[Kitguru]Nvidia`s big Pascal GP100 have taped out - Q1 2016 release

[tential]

Who says Nvidia is losing performance?

AMD cards are gaining performance from having poor performance at launch. Nvidia performance isn't decreasing.

And are we seriously debating the value at the high end? If you want a value card to last you years, then you aren't spending $650+. You're getting a 970/290, then upgrading when that card isn't good enough. But hey, lets make all the weird arguments we want about card longevity in the high end when we all recommend people AGAINST buying and holding high end cards for years vs midrange cards.

 

[tg2708]

Who says Nvidia is losing performance?

AMD cards are gaining performance from having poor performance at launch. Nvidia performance isn't decreasing.

And are we seriously debating the value at the high end? If you want a value card to last you years, then you aren't spending $650+. You're getting a 970/290, then upgrading when that card isn't good enough. But hey, lets make all the weird arguments we want about card longevity in the high end when we all recommend people AGAINST buying and holding high end cards for years vs midrange cards.

Is this not a good thing in certain cases? Spend the 600 across 2 years, the card of the second year will most likely have newer features and better driver support even if it loses a bit of performance compared to the former.

Also in newer games AMD is no longer playing catch up to the card it was meant to compete with in its time while nvidia performance is at a stand still with kelper. So it seems nvidia cards are an "of now card" rather than "of now and later card." This is how I see it feel free to take a stance.

 

[Silverforce11]

Who says Nvidia is losing performance?

AMD cards are gaining performance from having poor performance at launch. Nvidia performance isn't decreasing.

And are we seriously debating the value at the high end? If you want a value card to last you years, then you aren't spending $650+. You're getting a 970/290, then upgrading when that card isn't good enough. But hey, lets make all the weird arguments we want about card longevity in the high end when we all recommend people AGAINST buying and holding high end cards for years vs midrange cards.

Agreed, high end has never been about value or bang for buck.

For that, mid-range is where it's at. R290/X & 970 was the value king and still are these days.

 

[TaintedSquirrel]

AMD gaining performance is the functional equivalent to Nvidia losing performance.

Glass half full vs Glass half empty. The glass is still at 50% capacity.

 

[AtenRa]

What is with these double standards. There are people as you say and people as I say, yet I'm not allowed to represent myself?

These blanket statements are so full of feces it's not even funny anymore.

Pro AMD Guy:
ALL gtx 980 ti owners will regret their purchase.

980 Ti Owners:
Some of us won't.

Pro AMD Guy:
You think everyone is like you! LOLZ!!!!!11!!

It's like some of us with money can't exist or express our opinion because it hurts those without money. Woof. I went from being a sexual abuse survivor to someone that spreads STDs. You guys are too much, honestly.

Read again what you said,

I've noticed most people dropping $500+ on a GPU aren't the kind that care when something is faster. Since they'll most likely be upgrading to it.

It's the penny pinchers that seem to care most what those $500+ people are doing.

You have ZERO data to claim "most people" and "its only the penny pinchers". I told you that what you think yourself and a few friends is not what "most people" are doing.

Yes there are people like you who buy the fastest GPU every year, but that is not what most people who buy high-end GPUs are doing.

Come on, serious?

My upgrade pattern for the last few years:
HD 4870 > 4870X2 > 5870 > 5870 2GB > 7970 > 7970 CFX > EDIT Oops forgot- GTX 660 Ti SLI > 780 > 980 Ti

Why is it so hard for some to accept - WE HAVE MONEY!!! We really don't care about the performance in X-years, when something better* comes out, WE UPGRADE!!!

*Subjective: "better" is a term deemed by individuals.

You could only talked about your self and nobody would argue.

But again you are talking about WE, WE is not most people.

 

[littleg]

Read again what you said

I think you need to read again.

I've noticed most people dropping $500+ on a GPU aren't the kind that care when something is faster. Since they'll most likely be upgrading to it.

He doesn't say most people, he says most people dropping $500 on a GPU

 

[AtenRa]

I think you need to read again.



He doesn't say most people, he says most people dropping $500 on a GPU

And you should read again what i said,

Yes there are people like you who buy the fastest GPU every year, but that is not what most people who buy high-end GPUs are doing.

 

[railven]

Read again what you said,


You have ZERO data to claim "most people" and "its only the penny pinchers". I told you that what you think yourself and a few friends is not what "most people" are doing.

