AMD sees huge improvements in their Q2 2016 financial results

Good for them, now if they can keep this ball rolling, they can start to pour more money into their R&D.

MAx respect for them for keeping their products at sucha good price/performance level, even with their financial dire straits up to now.
 
I just saw that the RX 480 Nitro 4gb is on pre order for £199. I totally and utterly don't need it at all but I want it ! The prices on their 480 have been good so far and the AM cards are not pee taking either.

However, there's not really much point in buying a 480 to go with a stock P2 1055T just for Linux so I think I will opt for the 460 @ around £80. It should be more than enough to play most of my Steam games under Linux (the ones that are on there, L4D for example) so yeah, no need for a 480.

However if Zen is as good as I hope then I may build an APU TV center next year :)
 
Yes it's good that AMD has make some money what is stopping them from putting the price up cause we all know what companies are like once they start making money they want even more.

If this happens people are going to moan about that too.
 
Yes it's good that AMD has make some money what is stopping them from putting the price up cause we all know what companies are like once they start making money they want even more.

If this happens people are going to moan about that too.

Vega will not be cheap. Any one thinking it will is deluded. HBM2 costs a fortune. If you want to know how much? consider why Nvidia have just released another Titan X with no HBM. I would say Vega will cost at least £600 possibly more simply down to the materials AMD have chosen. Even after it got pounded by the Titan X AMD still did not lower the price of the Fury X. This speaks volumes IMO. It was probably because they simply couldn't.

You can also expect Zen to be really expensive too. AMD themselves have already politely pointed out that this sort of tech and investment (in designers and so on) does not come cheap.

I think it's a double edged sword really. Every one expects AMD products to be as good as the competition or better but that they should also be really cheap.
 
Vega will not be cheap. Any one thinking it will is deluded. HBM2 costs a fortune. If you want to know how much? consider why Nvidia have just released another Titan X with no HBM. I would say Vega will cost at least £600 possibly more simply down to the materials AMD have chosen. Even after it got pounded by the Titan X AMD still did not lower the price of the Fury X. This speaks volumes IMO. It was probably because they simply couldn't.

You can also expect Zen to be really expensive too. AMD themselves have already politely pointed out that this sort of tech and investment (in designers and so on) does not come cheap.

I think it's a double edged sword really. Every one expects AMD products to be as good as the competition or better but that they should also be really cheap.
Then you start to question the reason behind investing so much into an expensive technology that may not even allow AMD to topple the (new) Titan X's crown. It appears, then, to be an evolution for the sake of evolution. Maybe they need HBM2 to bring their power consumption down, but I'm sure there is more to it than that. Of course, Vega could match a Titan X(2), but I highly doubt it.
 
For me, it's going to come to a point where I just stop buying new graphics cards. I recognise that inflation is a thing and that as technology improves there is going to be an increase in price. But the Fury X was overpriced and the 1080 is overpriced. That's not a good sign. When the day comes when a mid-tier GPU is priced in the high-end market, I'll be sad. Gaming with every setting at the max is just not worth £1000 every year. FWP.
 
Then you start to question the reason behind investing so much into an expensive technology that may not even allow AMD to topple the (new) Titan X's crown. It appears, then, to be an evolution for the sake of evolution. Maybe they need HBM2 to bring their power consumption down, but I'm sure there is more to it than that. Of course, Vega could match a Titan X(2), but I highly doubt it.

Let there be absolutely and utterly no doubt, Vega will be all about 4k.

Any one buying it for less will never get the full potential out of it. Once you start getting to a level where the GPU itself is powerful enough to chuck 4k around memory bandwidth starts to become crucial. As Kaap pointed out here or somewhere else, with the lame clock speed, high chance of throttling (as even the 1080 does on the reference cooler) and GDDR5X it's quite clear that Nvidia are not being too serious with this "new" Titan X.

In actual specs IE - what it can do I have little doubt that Vega with HBM2 will absolutely smash the Titan X in Vulkan and DX12. The problem is it's still a little too early for people to actually care about that.

Whilst Nvidia have been living in the here and now since Fermi (IE making smaller dies with hardly anything going on and massive clock speeds to throw force at DX11) AMD have been sitting in their corner playing with GCN and tank cores with far more going on and because of that they have lower clock speeds and perform like poo in DX11.

