Nvidia Reveals Pascal: GTX 1080 And 1070 To Beat Titan X, GDDR5X Debuts

[Joseph Jasik]

I am not sure I am keen to buying more horsepower capable cards, only to help devs skirt optimization process because the card will handle the 30-something fps... utter crap!

 

[logainofhades]

There are a few selling for under $300.

http://pcpartpicker.com/parts/compare/asus-video-card-turbogtx970oc4gd5,evga-video-card-04gp42977kr,evga-video-card-04gp43973kr,zotac-video-card-zt9010110p/

 

[WarWolverineWarrior]

GDDR5X is a ripoff. I bet it's nowhere as fast as HBM2.

 

[turkey3_scratch]

GDDR5X is a ripoff. I bet it's nowhere as fast as HBM2.

That's ridiculous to even think that GDDR5X is going to be a burden for the GTX 1080. I think the Nvidia engineers know what they're doing.

 

[littleleo]

I'm really hoping the Zen & the Polaris are good products, they don't have to be great just good will do. They don't have to be home runs just a double or triple will be awesome. There are lots of AMD fans out there that will give them a good look.

Every PC in my house used to have a AMD CPU now they are all Intel. I hope the ZEN CPUs really offer a good options to the Intel CPUs.

I've never been an ATI fan due to all the driver issues I used to have to deal with. Especial before you could get the drivers off the manufacturer's website. You guys are all spoiled we had to call ATI tech support and tell them the VGA card model & the program used and they would send a floppy via USPS with the driver you needed. If you were lucky you could get it to your customer in about 1-2 weeks. I still have nightmares about customer's calling with ATI drivers issues. I never had that issue with any other brand like I had with ATI.

 

[TJ Hooker]

@Mousemonkey what you quoted doesn't actually seem to say that 14 and 16 nm are the same. The way I read it, they're saying that not all dimensions scale equally with process node, i.e. fin and metal pitch are the same between 14 and 16 nm. But I believe there's still another parameter - gate pitch - which is not mentioned in that quote. From what I've read, transistor density is (partially) proportional to gate pitch, so *if* 14 nm has smaller gate pitch than 16 nm, it could still have superior performance.

Now, I have no idea if that's the case, and overall I'm far from an expert on this. But there seems to be a piece of information missing. Also, you made the same comment in another thread, where I asked similar questions, so I'm hoping you can clarify this time around.

 

[dalauder]

2x the performance & 3x the power efficiency: I think they take these statements separately. So you can run the processors at a higher clock with the increased power efficiency enabling 2x the performance.

I'm waiting until AMD drops their next architecture to push those GTX 1070 prices down below $300 on Black Friday sales. Then I might have to bite. My GTX 570 is becoming too anemic to bear.

 

[rhysiam]

Tom's Hardware and @kcarbotte, you really need to correct the misrepresentations in this article. You're spreading misinformation.

The bits about power draw (it's not lower than maxwell, it's higher - just more efficient) and clock speed (you've quoted an "unbelievable 2.1 Ghz", which we know is the result of a significant overclock) aren't helpful. The bigger problem though is the performance claim. The first question everyone has about a new card is, "how fast is it" (closely followed by "how much does it cost"). Yet the only statement about the performance of the 1080 in your entire article is the one claiming "twice the performance of a Titan X" (along with TFlops but nothing to compare them against). You missed the most important part of that sentence, which made clear it was referring to performance of Pascal's "special features".

For comparison, AVX2 in certain situations can approach twice the performance of the initial AVX. But if I start a thread stating that an i5 4670K (with AVX2) is almost twice as fast as an i5 3570K (AVX) without context, I'd rightly get pilloried for spreading misinformation. Yet that's exactly what you've done in this launch article.

Nvidia actually claimed the 1080 was faster than 2x980s in SLI. That's far, far short of double Titan X. They also put a slide up (which you have in your article) which clearly shows they're not expecting anything like double the performance of a Titan X. Here's your own slide: http://media.bestofmicro.com/G/P/579913/original/new-king.jpg
Who knows how Nvidia is calculating "relative performance" in that chart, but it's a safe bet that if they're really expecting double TitanX/980ti performance in games, it would be reflected there.

