AMD never could afford developing their own process, and given their current volumes they cannot afford to develop even a node variant for them. Given that two situations, having or not a fab wouldn't make any difference.
Sometime in the future, the era of AMD being a direct competitor to Intel will be looked at as just that; an era. Because it's over. And it ended the second they decided to spin off their fabs.
my point being the vast majority out there will be more than satisfied with kaveri. With the vast majority being so...vast, sometimes we forget about countries outside the US and developing nations that can't afford the intel premium, where amd could serve.
Funny to see you talking about Intel premium. AMD has atrociously low gross margins, the company is losing money, without scale to sustain more than a small foundry, bleeding cash and the biggest silicon they got on the bottom market is 245mm^2.
Intel, OTOH, has margins above industry average, has healthy gross profits, healthy cash flow levels, immense scale of production enough to sustain a network of foundries and the kind of stuff they sell on the bottom market is not far from the 100mm^2 mark.
We have one company with products economically suitable for the bottom market, and actually making money on it and another company being financially trashed every quarter because their products aren't cost competitive. It's not hard to see that current price levels aren't sustained by AMD, but by Intel. AMD, in fact, desperately needs a price increase on their APU line up.
But despite these facts, you and a few other people aren't shy to call Intel out for price premiums and stuff like that. I don't get it.
AMD showed a profit last quarter, and is expected to do the same this quarter. They've also kept their long term debt growth to a minimum, and managed to drop their debt to assets ratio in Q3. They will need to continue to do this if they hope to stay viable into the future.
Whether they can sustain profitability over the next few quarters is of course the big question. And they will need said profits for R&D and such.
AMD's quarterly, and presumably annual statement will be released on January 21st, so we will have a better picture r.e. if the board's strategy of meaner and leaner will continue to bear fruit. But you can see why Kaveri is such a big deal to them, and why the PS4 and XBox One wins could not have come at a better time.
Not to mention the R Series doing as well as it is, although NVidia will very likely take the crown back soon, when their next batch of products is released.
http://en.wikipedia.org/wiki/GeForce_800_Series
tick, tock!
Is the R(9) series doing that well? USA yes, rest of the world? Not so much. AMD lost marketshare again in Q3 as well.
Where I am most 280X and 290 SKUs are completely sold out. Our leading online retailer has 11 280X skus - only 2 of which are available, and 8 290s sku of which 3 are available.
Meanwhile I have a choice of every 780 sku they stock. So yes, it is doing that well and not just in the US
Is the R(9) series doing that well? USA yes, rest of the world? Not so much. AMD lost marketshare again in Q3 as well.
AMD recently had to borrow another 500M$. And they have a delayed 200M$ WSA payment for Q4. So it needs a minor miracle to pull a profit in Q4. Plus you can say Q3 is artificially higher for the same reason.
In europe you can buy them by the 100s in regular online stores. Same in asia.
and NVidia's mobile strategy is now more important than ever.
Are you referring to discrete GPU marketshare in Q3?!? Yeah, I saw your chart, but dGPUs are becoming less and and less of the total overall market. A GPU is a GPU, no matter where it is used (APU, XBox, R series, etc.).
There's a reason Intel has entered the iGPU market...
And do keep in mind AMD was 'in between' releases in Q3, so I would expect NVidia's dGPU market share to grow at that point, as it was NVidia's turn at the top of the heap.
SINCE that time, PS4, XBox One, the R series hit the market. So I wouldn't hang too much on NVidia's gaining dGPU market share in Q3. NVidia losing out on the current generation of consoles has left the company scrambling, and NVidia's mobile strategy is now more important than ever.
Kaveri will be joining XBox One,PS4 and the Hawaiian Islands cards in the GPU sales department this quarter as well. So for the short term at least, AMD is strongly positioned on the GPU front, which is why Toms and other tech/financial sites can easily see AMD gaining 40% GPU market share within 6 Months.
Unless you can share projections that state otherwise?
Keeping that 40% market share throughout 2014, of course, will be another matter. I can see NVidia, etc. getting some nice tablet wins and such between now and Christmas.
I have to wonder if you're exaggerating here. I mean if I go on overclockers.co.uk there is definitely availability, but alot of 280/290 SKUs are pre order only. Not exactly what I'd call plentiful, let alone availability by the 100s
Remember all prices include 25% VAT.
Hell, you have 25% VAT in Denmark ?? and I thought 23% here was high.
If you want to count IGP, that 40% number goes completely wrong for you.
Q3 2013.
The entire console "wins" is barely offsetting the losses that AMD suffered in revenue in its other segments. And Kaveri wont change that, just continue the slow decrease in marketshare and revenue for the CPU division.
AMD showed a profit last quarter, and is expected to do the same this quarter. They've also kept their long term debt growth to a minimum, and managed to drop their debt to assets ratio in Q3. They will need to continue to do this if they hope to stay viable into the future.
With Kaveri, AMD chose to use GlobalFoundries’ 28SHP (28nm super high performance) process technology that was designed with various types of chips in mind. Improved transistor density and thinner elements of 28nm process tech allowed the company to integrate whopping 2.41 billion of transistors into a 245mm2 Kaveri die, which is 85% higher transistor count compared to Trinity/Richland design (246mm2, 1.303 billion transistors, 32nm SOI). Now the company has four Steamroller cores, hUMA [heterogeneous unified memory architecture] memory controller, Radeon graphics processing unit with 512 stream processors (AMD calls them eight GPU compute units) and a lot of various special-purpose hardware inside its flagship APU.
We already know this. It was stated on an AMD slide from their Kaveri presentation.BACK ON TOPIC, XBit labs had this to say about Kaveri's 28NM SHP (Super High Performance) process. A quote from the article:
Link to article:
http://www.xbitlabs.com/news/cpu/di...d_Using_APU_Optimized_Process_Technology.html
We already know this. It was stated on an AMD slide from their Kaveri presentation.
That's the first time in this thread that I've seen the abbreviation for SHP (super high performance) spelled out, and also the first confirmation I've seen in this thread that Global Foundries is doing the manufacturing (I may have missed that, but I don't remember a definitive answer to that question). Plus, I share the articles as they appear. You never know when some odd comment might appear in an article...
That's the first time in this thread that I've seen the abbreviation for SHP (super high performance) spelled out, and also the first confirmation I've seen in this thread that Global Foundries is doing the manufacturing (I may have missed that, but I don't remember a definitive answer to that question). Plus, I share the articles as they appear. You never know when some odd comment might appear in an article...
Lol, people have been saying this since Moore's Law was first stated. What a joke.The next 10 years are not going to be at the same rate of chip development we have seen over the past 20 as the process continues to get smaller. When we get beyond 14nm into the 10nm...7nm..2nm process, its going to finally highlight the problem with moores law. AMD I firmly believe will eventually catch right back up to intel in due time as it takes longer and longer to develop the next generation of chips unless something revolutionary happens. I think AMD knows this and this is the reason why they arent so hell bent on pressing the issue with advancing the process as fast as intel is and focusing on more interesting designs like with the APU. Thing's like HSA, huma, and mantle.