I can't see AMD getting back into the game until DX11 is not used any more. Once that happens the RX 480 will take a large step out in front of the 1060 and the 1060 will simply be left behind. However, by then Nvidia will probably have *another* card that wins any way.

AMD are playing the long game. Whether or not it will work out for them? time is going to tell. I would imagine having all of the consoles under their belt things may well end up suiting them and pushing out Nvidia.

Also you need to consider Moore's Law. Eventually Nvidia are going to end up where Intel are now, closing down factories and laying off staff because people just won't keep buying 5% incremental upgrades.

So they're absolutely milking it while then can, I would imagine so that they have the funds to actually invest into something new other than GPUs that will keep their company alive in the long term.
 
Let there be absolutely and utterly no doubt, Vega will be all about 4k.

Any one buying it for less will never get the full potential out of it. Once you start getting to a level where the GPU itself is powerful enough to chuck 4k around memory bandwidth starts to become crucial. As Kaap pointed out here or somewhere else, with the lame clock speed, high chance of throttling (as even the 1080 does on the reference cooler) and GDDR5X it's quite clear that Nvidia are not being too serious with this "new" Titan X.

In actual specs IE - what it can do I have little doubt that Vega with HBM2 will absolutely smash the Titan X in Vulkan and DX12. The problem is it's still a little too early for people to actually care about that.

Whilst Nvidia have been living in the here and now since Fermi (IE making smaller dies with hardly anything going on and massive clock speeds to throw force at DX11) AMD have been sitting in their corner playing with GCN and tank cores with far more going on and because of that they have lower clock speeds and perform like poo in DX11.

I can't see AMD getting back into the game until DX11 is not used any more. Once that happens the RX 480 will take a large step out in front of the 1060 and the 1060 will simply be left behind. However, by then Nvidia will probably have *another* card that wins any way.

AMD are playing the long game. Whether or not it will work out for them? time is going to tell. I would imagine having all of the consoles under their belt things may well end up suiting them and pushing out Nvidia.

Also you need to consider Moore's Law. Eventually Nvidia are going to end up where Intel are now, closing down factories and laying off staff because people just won't keep buying 5% incremental upgrades.

So they're absolutely milking it while then can, I would imagine so that they have the funds to actually invest into something new other than GPUs that will keep their company alive in the long term.
If big Vega is indeed all about 4K, I won't need big Vega.

For arguments sake, if the 490X is big Vega and offers 120% performance, I'd happy with a 490 at 100-110% performance. But if the 490 is intended to match the 1080 in both price, performance, and features (GDDR5X, for example), that may not be for me. I don't think I want a €750 AMD GPU that isn't even the flagship like nVidia is doing. The reason I migrated to AMD is to be a part of a more affordable system that still innovates and takes risks. Investing in Vulkan, Mantle, DX12, VR, HBM, Freesync, etc. are bold moves. The RX480 is a bolder GPU than the 1060, IMO. The 1060 is a sure thing, absolutely, but it's also a little boring. It's a safe bet that ticks all the boxes, and is certainly more impressive than the 960—which was only as good as a 680/770 until nVidia stopped supporting Kepler—but I still feel the RX480 was a bigger leap forward for AMD than the 1060 was for nVidia. The 1060 will give you higher frame rates in DX11 games (the majority of titles today), lower temperatures, lower power consumption, and more AIB variety, but it doesn't excite me like the RX480 does. I actually want to own an RX480. The 1060 is basically a 980 at a cheaper price. Temperatures or power consumption were never an issue with the 980, or even VRAM. It was just overpriced. That's basically all nVidia are introducing to their lineup with the 1060.

If the 490 is another RX480, where in DX11 tests it falls behind a 1080, but is also cheaper, that's cool with me. I may be reading things wrong, and this is just my opinion that could well be off—I was right about the 1060, 1070 and 1080), but I expected a little better from the RX480 in regards temperatures—but what I personally want is this:

RX490 - €600-650 for AIB like Sapphire Nitro, 200W TDP, 8GB HBM2, DX11 performance above a 1070, and DX12/Vulkan performance around 1080 levels. That's perfect for 1440p/60hz with settings cranked to the max, 1440p/144hz with settings like MSAA slightly lowered, or 4K/60Hz with settings lowered further.