Yes there are people like you who buy the fastest GPU every year, but that is not what most people who buy high-end GPUs are doing.

If I could back up my claim, I would. It's my gut feeling that most people who buy in that bracket of GPU upgrade frequently, possibly every numerical generation.

If we took a poll here, I'd still argue most people buying in that bracket/tier would agree with me.

You could only talked about your self and nobody would argue.

But again you are talking about WE, WE is not most people.

It's just ironic that you have an issue with me trying to argue most people in my bracket are similar to me, but had no issue when ALL 980 Ti owners were balled into one.

 

[stahlhart]

AtenRa, railven:

SHUT UP, and take it to PM, NOW.

-- stahlhart

 

[Erenhardt]

Are there any ETA's on the new GPUs on the new manufacturing process?
Is it worth it to upgrade again to 28nm GPU?
Both, Samsung (GloFo aswell?) and TSMC have their 14/16nm process sorted out and are fabbing for Apple. I guess it shouldn't take too long to ramp up the production beyond what apple can swallow, so we can have our GPUs on the production lines.

Is HBM2 a viable option for 2016 for a mainstream products? Will we see a sudden spike in VRAM capacity?

I have an itch... but I can wait... a bit, I hope...

 

[Azix]

Pro AMD Guy:
ALL gtx 980 ti owners will regret their purchase.

.

what i said was "Within the next 6 months to a year 980 ti owners will regret buying that card."

can be few, most or all. All you can definitely get from that is some will

 

[ShintaiDK]

what i said was "Within the next 6 months to a year 980 ti owners will regret buying that card."

can be few, most or all. All you can definitely get from that is some will

So you pretty much made sure your "prediction" would succeed. Even if you can find 1 980TI owner only. Same nonsense as always isn't it.

Replace 980TI with Fiji and get the same result in terms of regret.

 

[Azix]

So you pretty much made sure your "prediction" would succeed. Even if you can find 1 980TI owner only. Same nonsense as always isn't it.

Replace 980TI with Fiji and get the same result in terms of regret.

I also said he would pay me if I won and if I lost.

you guys make too much of this

 

[ShintaiDK]

Are there any ETA's on the new GPUs on the new manufacturing process?
Is it worth it to upgrade again to 28nm GPU?
Both, Samsung (GloFo aswell?) and TSMC have their 14/16nm process sorted out and are fabbing for Apple. I guess it shouldn't take too long to ramp up the production beyond what apple can swallow, so we can have our GPUs on the production lines.

Is HBM2 a viable option for 2016 for a mainstream products? Will we see a sudden spike in VRAM capacity?

I have an itch... but I can wait... a bit, I hope...

Samsung from the looks of it only got a low power version. And if we are to believe the A9 SoCs. Then its stinks. With TSMC being something like 50% better in performance/watt. TSMC 16FF seems the way to go.

While they do make A9s. The question is cost. Apple seems to be willing to pay 100 or even 200$ for every ~100mm2 A9. AMD and NVidia wouldn't be able to pay that. Sometime in 2016 tho the price should be low enough for GPUs.

We have to see if the HBM2 delay rumour is true or not. One thing is sure tho, we will see massive VRAM inflation. It seems like 4GB for lowend, 8GB mainstream, 16GB highend and 32GB professional.

But 16FF+HBM2 cards still look to be Q3/Q4 2016 even if on track.

 

[tential]

So you pretty much made sure your "prediction" would succeed. Even if you can find 1 980TI owner only. Same nonsense as always isn't it.

Replace 980TI with Fiji and get the same result in terms of regret.

Id be disappointed with fiji if I owned it right now... So I don't get why it's the long run that matters when a gpu is fastest in games when it comes out. My high end gpu 2-3 years from now won't handle 4k like when it did when it first came out.

 

[tential]

If I could back up my claim, I would. It's my gut feeling that most people who buy in that bracket of GPU upgrade frequently, possibly every numerical generation.

If we took a poll here, I'd still argue most people buying in that bracket/tier would agree with me.




It's just ironic that you have an issue with me trying to argue most people in my bracket are similar to me, but had no issue when ALL 980 Ti owners were balled into one.

