatives to already costly Ms, like 670MX was more expensive than 670M. As an economist, I see this as a pure robbing or extreme exploiting of the near-monopoly situation.
-
LOL, what the heck? It says DDR3 vRAM. What's the point then? It was supposed to offer GDDR5 which would improve performance a good 20% or more because of the limited 64-bit bus.Mr Najsman, Ionising_Radiation and Talon like this.
-
That is what I had thought as well, hopefully just a misprint?
-
Sure but they weren't actually seen in any laptops yet.
-
980M is weaker than a 970 no matter how you slice it.
980M is not certified for occulus, mainly because overclocking isn't a guarantee, and it cannot match a 970's memory bandwidth (yes, even the gimped 224-bit memory bus crap on the 970 is superior at stock to even a 980M's general max memory OC of 6000MHz, as long as the final memory section of the 970 isn't being used). -
Two Typos?
- Graphics GeForce 940M
- Graphics VRAM 2GB DDR3
http://www.msi.com/product/notebook/CX62-2QD.html#hero-specification -
transphasic Notebook Consultant
_____________________________________________
So many questions, and so few answers to play around with here with what Nvidia has up their sleeves for this summer.
Since the slightly-upgraded 980MX release is not going to be Pascal, then it would make sense for them to release it much sooner than June- more like March, so as to not distract people from their new crowning achievement in 14nm process in Pascal.
However, since they have released some MX versions before just for Apple IMacs, will they do the same once again here with the 980MX?
Since Pascal is going to be a big game changer for gamers, that would seem the most likely scenario, it would seem.
I am going to hold off on my next Laptop until Skylake is paired with Pascal, and see how it measures up benchmark-wise with current systems. -
Might as well wait for Kaby Lake at this point.
-
No no no this is the "old" Broadwell model. The model I was referring to is still unreleased and will be based on Maxwell.hmscott likes this.
-
But being forced to use Windows 10... ugh.
-
Its going to happen eventually, might as well get used to the idea now or move to a Linux box. Sad reality.
With that said, I think Microsoft is trying to dump support on the manufacturers. Microsoft wants to put all their focus on one operating system. Microsoft wants Windows 10 on every device in order to have a unified platform for every device we use. Ambitious, yes. Stupid idea, maybe. But I highly doubt that Intel and AMD will drop support for the older operating systems and will instead support them on their own. Why? Because businesses are not upgrading to 10 any time soon and they won't let their bottom line get hit by it.jaybee83, Ashtrix, TomJGX and 1 other person like this.
Nvidia`s schedule: Geforce 940MX, GTX 970MX and 980MX in Q2 2016
Discussion in 'Gaming (Software and Graphics Cards)' started by Cloudfire, Jan 16, 2016.