AMD Opening Up Mantle to Additional Developers

Status
Not open for further replies.

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20
OpenGL does the same stuff as they showed already. DX12 may not do it for a while (2015?) but OpenGL does it already and has for a few years. So devs, write for AMD's small subset of their own cards (only GCN) or just use OpenGL which runs on everything easily portable to everywhere. I really thought AMD would have dropped this by now and started spending more of this money on DRIVERS for all their current cards! You know, to avoid a phase4 soon etc or more driver phases for the next chips... :(

http://blogs.nvidia.com/blog/2014/03/20/opengl-gdc2014/
Watch the vid at bottom AMD ;) One of your people was involved in this show, so why keep pushing proprietary for something that is already working on all cards TODAY? Proprietary crap works when you get there first and are special (doing something others can't, thus a NEED for your tech, like Cuda, it was their only choice back then, so it got entrenched). It doesn't work when there is an established product already out that does everything your proprietary stuff does cheaper. This would make more sense if OpenGL didn't exist, and only options were Mantle or DX12 which is ages away (thus Mantle would have a chance to get entrenched). SteamOS will be pushing OpenGL, and same for Android etc if a game needs a high number of draw calls etc.
 

TechnoD

Honorable
Jun 25, 2012
293
0
10,810
20
OpenGL is, as others have stated, a better option (multi-platform). However, I do believe AMD is pushing mantle for profit only; when a user starts a game and sees "open-gl" they don't really know/care what it is. However, when a user starts a game using mantle and sees a mantle logo with AMD logo, they associate the software with AMD, and thus, AMD gains brand recognition. Other than that i don't see any other reason for mantle.
 
Mantle is faster than OpenGL. Since it is developed for just one card it is able to work with much less overhead than OpenGL or any other general use API. This not only is letting them have faster performance but makes porting games with higher quality settings enabled to consoles easier. If it wasn't better, then the many major companies turning to use it, wouldn't be turning to use it.
 

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20
AMD might like this idea, but I fail to see how anyone will want to use it without an AMD check tied to using it. It just makes no sense for a dev to show any interest without a check first. No matter how good or bad it is, Mantle requires EXTRA work over what you already have to do to support your main users (DX, OpenGL etc, whatever you use for the rest of the world not on GCN cards).

Extra work/cost for no financial gain in a crap economy? Not even if I am in love with Mantle and how it works. Not even if my (future?) wife was the one who came up with it...LOL. OK, maybe then...But only because she'd have something I want daily...ROFL. All AMD has for my EXTRA work is what, a pat on the back and thank you for helping them tackle NV? Come to me with an idea that makes me MORE money and I'll adopt it. Come with more work for ZERO extra money and you'll be denied.
 

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20


Please watch the vid link I pointed to. 15x faster draw calls. AMD set out to solve a problem OpenGL already had licked ages ago. As he shows in the video, people just didn't know how to fix it, but it was in there as he demonstrated. Just changing one dev's code with a few lines caused a 4x draw call speed up and he comments you can double that easily with more than the simple short phone call they had.

You're not making sense. Nobody is turning to it that has NOT been paid (IE EA got $8mil to use it etc). We will see how many sign up to do FREE WORK to help AMD's cards (a small subset of them, only GCN) look better when their time could be spent on EVERYONE else at once with OpenGL. The people in the audience for that vid skipped the VR session next door (oculus stuff) to see the DRAW CALL session for OpenGL. Even the guys on stage were surprised people did that. Clearly a lot of interest in finding out about some stuff in OpenGL already that they were NOT aware of. The guys doing the speech admitted they'd be next door in VR session if they weren't giving the speech...LOL. They thought the room would be empty, but it was full. NV should have done this speech the day AMD said we needed Mantle.
 
If you think Mantle is pointless then you haven't done enough research. People point to "draw calls" but it's far more than that. Also, OPENGL does not solve all the issues which is why developers asked for a new API built from the ground up to solve some headaches.

With DX12 coming I'd normally say Mantle might be doomed but since the Steambox is likely to be a huge success (albeit slowly) and can only use Mantle or OpenGL I think Mantle will be successful.

No DX on Steambox but both NVidia/AMD so it will be interesting to see if programmers just stick with OpenGL (especially when it improves) or what.

This is pretty useful: https://www.youtube.com/watch?v=sSY2KXBoro0
 

BriboCN

Honorable
Aug 29, 2012
97
0
10,640
2
I think those judging Mantle against OpenGL and DirectX need a bit of perspective. Although it is absolutely fair to compare Mantle against its competitors, it is important to remember that OpenGL initially released in 1992. DirectX released in 1995. The fact that Mantle is at least on par with, and in many cases beats OpenGL and DirectX deserves consideration.
I think the fault of Mantle thus far is in its marketing. Like other members of the video gaming industry they have chosen to market their product based on an ideal, and absolutely unrealistic, set of hardware and circumstances. This approach works wonderfully for selling individual graphics cards, but terribly for an API designed to be used on an entire product line with an infinite number of hardware configurations.
I believe that if AMD was absolutely honest about Mantle it would still be successful.
If they told the gamer crowd, you paid top dollar to get the best performance possible, and we can give you an extra 3-5% gain for nothing. It would be embraced by that crowd.
If they tell the casual crowd, we created this API so you can run the games you want and won't have to upgrade your PC this year. It would be embraced by that crowd as well.