There's loads of people on the forums going crazy with the double Titan X performance claim, and Toms HW and this article is culpable. You really need to correct or ammend it.

 

[nikoli707]

in real world fps gaming it is highly doubtful that even the 1080ti wont be equal to sli titan x's, and thats still about 10 months away. i dont think there has ever been a generational flagship successor from either amd or nvidia that was anywhere close to twice as fast as the previous generation flagship. usually it falls in the 35-55% area.... think 580 > 680 > 780ti > 980ti... and 20-30% area for its siblings... think 480 > 580 or 680 > 780ti or 980 to 980ti. more recently nvidia has picked up the trend of releasing this mini generational jump which give relatively small performance increases with massive power reduction... 780/780ti > 970/980 or 760 > 960. amd hasn't really done this except with the 7970=280x > 380x and the 7950=280 > 285/380.... though the performance differences are much smaller since their core architectures are nearly identical besides the memory configuration.

rhysiam above hits the nail on the head and makes the obvious comparison... in some random algorithm/code it may indeed be twice as fast... heck even a gtx580 is twice as fast as a titanx is some workstation applications.

 

[Mousemonkey]

@TJ, you could always do some research yourself, Google is free to use after all. 😉

 

[NeutralIsGood]

AMD, make your move, this time I'll wait. Bought the 970 before AMD's 300 series launched and I regret it.

I know that feel. I wanted to say something about AMD and Nvidia but............. Nvidia vs AMD = Flame and Spam Wars.

 

[turkey3_scratch]

@TJ, you could always do some research yourself, Google is free to use after all. 😉

It's sure easy for you to go around telling people to do research rather than answering their highly intellectual question. Unless, of course, you don't have an answer? It's pretty clear he did do the research. We all know Mousemonkey that if it was flipped, Nvidia on 14nm and AMD on 16nm, you'd be on the other side of the bus arguing anyway.

 

[Mousemonkey]

The fact would remain the same and for someone who was moaning about posters who want to be spoon fed, why do you want me to spoon feed you? Are you incapable of doing any research yourself? Is that how you are getting information about PSU's?

 

[turkey3_scratch]

The fact would remain the same and for someone who was moaning about posters who want to be spoon fed, why do you want me to spoon feed you? Are you incapable of doing any research yourself? Is that how you are getting information about PSU's?

Or you could answer TJ Hooker's question. It's between you and him, not me; I'm just trying to see what the outcome of the discussion between you two becomes. I'm only trying to ignite this intellectual discussion between you and Hooker while I am a mere onlooker. You seem as if you want to refrain from answering TJ's question. I never said I researched it; I did not. By TJ Hooker did, and he needs his question answered.

 

[Mousemonkey]

I provided a link which is a starter and if you or anyone else wants more than that then go and look it up. Having a day job means that I don't have the time to do the research for you.

 

[turkey3_scratch]

I provided a link which is a starter and if you or anyone else wants more than that then go and look it up. Having a day job means that I don't have the time to do the research for you.

TJ Hooker did read it, and he asked a question, one which by your persistent avoidance shows you have not done the research to answer him but rather try to turn the tides upon me. It's not about me, it's about him.

 

[king3pj]

I'm pretty tempted to sell each of my 970s for the $225 they are going for on eBay. $150 out of pocket for a 1080 doesn't sound too bad. Hell, I could actually make a small profit if I sold them and bought a 1070. Since SLI 970s are about as fast as a 980 Ti and the 1070 is supposedly faster than a Titan X this might not be a terrible idea.

The only downside is that I would probably be without a GPU for at least a month. I would have to sell before the new cards release to get the best return and stock of the new cards will likely be limited for a little while.

I'm mostly happy with my SLI 970s at 1440p. I have a 60Hz monitor and play most games with VSYNC on. In most new games I play each card is still only at about 50-60% core usage to maintain 60 FPS.

