AMD CPU speculation... and expert conjecture

Page 36 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
^^ I'm just saying it in advance, because EVERYONE will be informing people to buy PD/EX because "games will be more optimized at the end of the year, because consoles will have more cores".

Trust me, in 2-3 months, you WILL see recommendations like this on a fairly decent basis, picking up as you approach black monday (when I assume the releases will be). And when the people who made these recommendations turn out to be wrong (for reasons I've long since explained), I will be here, saying "I TOLD YOU SO!!!".
 

Cazalan

Distinguished
Sep 4, 2011
2,672
0
20,810


Should just remove the Steamroller from the title. There's like 4 more chips due out before it.

AMD speculation... and expert conjecture
 

blackkstar

Honorable
Sep 30, 2012
468
0
10,780


If this is true, it is about AMD shifting games from wanting a single powerful thread into lots of weaker ones. It is not about taking advantage of Jaguar, it's about taking advantage of several weaker cores.

Doing so would benefit AMD greatly. Currently, the ~$125 range is occupied by dual core i3 and 3 module/6 core FX 6300. If games wanted 6 or 8 weaker cores as opposed to two stronger ones, it would put Intel in a very bad position for the budget gaming market.

It is a dream come true for AMD. It will basically force developers to design games around 8 weak cores. When that translates into 8 stronger than Jaguar cores, it'll put AMD into a very good position.

http://www.tomshardware.com/reviews/geforce-gtx-660-ti-memory-bandwidth-anti-aliasing,3283-13.html

Pitcairn is also better than even GK104 when it comes to memory bandwidth intensive situations. By gaining control of all high end systems, AMD is basically forcing developers to focus on bandwidth intensive graphical improvements over shader based (where Nvidia is strong), and it's forcing them to prioritize lots of weaker threads over a hand full of strong ones.

If this is true, AMD basically just crushed TWIMTBP out of the park, and they're going to get paid to get games optimized for their hardware. And then what would Nvidia do? Approach game devs and want them to optimize for Nvidia GPUs? Because prioritizing PC market over consoles has been the trend for the last 5 years...
 

jdwii

Splendid




I agree everyone will say this but i do believe the PC may get more ports if its easier to make games for, what do you think about the 7970M in a console if it can go to a laptop should run fine in a console right?

I'm sure by now the next systems will have close spec's and the reason the 8 core is in their is for multitasking skype plus gaming, maybe even for kinect

Might be Amd's way or hope more games are more multithreaded or at least using 4 cores.
 


It is a long term plan they have in their hands, true. And it is a scary thought that they could piggyback TWIMTBP like this if you're correct.

Let's see how the marketing hype/mob from nVidia and Intel play against consoles next generation if they have AMD hardware in it.

Cheers!
 
that might just be too much thinking into it. My opinion is : its likely small things will change but it won't really be massive like you are expecting.
 

tbh, i was kidding. later i asked around and found out that core i3's are selling quite well.
this is a common trend among amd fanboys. every time new games, new hardware come out and they started pushing 4+ core cpus on anyone who wants to build a gaming pc. i am still waiting for the thubans to become relevant. although, if the new consoles indeed can push the cpus, i'll be quite happy.

it's not directly related to consoles, but i'd heard this 'gaming will use more cores' line back when 'world's first 6 core mainstream desktop cpu' came out. and every year i hear 'we'll see in a year' or something similar.
years are still passing and i am waiting for the dualcore cpus to take some serious beating in general gaming. by dual cores i mean intel dual cores, including core i3.

you seem to be under the impression that it's amd who decides how the console cpus will be designed and how console games will be developed. afaik it's the opposite. i hope people better versed in console development can shed some more light on this.
besides, the octo core jaguar architecture could very well be a derivative of the server version of jaguar but heavily customized for consoles. just saying.

afaik, no. the hexcore (and higher) fx and phenoms are still low-selling. i am disregarding recommendations because those don't directly translate into cpu sales.
some games, especially multiplayer fps, already gain a lot from multicore cpus vs dual core cpus. intel doesn't seem to be bothered by it much, revenue-wise.

i highly doubt there will be a retail, mainstream desktop version of 8 core jaguar this year, at least. the rumors i've been reading points to 4 cores max, with radeon (gcn/hsa) igpu i.e. apus.
what good position?
.....

no one is crushing anything. if twimtbp and physx crushed amd, then 'gaming evolved', amd's hd3d standard, opencl would never have seen even minor success and vice versa.
nvidia has maintained much better isv relationships than amd so far. amd is getting off their asses but it will take a while to catch up. gaming evolved is a decent start.
when new consoles come out and if they succeed financially, then devs will switch their attention to those and pc will once again become less priority.