RX490X - €700-750 for AIB like Sapphire Nitro, 250W TDP, 12GB HBM2, DX11 performance at heavily overclocked 1080 levels, and DX12/Vulkan performance at above Fury ProDuo levels (with good scaling factored in). That's 1440p/144Hz with settings cranked or 4K/120hz with settings slightly lowered.
 
Last edited:
€700 is $650, the same price as a Fury X. Gone are the days when I could get a $650 GPU for £500 or €580.

Doesnt work like that anymore sadly. I miss those days too.

Now when we see RRP of $650 they simply just change the currency but the numbers remain €650 or £650.

...Or if in Norway 9001NoK ;)
 
700 for a 490x is a bit much.

Compared to $1200 for a Titan X?

It's a bit ironic isn't it fella? you have three 980tis in your sig rig and those would have cost you about $700 each. Your next logical upgrade is 1080ti at around $900.

It's like I said, people almost expect AMD to offer incredible value whilst Nvidia can seemingly charge whatever they like. If people didn't have that attitude AMD would be far more competitive than they are.
 
Compared to $1200 for a Titan X?

It's a bit ironic isn't it fella? you have three 980tis in your sig rig and those would have cost you about $700 each. Your next logical upgrade is 1080ti at around $900.

It's like I said, people almost expect AMD to offer incredible value whilst Nvidia can seemingly charge whatever they like. If people didn't have that attitude AMD would be far more competitive than they are.
I think AMD should be the more affordable and value orientated option. At launch, the 290 competed against the 780 whilst being $250 cheaper. The prices of the 780 dropped once the 290 was released, but it remained the more affordable option. The same can be said for the 290X which was $150 cheaper than the 780ti. Now the 290X will beat a 780ti in almost any modern test and a 290 will beat a 780.

As to whether people should keep buying nVidia's expensive GPU's whilst criticising AMD for attempting to do the same, I agree, that's not fair. But I don't want AMD to do that. I want to see AMD make another 290X, not another overpriced Fury X. I would like a 290/290X with a more competitive TDP. I'm not interested in paying €900 for a graphics card. I'm not interested in 3-way SLI. I want a 1440p beast for €600. NVdia offer that with the 1070, but it's not really powerful enough to warrant upgrading from my current GPU and swapping to a G-Sync monitor.
 
Compared to $1200 for a Titan X?

It's a bit ironic isn't it fella? you have three 980tis in your sig rig and those would have cost you about $700 each. Your next logical upgrade is 1080ti at around $900.

It's like I said, people almost expect AMD to offer incredible value whilst Nvidia can seemingly charge whatever they like. If people didn't have that attitude AMD would be far more competitive than they are.

If the GTX1080Ti comes out it will be more than $1200 here more like $1500 the GTX1080 here are $1000-1200 dude.
 
I think AMD should be the more affordable and value orientated option. At launch, the 290 competed against the 780 whilst being $250 cheaper. The prices of the 780 dropped once the 290 was released, but it remained the more affordable option. The same can be said for the 290X which was $150 cheaper than the 780ti. Now the 290X will beat a 780ti in almost any modern test and a 290 will beat a 780.

As to whether people should keep buying nVidia's expensive GPU's whilst criticising AMD for attempting to do the same, I agree, that's not fair. But I don't want AMD to do that. I want to see AMD make another 290X, not another overpriced Fury X. I would like a 290/290X with a more competitive TDP. I'm not interested in paying €900 for a graphics card. I'm not interested in 3-way SLI. I want a 1440p beast for €600. NVdia offer that with the 1070, but it's not really powerful enough to warrant upgrading from my current GPU and swapping to a G-Sync monitor.

That is the whole problem right there.

IDK if you remember but back in the Athlon XP days AMD were not beating Intel at the high end. They were spanking them on up to mid range and with what most people would buy but Intel had their Xeons up their sleeve.

So AMD decided they would take on Intel right up to the high end. It nearly sent them out of business. Why? because no matter how good the AMD CPUs were (and they were, spectacular even) people simply would not pay Intel prices for them. Which if you are competing you have costs for and they're the same no matter who you are.

You can't expect AMD to be able to make GPUs that compete with, or beat, Nvidia ones without them coming in at a cost. AMD's last attempt would have cost far more than the GPU it was competing with (the 980TI, I'm talking about Fury X) but they could only charge £530 for it. Which is a lot of money, but they probably had far less profit than Nvidia due to HBM and a liquid cooler..

