• 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.

Analysts Start to Fear the Worst for AMD.

So here is my 2 cents.

I agree that AMD are in a very difficult spot and are burning through CASH at an alarming rate. This is driving them to cut costs and investments in R&D which is never a good story in a very competitive industry. They also got hammered by oversupply in the channels on the back of the whole bitcoin mining craze which they made a ton of cash off of in the short term but has impacted demand more than they expected due to used supply.

I suspect they have thrown a ton of cash at things like HBM as well - a typical ATI/AMD move like with tessellation, DX10/10.1, etc. jumping the gun - while GDDR5 is at the end of the road its not yet a significant performance bottleneck and shoving it onto the next round of cards will likely provide little real tangible benefit (that isn't to say there won't be performance increases at all) - I really don't understand them sometimes. We do need to get to it eventually but I'd be surprised if it couldn't have been done far more cost effectively.
 
So you mention the amount of nonsense being posted and then say that they paid too much for ATI to simply forget it. The one thing you failed to think of, is that if someone else buys out AMD, they wont have paid anything for ATI at all.
If the new parent company deems the GPU division isn't worth keeping then they will either sell it off or just close it down.

So you'd buy them based on their CPU sector only?

You're a brave man.
 
I suspect they have thrown a ton of cash at things like HBM as well - a typical ATI/AMD move like with tessellation, DX10/10.1, etc. jumping the gun - while GDDR5 is at the end of the road its not yet a significant performance bottleneck and shoving it onto the next round of cards will likely provide little real tangible benefit (that isn't to say there won't be performance increases at all) - I really don't understand them sometimes. We do need to get to it eventually but I'd be surprised if it couldn't have been done far more cost effectively.
HBM planning for the future, smaller,less power, more band width, nvidia will use it next year.i don't want to see amd fail, we have only two manufactures so we certainly are not spoiled for choice. i always hope nvidia and amd both have stellar top end products, its more fun ,all round.:)
 
From personal thoughts

Their CPUs are Ok but I always recommded Intel over them.

I always recommded their GPU over nvidia, because for the most part its the same performance and cheaper

However if they do run out I would like to see the following

Samsung use their cash to help fund amd budget not take over just fund.

The huge wealth if Samsung the r&d of amd, Samsung's 14nm production together their could be a chance to fix the 14nm issue with large dies. And make a stinking GPU
To beat nvidia cheaper.

Their CPU division holding not much hope for high end low/medium they do good
 
How are they "jumping the gun" with HBM? They often co-develop the new technologies in GPUs and they are going to need it to update their SoC and custom offerings. Nvidia were right alongside them but made a misstep and got beaten to the punch.
 
How are they "jumping the gun" with HBM? They often co-develop the new technologies in GPUs and they are going to need it to update their SoC and custom offerings. Nvidia were right alongside them but made a misstep and got beaten to the punch.

They often co-develop tech that comes to nothing though. They use this 'nothing' as a spec sheet filler.


What exactly have AMD beaten Nvidia to the punch to ?
 
How much of a tangible benefit is it going to be with the coming round of GPUs?

If their new flagship beats the new Titan (unlikely), without turning itself into a volcano, the cards will fly off the shelves. And if the new CPUs in 2016 are also good, the company will fully recover. Those are quite big ifs...
 
If their new flagship beats the new Titan (unlikely), without turning itself into a volcano, the cards will fly off the shelves. And if the new CPUs in 2016 are also good, the company will fully recover. Those are quite big ifs...
thats why shares in amd MAY be worth a punt, anyone got any inside information ?(trust me with info) :D
 
If their new flagship beats the new Titan (unlikely), without turning itself into a volcano, the cards will fly off the shelves. And if the new CPUs in 2016 are also good, the company will fully recover. Those are quite big ifs...

If their flagship doesnt beat Titan X there will be a hell of a lot of belly laughing. By the time its released it will be almost 2 years since their previous single gpu (290x) so I would have thought it should beat it. It had better beat it for their sake. (In before someone says it will be cheaper)

'Fly off the shelves' is not a term I would associate with any AMD gpu, dont know much about their CPU's.
 
They often co-develop tech that comes to nothing though. They use this 'nothing' as a spec sheet filler.


What exactly have AMD beaten Nvidia to the punch to ?

AFAIK they were the first to GDDR5,co-developed GDDR3 and the first to implement modern 2D/3D switching clocks in a desktop graphics card.
 
Last edited:
HBM memory is mainly going to benefit the ultra high resolution crowd who are a small portion of even this forum, so even if 390X does comfortably beat Titan it's not going to turn AMD's fortunes around by itself, it's the mid-high end (1080P gamers) where the most money will be made.