I have had to turn settings down in a few games that were using the full 4GB of VRAM though. Most recently this happened in Forza 6 APEX but it has also happened in The Division and a couple other games. The raw horsepower of SLI 970s still seems to be enough to max out anything I want to play at 1440p but something a little faster with 8GB of VRAM is pretty appealing.

 

[Bartendalot]

I currently have i5-3570K, 16GB of DDR3 RAM and most importantly:G1 Gaming GTX 980 SLI.

So now that hard part, upgrade now to a single 1080p? Or wait?
Also must I full upgrade my rig / build a new system to use this card properly?

I just don't know. For that price range (which is still very expensive but cheaper than what Titan X was) getting x2 performance of Titan X is just tempting.

But maybe I should wait for 1080ti or whatever comes next? Pascal is still in its diapers basically, is skipping it wise?

You should be fine in terms of CPU bottlenecking in most situations if that is what you are asking. You don't need to upgrade anything unless you need to scratch that itch

In terms of waiting, there will most likely be a 1080 ti coming towards the end of the year or early 2017. The 1080 looks like it is going to be a great card but the 1080ti will be that much better.

I currently have an x58 with a 980ti and will wait to upgrade until the 1080ti comes out. It's really up to you. It's not like I am having trouble running anything at 1440 as is. The next upgrade will probably also see me get a new monitor that can take advantage of HDR @1440 with high refresh rate. Or maybe 4k? I'll wait until benchmarks come out...

 

[Mousemonkey]

It would be nice if the graphics market had more then 2 big players,maybe a 3 wheel would promote lower pricing and keep Nvidia/AMD more competitive. who was the last big 3rd graphics company? i forget but had to be long ago.

3dfx, turn of the century.

 

[Nuckles_56]

It does, intel, the chipzilla itself

 

[bit_user]

It has been noted elsewhere that on screen information during some of the show was clocked at 2114MHz. That is a significant jump over the base clock of 1607MHz, or even over the 1733MHz boost clock.

That definitely fooled me. I thought he was running a stock card at stock frequencies.

Overclocking during the launch demo of a new product is pretty sketchy, IMO. Even if they're honest about it.

 

[bit_user]

who was the last big 3rd graphics company? i forget but had to be long ago.

3dfx, turn of the century.

3dfx was bought by NVidia. I heard most of their design talent originally came from SGI.

And I think some other SGI folks went to ARTx, which designed the GameCube's GPU and was bought by ATI.

It does, intel, the chipzilla itself

Yes and no. He's obviously talking about mid-range and high-end. Intel's best Iris Pro parts are only about as good as an entry-level discrete card.

Intel could jump into this market, easily. I think it's just too small for them.

I think it's most likely that one of the mobile players (Imagination, ARM/Mali, Qualcomm, Vivante) decides to scale up, mostly to go after HPC. But I still don't know if they'd bother trying to build and market a graphics card. If Sony or MS used another GPU for their next console, then that could be the entry point. Otherwise, it's a lot of investment for a questionable payoff.

At least Imagination makes PCIe graphics cards, but they're marketed towards content creation (HW accelerated raytracing). Interestingly, they were once in the PC graphics space, with their original PowerVR GPUs.

 

[Clarky-boy]

Not overly impressed..! its still a massive card and ugly IMO..!
Personally I'm going to hold off for the HBM version in due course - should be smaller hopefully.. something like the size of the Nano packing this punch or more would be perfect.

 

[LOLROFL]

Shut up and take my money !

 

[TJ Hooker]

@TJ, you could always do some research yourself, Google is free to use after all. 😉

If you're the one making a claim, especially one that is counter-intuitive (i.e. that 2 different process nodes are actually the same), it's on you to back that claim up with evidence. At least if you want that claim to carry any weight that is.

Edit: To be honest, it's not even that I necessarily think you're wrong, but I like to play devil's advocate sometimes. Not to mention, if you did have some additional info/source handy, I wouldn't mind learning more on the subject just for my own interest's sake. But finding info on the subject isn't trivial.