Soldato
- Joined
- 2 Jan 2012
- Posts
- 12,407
- Location
- UK.
[Digitaltrends.com]
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
The end of the battle between AMD and Nvidia won’t be great for PC gamers
- Thread starter Boomstick777
- Start date
+1
As nice as the Titan X is, it is also a dead end and obsolete.
When the 390X arrives even if it does not beat a Titan X, it is still the first of a new wave of cards using HBM.
It will probably take NVidia until next year to catch up and start releasing cards with HBM.
Permabanned
- Joined
- 12 Sep 2013
- Posts
- 9,221
- Location
- Knowhere
I don't totally agree with the article but AMD is in a precarious position at the moment. They need to get a gpu range released so they don't lose more of there market share and they need to hold out until there Zen cpu range appears and then if that is a good product hopefully things will become easier and they can restart the pushback on the gpu side of things.
Maybe it's got like it has because they've become too unfocused and have spread themselves too thin, Let's hope they hold out and get in the game again or else as the title say's it'll be a bad day for PC gamers.
Maybe it's got like it has because they've become too unfocused and have spread themselves too thin, Let's hope they hold out and get in the game again or else as the title say's it'll be a bad day for PC gamers.
Useless article, clickbait frankly.
The only issue is the likely lack of anything beyond 16nm, it becomes profitless unless somehow someone can develop a cheap method (which i doubt).
Both companies have issues, but i think Nvidia has the most to lose.
The only issue is the likely lack of anything beyond 16nm, it becomes profitless unless somehow someone can develop a cheap method (which i doubt).
Both companies have issues, but i think Nvidia has the most to lose.
If NV were more on the ball they would've used the 28nm extension to put out a HBM card, now they'll have to risk it on a brand new finfet process. I hope it doesn't go wrong like the 480.
http://www.kitguru.net/components/a...ners-in-2015-to-use-euv-tools-for-10nm-chips/
You never know, umpteenth time might be the charm.
Last edited:
There will be more differentiation soon than there has been for years, the article is ********.
HBM (biggest change in an eternity) will see AMD with this advantage for an entire year, and when NVIDIA do release it, it's likely to be one and done (with them moving to the competing HMC which they'd originally planned to use).
NVIDIA will be moving to TSMC's 16nm FINFET at some point next year, with AMD moving to GF's 14nm FINFET node process (effectively Samsung's process). Again, no more same foundry, same node for both.
AMD have bet heavily on Mantle, and will likely reap the benefits in DX12 (heavily Mantle influenced) until NVIDIA can release Pascal / collude with MS to cripple AMD. Vulkan should massively favour AMD, both on Windows (where it's more likely to be an option than the norm initially) and particularly on OSX / Linux. NVIDIA are betting on dodgy closed libraries which seem to result in awful performance for both parties. Choice here.
G-Synch vs a free VESA standard which is cheaper, universal and arguably superior.
HSA vs ....... effectively nothing in the consumer space.
HBM (biggest change in an eternity) will see AMD with this advantage for an entire year, and when NVIDIA do release it, it's likely to be one and done (with them moving to the competing HMC which they'd originally planned to use).
NVIDIA will be moving to TSMC's 16nm FINFET at some point next year, with AMD moving to GF's 14nm FINFET node process (effectively Samsung's process). Again, no more same foundry, same node for both.
AMD have bet heavily on Mantle, and will likely reap the benefits in DX12 (heavily Mantle influenced) until NVIDIA can release Pascal / collude with MS to cripple AMD. Vulkan should massively favour AMD, both on Windows (where it's more likely to be an option than the norm initially) and particularly on OSX / Linux. NVIDIA are betting on dodgy closed libraries which seem to result in awful performance for both parties. Choice here.
G-Synch vs a free VESA standard which is cheaper, universal and arguably superior.
HSA vs ....... effectively nothing in the consumer space.
Associate
- Joined
- 1 Mar 2014
- Posts
- 2,419
Next time please add [Opinion Piece] thanks 
Soldato
- Joined
- 27 Dec 2009
- Posts
- 2,727
- Location
- Gillingham, Kent
I get really frustrated where every time there's a gap between one GPU maker introducing a new series and the competitor introducing theirs, people always go on about how doomed the later company is. I remember when the DX11 AMD 5xxx series launched and Nvidia were a couple of months later with their DX11 offering - the 4xx series. During this time I remember several people and articles proclaiming AMD's advantage and how on the ropes Nvidia were - total rubbish, as we all know with hindsight. Just one company moving a little quicker than the other.