Assuming 370X/380X etc will just be rebadges of current chips (or slightly updated variants), AMD are still going to have a very hard time if NVidia release a 980Ti and cut the price of GTX970/980 drastically.
 
AFAIK they were the first to GDDR5,co-developed GDDR3 and the first to implement modern 2D/3D switching clocks in a desktop graphics card.

Not sure about gddr5, gddr3 I'm sure was ATI, not AMD and meh for 2d/3d switching.

Philips invented the CD and I still wouldnt buy anything they make.

My point, whatever AMD are doing right now is not working for them.
 
If their new flagship beats the new Titan (unlikely), without turning itself into a volcano, the cards will fly off the shelves.

Only until Nvidia release their faster than Titan X, and cheaper 980 Ti, and also their rumoured performance driver, then they'll stop flying off the shelves, as everyone will be clamouring for those instead, and AMD will be back to where are they now, all depends on when Nvidia drop the 980 Ti, if its about same time as the 390, or just before, then AMDs totally ****** before they've even started, if its slightly later, then AMD will only have that time to get a few sales in.

The 390 will not do anything for them imo, as Nvidia are just too far ahead, they already have cards all ready and waiting to counter them.
 
Last edited:
Good point mmj about the price drops, 970/980 will have been out for around 9 months by the time we see the 3 series and nVidia have already made massive profits on them, they won't mind dropping the price or offering game vouchers if need be to keep the sales up.
 
If their flagship doesnt beat Titan X there will be a hell of a lot of belly laughing. By the time its released it will be almost 2 years since their previous single gpu (290x) so I would have thought it should beat it. It had better beat it for their sake. (In before someone says it will be cheaper)

'Fly off the shelves' is not a term I would associate with any AMD gpu, dont know much about their CPU's.
the 290 and the 290x sold well, nvidia were and are still top dogs.the mining craze was a bit of a double edge sword for amd,they sold a lot of cards then the mining price fell and flooded the market with cheaper second hand cards, leaving them with 290/290x stock.
 
Not sure about gddr5, gddr3 I'm sure was ATI, not AMD and meh for 2d/3d switching.

Philips invented the CD and I still wouldnt buy anything they make.

My point, whatever AMD are doing right now is not working for them.

If I remember correctly it was AMD who started using GDDR5 first with there 4870 and later 4890 cards (the 4850 stuck with GDDR3. This alowed to them to have a smaller bus and so a smaller die. This in turn lead to much cheaper card next to the GTX280 which still used GDDR3 on a 512bit bus (which made it the biggest TSMC made die untill the GM200 came along) even if it made it was a little slower than the GTX280.

It wasn't until the GTX480 when nvidia started using GDDR5 as standard
 
^ Theres that as well.

There is also the issue that many people are happy with the 980s and Titan X, even if the 390 is a bit cheaper but faster there is a cost to the upgrade that likely makes it not worth it for most, especially since they could simply, add a second 980 for example and get better performance for less money.

Plus by then people will be wondering when the next Nvidia GPU will be out. Why buy a 390x in the summer if in 6 months time there is a 16nm HBM Nvidia Pascal?

I think regardless of how well the 390x will be AMD simply won't see the sales it needs. They have lost too much market share already and it will very difficult for them to claim it back.

If AMD start a price war it will be them that loose, Nvidia have already made a load of profit on the 970/980 cards, biggest profit in GPU history supposedly. 390X will be more expensive to produce and if they cut their margins to nothing then even if the cards fly off the shelves it is pointless, their R&D will suffer and the next round will be even harder for AMD to survive.
 
There is also the issue that many people are happy with the 980s and Titan X, even if the 390 is a bit cheaper but faster there is a cost to the upgrade that likely makes it not worth it for most, especially since they could simply, add a second 980 for example and get better performance for less money.

Plus by then people will be wondering when the next Nvidia GPU will be out. Why buy a 390x in the summer if in 6 months time there is a 16nm HBM Nvidia Pascal?

I think regardless of how well the 390x will be AMD simply won't see the sales it needs. They have lost too much market share already and it will very difficult for them to claim it back.

If AMD start a price war it will be them that loose, Nvidia have already made a load of profit on the 970/980 cards, biggest profit in GPU history supposedly. 390X will be more expensive to produce and if they cut their margins to nothing then even if the cards fly off the shelves it is pointless, their R&D will suffer and the next round will be even harder for AMD to survive.

Good read, my thoughts on this as well
 
Back
Top Bottom