It's odd because those complaining loudest are those living in economies in which gpus are even more expensive relative to income. My gpu purchase is a drop in the water compared to other purchases..
My gpu is how I access the games. Playing with a slow gpu or a gpu not up to par isn't worth it anymore. I did a small upgrade to the 290 from my 7950 and I will probably upgrade yearly or every other year. Playing with compromises vs enjoying my experience. If I sold my gpu at a 200 loss every hear, that's 200 a year to game in hq.
That's nothing in the world of pc gaming.
I cant even play dmc4 right now because playing without 4k freesync breaks my heart.

 

[tential]

Agreed, high end has never been about value or bang for buck.

For that, mid-range is where it's at. R290/X & 970 was the value king and still are these days.

I mean I was at 1440p vsr with the 7950. Upgraded to the 290 for 1800p vsr. Will upgrade to arctic Islands for hopefully higher than 4k vsr.
It's all miniscule upgrades. If I wanted price/perf I'd stick with 1440p. I just want to game at high res downsample.
Especially with vsr, I've had to ignore value, and just get the job done. Otherwise, I just don't game because I'm not happy with the image and lose my immersion completely.
I mean, after my 4k monitor I may get sapphire fury trix. I mean, might as well have all 3 of the nicest sapphire cards released these last gens lol.

I just don't think a 980ti owner is unhappy anytime soon. You spend that kind of money knowing what could happen but you're happy now. My issue with amd is I'm not happy with the card for awhile. I'm happy with nvidia performance at launch.

 

[Headfoot]

Are there any ETA's on the new GPUs on the new manufacturing process?
Is it worth it to upgrade again to 28nm GPU?
Both, Samsung (GloFo aswell?) and TSMC have their 14/16nm process sorted out and are fabbing for Apple. I guess it shouldn't take too long to ramp up the production beyond what apple can swallow, so we can have our GPUs on the production lines.

Is HBM2 a viable option for 2016 for a mainstream products? Will we see a sudden spike in VRAM capacity?

I have an itch... but I can wait... a bit, I hope...

Nobody seems to know... My guess is that the previous guesses of Spring 2016 are too optimistic and we won't see them until late 2016. I hope I'm wrong though

 

[tential]

Nobody seems to know... My guess is that the previous guesses of Spring 2016 are too optimistic and we won't see them until late 2016. I hope I'm wrong though

I agree I think we won't see anything til holiday season. A little late, but that's been what I've planned for with my money.

 

[Teizo]

AMD gaining performance is the functional equivalent to Nvidia losing performance.

Glass half full vs Glass half empty. The glass is still at 50% capacity.

No. That would be Nvidia losing their 'lead' so to speak. It is not as if Nvidia would be under performing. It would just mean Radeon's are starting to get better in relation.

I understand the argument you are making, but as stated it sounds like Nvidia would be taking backward steps instead of Radeon making large forward steps.

 

[xpea]

If anything I am thinking the main performance difference between Titan X and GP100 Titan would be via clock speed increases. 16FF should provide a real nice power consumption improvement and there will be room to increase clocks. How nVidia is going to manage going from a DP-light design to a heavy one is something to watch... especially with no transistor cost gain despite the smaller node.

if rumors are true (GP100 is reported to have 17 billions transistors), we will see performance improvement by huge boost of transistor count + HBM2 bandwidth increase instead of clock speed

 

[jpiniero]

if rumors are true (GP100 is reported to have 17 billions transistors), we will see performance improvement by huge boost of transistor count + HBM2 bandwidth increase instead of clock speed

Likely fake. I don't think the market's ready for something like a $1500 Titan. I would think they would be more gentile with the price hikes.

 

[tential]

AMD gaining performance is the functional equivalent to Nvidia losing performance.

Glass half full vs Glass half empty. The glass is still at 50% capacity.

AMD is going 60 miles an hour
Nvidia is going 70 miles an hour

AMD now is going 70 miles an hour.

Nvidia didn't go slower... AMD just went faster.

It's not a glass half full half empty situation. It's not even remotely close.

 

[tential]

Likely fake. I don't think the market's ready for something like a $1500 Titan. I would think they would be more gentile with the price hikes.

Like the original Titan release right? That was a gentle price hike?
Or Titan Z?

 

[MrTeal]

AMD is going 60 miles an hour
Nvidia is going 70 miles an hour

AMD now is going 70 miles an hour.

Nvidia didn't go slower... AMD just went faster.

It's not a glass half full half empty situation. It's not even remotely close.

I suppose it depends whether you like using your card to play games, or to argue Green vs Red.