I am also still unsure whether we should consider the Nvidia 9xx series as a different generation to the AMD R9 2xx series or not. Both are on the same process and have the same very basic setup (28nm GPU with GDDR5 RAM), so are Nvidia really a generation ahead with AMD "needing" to come out with a new GPU to compete? Don't get me wrong, I'm not knocking Maxwell and it's power saving is very impressive, but in terms of grunt it's only on a par with the Nvidia 7xx/AMD R9 2xx series, not massively ahead of them. The only card that I do consider a decent improvement over the 7xx/2xx series is the Titan X, which is priced too highly for me to consider it a mainstream gaming card. I am aware others will view this very differently to me, but everyone's entitled to their opinions!
I am hoping the R9 3xx series is a proper step forward on performance, and then Nvidia respond with a 980Ti to get things back on an even keel and stop the doomsayers on both sides for a while.
I am also still unsure whether we should consider the Nvidia 9xx series as a different generation to the AMD R9 2xx series or not. Both are on the same process and have the same very basic setup (28nm GPU with GDDR5 RAM), so are Nvidia really a generation ahead with AMD "needing" to come out with a new GPU to compete? Don't get me wrong, I'm not knocking Maxwell and it's power saving is very impressive, but in terms of grunt it's only on a par with the Nvidia 7xx/AMD R9 2xx series, not massively ahead of them. The only card that I do consider a decent improvement over the 7xx/2xx series is the Titan X, which is priced too highly for me to consider it a mainstream gaming card. I am aware others will view this very differently to me, but everyone's entitled to their opinions!
I am hoping the R9 3xx series is a proper step forward on performance, and then Nvidia respond with a 980Ti to get things back on an even keel and stop the doomsayers on both sides for a while.
+ 1I get really frustrated where every time there's a gap between one GPU maker introducing a new series and the competitor introducing theirs, people always go on about how doomed the later company is. I remember when the DX11 AMD 5xxx series launched and Nvidia were a couple of months later with their DX11 offering - the 4xx series. During this time I remember several people and articles proclaiming AMD's advantage and how on the ropes Nvidia were - total rubbish, as we all know with hindsight. Just one company moving a little quicker than the other.
I am also still unsure whether we should consider the Nvidia 9xx series as a different generation to the AMD R9 2xx series or not. Both are on the same process and have the same very basic setup (28nm GPU with GDDR5 RAM), so are Nvidia really a generation ahead with AMD "needing" to come out with a new GPU to compete? Don't get me wrong, I'm not knocking Maxwell and it's power saving is very impressive, but in terms of grunt it's only on a par with the Nvidia 7xx/AMD R9 2xx series, not massively ahead of them. The only card that I do consider a decent improvement over the 7xx/2xx series is the Titan X, which is priced too highly for me to consider it a mainstream gaming card. I am aware others will view this very differently to me, but everyone's entitled to their opinions!
I am hoping the R9 3xx series is a proper step forward on performance, and then Nvidia respond with a 980Ti to get things back on an even keel and stop the doomsayers on both sides for a while.
Weather nvidia or amd make the best most powerful card in existence isn't really the point anymore. It comes down to cash flow, marketing and direction. If amd become financially unviable it doesn't matter if a great card is just around the concern. Concerning the rumoured buyouts of amd, it could be good or bad comes down to what direction the New owners decide to take things, or if they just want to asset strip intellectual properties.
The important point has been raised before; AMD are loosing money and cutting R&D budget but they need to spread their diminishing budget over many more areas than Nvidia. Nvidia is getting more and more profitable and is increasing their R&D budget.
It is much easier for Nvidia to stay ahead in this situation. Doesn't mean AMD can't keep up but it starts to depend more on less tangible things like AMD has some unique technology, or better engineers, management make better decisions or plain luck.
I think some people are glossing over the fact that this is pretty much a historic low point in AMD's market share, and market share is hard to win back. Even a very good 390X launch wont see AMD's share return to where it has been historically. There are olenty of happy 780/970/980/TX owners that wont feel the need to upgrade. that is the issue if you don't release a product on the same timeline as your competitors, it is no longer a comparison of is product A better than B, it is a question of whether one o should sell their product A, making a loss, and then buy another a new product B to get that small bump. If people loose 100 quid selling on their 980 then they have to feel that they will make more than that value up buying a 390x which is a very different proposition to a straight comparison of products.