If you really, truly want them to compete then you need to understand it will cost. Well, unless AMD can break the laws of physics and economy and so on.
 
That is the whole problem right there.

IDK if you remember but back in the Athlon XP days AMD were not beating Intel at the high end. They were spanking them on up to mid range and with what most people would buy but Intel had their Xeons up their sleeve.

So AMD decided they would take on Intel right up to the high end. It nearly sent them out of business. Why? because no matter how good the AMD CPUs were (and they were, spectacular even) people simply would not pay Intel prices for them. Which if you are competing you have costs for and they're the same no matter who you are.

You can't expect AMD to be able to make GPUs that compete with, or beat, Nvidia ones without them coming in at a cost. AMD's last attempt would have cost far more than the GPU it was competing with (the 980TI, I'm talking about Fury X) but they could only charge £530 for it. Which is a lot of money, but they probably had far less profit than Nvidia due to HBM and a liquid cooler..

If you really, truly want them to compete then you need to understand it will cost. Well, unless AMD can break the laws of physics and economy and so on.

It's possible I didn't explain myself properly.

I was disappointed with the Fury X. The price was the absolute max I'd be willing to pay for a GPU, but the performance and design was not what I was looking for. So when I say 'I don't want another Fury X', I mean I don't want to pay €700 for a GPU that does not match a €700 graphics card from a competitor. The 980ti beat the Fury X in almost every respect except for DX12, which was a year away from being introduced. The Fury X closed the gap at higher resolutions, which is why I chose the Fury (non X) over the 980 and I'm glad I did, but it was just not good enough.

I have no problem AMD charging €750+ for a GPU that actually is a beast. I just won't buy it. They released the 295X2 and ProDuo and I had very little interest in them; I just couldn't afford them. I could have bought a 295X2 when they were around £500, but I would have been forced to change power supplies and purchase all new custom cables. I had no preconceived notion that AMD were in some way inferior or did not deserve that much money. I just couldn't afford it.

This is purely speaking from personal preference, I want a €600-650 GPU that handily takes on the 1070 in DX11 and handily takes on the 1080 in DX12, all with a TDP of around 200-225W. That's basically a refresh of the 290 and Fury with a lower TDP. If that's too much to ask for, that's too much to ask for. If AMD want to make a €800 behemoth that puts a huge smile across our faces, wonderful, but I won't buy it. It's not because I won't pay that much for an AMD card; it's because I won't pay that much for a graphics card. I wouldn't even dream of paying what nVidia are asking for the 1080. It's under powered for what they're asking, IMO, especially when I know that a 1080ti could come out that offers a better price to performance ratio.

If AMD are really aiming to make Polaris the more affordable option that maintains healthy competition—and that is exactly what it is doing—then why can't they do that with Vega? Does it have to beat nVidia? Not in my opinion. The RX480 didn't have to beat the 1060. All the card had to do was fulfill a particular set of needs. Getting back to Vega, When I say things like, 'matching 1070 in DX11/matching 1080 in DX12', it is purely as a point of reference. If it's behind, so be it. If it is ahead, so be it. Just make sure the price is reflective of that, as much as possible. Maybe my point of reference as far as prices is out of date and skewed, but it's what I hope for.

The 7970 was $550 and the 680 was $500. The 680 was, at first, the faster card and thus better value. A few months later that changed. That extra $50 on the AMD side bought you something—longevity. A win for AMD.
The 290X was $150 cheaper than the 780ti, but it wasn't quite as powerful. It also was hotter and more power hungry. That saving was worth it, though, due to the longevity and maturity of the card. Another win for AMD.
The Fury X was $650 and did not offer any tangible benefits over the $650 980ti for another year or unless at really higher resolutions, and even then it was held back by VRAM. A win for nVidia.

I must admit that I know very little about CPU's further than 3 years back and things may be very different there. Sorry for the wall of text. I just felt I may not have been understood fully. I want AMD to make an €800 GPU that is absolutely phenomenal. I won't be buying it, though. I'll be buying their lower end GPU. I'm hoping it'll be their 1080 competitor, but it may have to be the 1070 competitor as, like I said, my perspective of prices may be out of date.
 
Back
Top