 
Hard to comment on SR, given how its a year off and we have no news.

But we DO have AMD's quarterly financials:

http://www.amd.com/us/press-releases/Pages/press-release-2013jan22.aspx

AMD (NYSE:AMD) today announced revenue for the fourth quarter of 2012 of $1.16 billion, an operating loss of $422 million, and a net loss of $473 million, or $0.63 per share. The company reported a non-GAAP operating loss of $55 million and a non-GAAP net loss of $102 million, or $0.14 per share.

For the year ended December 29, 2012, AMD reported revenue of $5.42 billion, an operating loss of $1.06 billion and a net loss of $1.18 billion, or $1.60 per share. The full year non-GAAP operating income was $45 million and non-GAAP net loss was $114 million, or $0.16 per share.

"Better then analysts expected", though.

EDIT

Its worth noting, AMD's yearly losses are now about the same as its cash on hand (~$1.2 Billion). Which means, even if AMD's financials do not get worse (they've been declining for two years now), AMD only has enough cash on hand to survive another year before they are flat broke.

Point being: AMD needs revenue, ASAP.

EDIT2

http://news.cnet.com/8301-1001_3-57565281-92/amd-limps-through-q4-and-2013-doesnt-look-much-better/

AMD's outlook for the upcoming 2013 fiscal first quarter is bleak. AMD expects revenue to decrease by a further 9 percent, give or take 3 percent.

AMD is in very, very, bad financial situation right now.
 

Cazalan

Distinguished
Sep 4, 2011
2,672
0
20,810


They would have been profitable if they didn't pay off GFlo. Time will tell if that move will ever pay off. Unlikely with Apples move to TSMC.

Microsoft/Sony/Nintendo have a vested interest to keep AMD afloat, but expect more downsizing.
 


And...are they making a profit on those parts? Are those parts high profit-margin, or just eating wafer space that could be used for higher grossing parts? How much of a loss will AMD take in 5 years when they have to continue to produce an (by then) aged architecture? Nevermind the assumption the consoles will sell well, which may or may not be the case.

AMD doesn't need market share; they need REVENUE. And if revenue does not pick up, expenses dropped, or cash raised (partial sell off), AMD will be out of cash (bankrupt) this time next year. [And no, when you cash on hand = x, and you yearly losses >= x, you are clearly close to bankrupt.]
 

lilcinw

Distinguished
Jan 25, 2011
833
0
19,010
Most of AMD's losses were due to one time (supposedly) charges including $700 million to GF to break/amend contract. Before all of that was charged they actually made a slim profit for the year.

I'm not saying everything is roses and sunshine for them but they should be able to last more than one year before they are looking at bankruptcy.
 

Cazalan

Distinguished
Sep 4, 2011
2,672
0
20,810



The consoles may tank but if they keep up with historic sales you're looking at 300+ million units over 5 years. If CPU+GPU you could modestly say $60/unit. That's 18 billion.

They still need to downsize and sell property but it can provide cash flow to keep on top of their loans. No doubt they are hemorrhaging cash and the decision to cut off GFlo has cost them dearly.
 

Cazalan

Distinguished
Sep 4, 2011
2,672
0
20,810


If you can trust their non-GAAP earnings the very worst has already happened. They also had to payout for the severance packages. They're still in a tough place but they're not going bankrupt anytime soon either.

Longer term they have to deal with 28nm product trying to compete with 14nm product. The gap is widening and Intel isn't letting up.
 

jdwii

Splendid



Its probably going by core and that is just plain retarded.


As for console sales its obviously going to make Amd money, to say otherwise is just nonsense.
 


Lets do math!

18 Billion over 5 years = 3.6 Billion/year.
3.6 Billion/year / 4 quarters = ~900 Million per quarter, using your estimate.

But remember this: its not like the die space these chips are using would NOT have gone unused if AMD hadn't won the contract. So the real question, is as follows:

"Is the die space being used for a higher-grossing part then would have otherwise been used?"

If not, then revenue would actually DECREASE by virtue of less profit per part sold. I've seen multiple companies over the years tank because they were contractually obligated to continue to produce parts they didn't turn a profit on. Simple example: Assuming the PS4 uses a BD based arch as has been rumored, how much is AMD going to lose by continuing to produce BD chips into 2018-2019, instead of producing newer, higher grossing designs?

These are the questions that concern me. I'm worried the console "wins" are going to be "losses" in 3-4 years.

[Isn't economics fun?]
 
Status
Not open for further replies.