Jump to content
3DCoat Forums

The Nvidia 970


Recommended Posts

  • Advanced Member

eh I still bought two of them anyway. Still worth the money, but I do agree with the anger regarding the forced bottlenecking (maiming) of the card in order to keep the premium version more appealing. The strix 780 isnt worth it either, though much of it in part due to the fact that bitcoin miners really messed up the market for older GPUs, which can now sell for a lot more money than newer cards with equal or better specs.

Edited by RabenWulf
Link to comment
Share on other sites

  • Advanced Member

when i was buying the strix was only $50 more expensive and would've had 6gb of ram, which for gpu rendering would've been nice =) 

 

or the 780ti which only has half a gig less of ram. would've been a better choice for me. (disregarding the cost in this case)

Link to comment
Share on other sites

  • Advanced Member

I feel a bit better for forking out the extra $ for a 980.  Although I haven't seen how much this affects rendering, all of the sites that I have read just talk about games.  I did read that even with the slow down for the last half gig of memory it will still be faster than if you are rendering with a CPU.   I don't think it's the end of the world for 970 owners, but it's still shenanigans on Nvidia's part.   :angry:

Link to comment
Share on other sites

  • Reputable Contributor

This is one reason why I've been waiting for the next generation/architecture card. NVidia seems to have a habit of crippling something very important for CG content creators, in order to get power consumption down. They either hamstring the Memory Bus, make the CUDA Cores far less effective than the Fermi cards....even if there are a lot more of them. The GTX 580 3GB has been a nice balance of affordability, good performance handling large meshes, and still a beast in GPU rendering. So, I'm going to keep waiting until NVidia gets it right.

Link to comment
Share on other sites

  • Advanced Member

I feel a bit better for forking out the extra $ for a 980.  Although I haven't seen how much this affects rendering, all of the sites that I have read just talk about games.  I did read that even with the slow down for the last half gig of memory it will still be faster than if you are rendering with a CPU.   I don't think it's the end of the world for 970 owners, but it's still shenanigans on Nvidia's part.   :angry:

 

From what I am hearing, the 980 only gives you a and I quote "10% improvement for a 60% more price increase". Which is probably why nvidia tried to throttle the 970. That said, given that info why not go 2x970 SLI then? Seems like you would get more for less that way.

Link to comment
Share on other sites

  • Reputable Contributor

From what I am hearing, the 980 only gives you a and I quote "10% improvement for a 60% more price increase". Which is probably why nvidia tried to throttle the 970. That said, given that info why not go 2x970 SLI then? Seems like you would get more for less that way.

SLI is only beneficial for games. Does nothing for CG apps. GPU renderers will see a second card and use it accordingly. No need to use an SLI bridge. Nor do 3D apps use SLI to increase performance in the viewport.

Link to comment
Share on other sites

  • Contributor

We don't have flame wars on this forum. Especially if one thinks you are referring to our dear friend AbnRanger. But I see you are referring to the hapless clown that started the thread. So your remarks were merely intemperate (consider his poor parents).

Edited by Tony Nemo
Link to comment
Share on other sites

  • Advanced Member

From what I am hearing, the 980 only gives you a and I quote "10% improvement for a 60% more price increase". Which is probably why nvidia tried to throttle the 970. That said, given that info why not go 2x970 SLI then? Seems like you would get more for less that way.

That would be true, but I had other considerations that made two 270's unworkable.  The biggest issue is that to get GPU rendering working really well you need another GPU to handle your monitors.  My box will only allow me to have two double slot GPU's.  I have a third slot on the mother board but the case doesn't have enough room.  The other consideration was that I was upgrading other parts of my computer and only had so much money to go around, the two 270's would have been more expensive than the one 980.  I'm happy with it, it's so much faster than my old 460 and uses less power and it has enough speed to get me through for a couple of years until the next round of upgrades.  :)

Link to comment
Share on other sites

  • Reputable Contributor

We don't have flame wars on this forum. Especially if one thinks you are referring to our dear friend AbnRanger. But I see you are referring to the hapless clown that started the thread. So your remarks were merely intemperate (consider his poor parents).

Tony...are you OK, man? :) Nobody is flaming another in this thread. Lobito is talking about the guy in video. BTW, it's obvious that someone added the text to some random video on the interwebs. It's just someone's idea of a joke.

  • Like 1
Link to comment
Share on other sites

  • Contributor

Ha ha ha! I wanna know what that dude is really talking about. Must be pretty funny! What language is he speaking?

Edit: I did a little research and found the original video. He is speaking Spanish. I guess I'll find someone to translate it for me.

Link to comment
Share on other sites

  • Contributor

Tony...are you OK, man? :) Nobody is flaming another in this thread. Lobito is talking about the guy in video. BTW, it's obvious that someone added the text to some random video on the interwebs. It's just someone's idea of a joke.

So he isn't an engineer or an idiot but someone relating a funny anecdote to his companion. I hope Timmy shares his discovery.

Link to comment
Share on other sites

  • Reputable Contributor

About the mystery. Someone wrote in one of the comments that this is a Spanish TV programme and the guy is simply telling some jokes.

Timmy, if you discover what he's saying, do drop us a word please. :)

I bet it's pretty R rated. Sort of like Benny Hill, but more colorful. :D

Link to comment
Share on other sites

  • 3 weeks later...
  • Advanced Member

We haven't heard the last of NVIDIA's GeForce GTX 970 memory allocation controversy, not by a long shot. Owners of the card, after having compiled technical information and details over weeks, filed a class-action lawsuit in a US Court (District Court for the Northern District of California). The lawsuit, titled "Andrew Ostrowsky (and others in similar situation) vs. NVIDIA Corporation and GIGABYTE Global Business Corporation," accuses the defendants of unfair, unlawful, and deceptive business practices, in three separate charges, and misleading advertising, demanding for Jury Trial.