It is much easier for Nvidia to stay ahead in this situation. Doesn't mean AMD can't keep up but it starts to depend more on less tangible things like AMD has some unique technology, or better engineers, management make better decisions or plain luck.
I think some people are glossing over the fact that this is pretty much a historic low point in AMD's market share, and market share is hard to win back. Even a very good 390X launch wont see AMD's share return to where it has been historically. There are olenty of happy 780/970/980/TX owners that wont feel the need to upgrade. that is the issue if you don't release a product on the same timeline as your competitors, it is no longer a comparison of is product A better than B, it is a question of whether one o should sell their product A, making a loss, and then buy another a new product B to get that small bump. If people loose 100 quid selling on their 980 then they have to feel that they will make more than that value up buying a 390x which is a very different proposition to a straight comparison of products.
- Joined
- 2 Jan 2012
- Posts
- 12,407
- Location
- UK.
Not sure how you read that and came away with that conclusion.
I read it as despite AMD having some decent hardware and reducing prices on their current cards to make them competitive against Nvidia's latest, Nvidia still continues to sell more. It was more about the shift in market then specific tech.
AMD can sell cards in the lower mid tier but Nvidia are shifting lot's of cards and even in real expensive tier. That's what AMD somehow needs to get back.
Article makes some good points, but then I'm not biased either way (Neither company is great) so I read with open mind lol. Even if 390X is great will it sell as much as Nvidia's cards.. This is the difficult situation AMD are in atm.
+1
Yeah this guy get it. This is about market shift in Nvidia's favour, despite AMD having competitive options. Nvidia continue to make record profit while AMD struggle. This is what the article is talking about.
Last edited:
I hope you are right but I think you are looking at things with red-tinted glasses.
HBM only helps in situations that are bandwidth limited, which is not the case at all times. Also I don't think AMD will necessarily get a whole year use. It all depends when Nvidia bring Pascal to market and when AMD finally get the 390x out the door - there might not be more than 6-8 months separation.
The DX12 stuff is mostly nonsense- Nvidia is part of the DX consortium and has known specs for years. They have had as long as AMD to prepare architectures and drivers. You paint mantle as an advantage for AMD, Mantle has been dropped like a sack of potatoes, I would consider that a big waste of effort. While AMD have been putting resources into Mantle Nvidia have been putting resources in to DX12.
While AMD have been demoing mantle games Nvidia have been demoing DX12 games.
http://wccftech.com/forza-5-demo-directx-12-titan-black-pc/
I hope you are right but I think you are looking at things with red-tinted glasses.
HBM only helps in situations that are bandwidth limited, which is not the case at all times. Also I don't think AMD will necessarily get a whole year use. It all depends when Nvidia bring Pascal to market and when AMD finally get the 390x out the door - there might not be more than 6-8 months separation. Nvidia have been working on stacked DRAM as long as AMD have, being first to market doesn't necessarily give you an advantage. AMD is late, no one know why, but integrating HBM might be a possible explanation *IMO*. Nvidia have been bitten before by trying to bring technologies that weren't ready yet to market, they are much more cautious now because they can afford to be.
I really hope HBM on the 390x makes a dramatic difference though.
The DX12 stuff is mostly nonsense- Nvidia is part of the DX consortium and has known specs for years. They have had as long as AMD to prepare architectures and drivers. You paint mantle as an advantage for AMD, Mantle has been dropped like a sack of potatoes, I would consider that a big waste of effort. While AMD have been putting resources into Mantle Nvidia have been putting resources in to DX12.
While AMD have been demoing mantle games Nvidia have been demoing DX12 games.
http://wccftech.com/forza-5-demo-directx-12-titan-black-pc/
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 31,182
You know that AMD themselves pretty much dropped mantle a long time ago right?
The very lacking amount of games it is included in, no public SDK, they even released a programming paper on how to program for mantle (although its almost entirely just about good low level GPU programming, meaning it applies to DX12 and vulkan) and then stated they wouldn't release an SDK for it.
It was pretty obvious from a development standpoint that mantle only came around to kick the ass of MS into releasing DX12, AMD themselves state to ignore mantle and use DX12 instead.