The lawsuit goes on to read that the amount in controversy exceeds US $5 million, and encompasses over 100 Class members, meeting the minimal diversity clause, with the plantiff and numerous Class members being citizens of different states than the defendants. The lawsuit accuses the defendants of misleading buyers of the GeForce GTX 970 graphics cards with memory amount (being 3.5 GB with a 0.5 GB "spillover," and not the advertised 4 GB), ROP count being 56 and not 64 (as communicated to the media at launch, and to buyers through them); and L2 cache amount being 1.75 MB and not 2 MB. If you are eligible to be a Class member, find details of the law firms involved in the lawsuit document.

 

http://www.techpowerup.com/210100/class-action-lawsuit-filed-against-nvidia-over-gtx-970-memory-issue.html

Link to comment
Share on other sites

  • Advanced Member

http://www.techpowerup.com/210149/it-wont-happen-again-nvidia-ceo-breaks-silence-on-gtx-970-controversy.html

 

 

Hey everyone,

Some of you are disappointed that we didn't clearly describe the segmented memory of GeForce GTX 970 when we launched it. I can see why, so let me address it.

We invented a new memory architecture in Maxwell. This new capability was created so that reduced-configurations of Maxwell can have a larger framebuffer - i.e., so that GTX 970 is not limited to 3GB, and can have an additional 1GB.

GTX 970 is a 4GB card. However, the upper 512MB of the additional 1GB is segmented and has reduced bandwidth. This is a good design because we were able to add an additional 1GB for GTX 970 and our software engineers can keep less frequently used data in the 512MB segment.

Unfortunately, we failed to communicate this internally to our marketing team, and externally to reviewers at launch.

Since then, Jonah Alben, our senior vice president of hardware engineering, provided a technical description of the design, which was captured well by several editors. Here's one example from The Tech Report.

Instead of being excited that we invented a way to increase memory of the GTX 970 from 3GB to 4GB, some were disappointed that we didn't better describe the segmented nature of the architecture for that last 1GB of memory.

This is understandable. But, let me be clear: Our only intention was to create the best GPU for you. We wanted GTX 970 to have 4GB of memory, as games are using more memory than ever.

The 4GB of memory on GTX 970 is used and useful to achieve the performance you are enjoying. And as ever, our engineers will continue to enhance game performance that you can regularly download using GeForce Experience.

This new feature of Maxwell should have been clearly detailed from the beginning.

We won't let this happen again. We'll do a better job next time.

Jen-Hsun

Edited by L'Ancien Regime
Link to comment
Share on other sites

  • Advanced Member

im thinking of buying a laptop with a 970 in it. on the websites its shown that it has 6gb of memory. Is that real memory or do i get this crippled stuff? Where could i check?

Link to comment
Share on other sites

  • Reputable Contributor

Nvidia has been purposefully deceptive before, with regard to the Kepler cards (GTX 6xx-770), and CUDA performance. They touted the near tripled number of CUDA Cores as a major step forward...when in fact it was a major step backwards in raw performance. I got burned by that and it's why I'm super cautious about making another purchase ever since. There were more alleged deceptive or fraudulent claims by NVidia, for the same card, in that class action suit. So, it's funny the CEO would only address the memory and not the other charges.

 

I think he should be forced to step down over all of this, because it's become a common practice. The crippled memory bus is another example of them touting all the benefits of a new card, while failing to mention they actually took a major step back in other key areas of the card. They know the average person is not going to know about the importance of the size/throughput of the memory bus. I'm not touching another 256bit card...period, regardless of what excuses they make for it. Power consumption takes a back seat on the priority list for most consumers.

Link to comment
Share on other sites

  • Reputable Contributor

The 780ti has a 384bit bus so why would they want to put a 256bit bus on the 980s gets me.

I remember the cuda core disaster... I was glad I did not buy a new card at that time but stayed with my older 550Ti with just 200 cuda cores but they are running full tilt and not crippled.

 

I am going to upgrade but that 256bit bus for the 980s does stick in your throat...

Edited by digman
Link to comment
Share on other sites

  • Reputable Contributor

Also as Nvidia makes stronger gaming cards they continue to cripple the cuda cores... 980s are  FP 64 ----1/32 FP32, down from the 780s  which is 1/24 FP32

My 550Ti is 1/8 FP32

 

Yes according to Nivida the 980 being a faster card the cuda cores stay on par... I do not mine some crippling as I am paying for a gaming card but 1/32 is just not a fair deal... 1/16 I could live with if they stayed at that level and no more further crippling... I lose my desire to buy a quadro card from Nvidia because of these practices even though a quadro card is not crippled and I pay 2,500 bucks for the 5200 8 gig card to get the non-crippling of the cuda cores...

 

http://www.anandtech.com/show/8526/nvidia-geforce-gtx-980-review/7

Edited by digman
Link to comment
Share on other sites

  • Advanced Member

http://www.pcpro.co.uk/other/1000216/nvidia-launches-drive-the-computer-self-driving-cars-have-been-crying-out-for

 

When you look at where Nvidia is going with its company I sort of question even  how important GPUs are to their future. I suspect that  their real corporate focus is on NVidia's role in the swiftly approaching revolution of the self driving car.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...