Mantle is in the position to thrive, they just need to fire their marketing department.
 

firefoxx04

Distinguished
Jan 23, 2009
1,371
1
19,660
144
I agree that AMD markets Mantle in the wrong manner.

At first everyone thought it would just be a performance boost for everyone.. truth is only those with weaker CPUs see the gain. Fortunately, anyone with an AMD cpu will see a gain more than likely simple because they are easier to bottneck on than any i5 or i7 cpu.
 
I haven't looked too much at the marketing of it, just thought the benchmarks and reviews were following the R9 290x cards.

It would make a lot more sense for them to run the marketing and development on a PS4 or a Radeon 7850 or R9 270 with a FX 6300 CPU. That would put it much more mid-range. Also help get more funding for the program if they did it on the PS4 side also.
 

ohim

Distinguished
Feb 10, 2009
1,195
0
19,360
26
Mantle is not proprietary (closed) to AMD, they said Nvidia is free to use it, but as you can expect that won`t happen unless Mantle gets really popular.
 

irish_adam

Distinguished
Mar 30, 2010
217
3
18,685
0
i really dont see what all these nay sayers problem is except that obviously they all run Nvidia cards

Of course your all right these game developers are all idiots, they have no idea how to make games or whats the best API for them. You guys should all group together and take the world by storm, you'll be multi millionaires in no time!

or you could apply some logic to your argument. If opengl fixed all the problems that mantle does a long time ago then why is support for it so shit? i mean if its so much better than directx then surely its in the studios best interests to use it, i mean its multi platform which would give them access to a bigger audiance. That is of course unless microsoft is paying each developer for every game that uses directx which i highly doubt as it would cost them a fortune over the years. Of course the makers of opengl is going to stand there and say that theirs is better and has done what mantle does all along, thats called marketing, they are hardly going to stand there and admit that their API is shit.

At the end of the day though, the more APIs the better really, even if you have an Nvidia GPU (which AMD did state that they could use mantle) It gives the other APIs a kick up the bum to make theirs a tleast as good. So how about instead of bitching about how stupid it is and how stupid all these developers are for getting on board, you celebrate the fact that at long last these APIs will get a major overhaul which lets admit has been a long time coming
 
At first everyone thought it would just be a performance boost for everyone.. truth is only those with weaker CPUs see the gain. Fortunately, anyone with an AMD cpu will see a gain more than likely simple because they are easier to bottneck on than any i5 or i7 cpu.
:ouch: WARNING: Internet Propaganda Fail :ouch:

4770K 10% FPS+ In BF4 with R9 290X & Mantle over DX
(59% increase for Kaveri CUs)

From Tech Report

That is all.
We now return you to non-reality-based programming.





 
That benchmark looks very believable. The 290x has no trouble getting high FPS. Mantle has been organized to work better with weaker CPUs and more cores. So it makes sense that the CPU is bottlenecking it, Mantel helps signifanctly to reduce the impact of the weaker CPU. Compared to the 4770k where the CPU is not a bottleneck, it has some of the other improvements from Mantel but does not benefit from the CPU related tweaks.
 
What the gentleman said: "only those with weaker CPUs see the gain"

I linked a Tech Report review that showed the 'weak' i7-4770K with a 10%+ gain in BF4 with Mantle over DX.

That's the point I made -- the exact, 180-degree opposite of the postulate espoused by the gentleman from Planet Claire. He drives a Plymouth Satellite faster than the speed of light. Some say he's from Mars. But, he's not.

I'm not arguing semantics one way or the other.
As far as I might know it's a combination of Mantle, a BF4 patch, a Catalyst driver, better use of the CUs and divine intervention.

 

Mac266

Honorable
Mar 12, 2014
965
0
11,160
70
hahahahahaha it must be mosetly AMD fanboys watching this thread, all the posts about 'openGL being better than mantle' are voted down :p
 

tobalaz

Honorable
Jun 26, 2012
276
0
10,780
0
I don't care if Mantle stays or goes so long as it gets MS and the Khronos Group off their asses and pushes them into refining Direct X and OpenGL to use less cpu overhead while becoming lower level maximizing GPUs in the process.
If it stays around, more power to them, if it doesn't, it was a swift kick in the ass that was desperately needed. Either way, we gamers win.
 

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20


Yes it's quite comic. Nothing I said is wrong (most others spot on also). It costs extra money to code for mantle period and you gain no extra cash doing it if AMD doesn't write a check. OpenGL does the same and even dx11 as NV's new driver shows vs. mantle games. If that doesn't do the damage DX12 will as MS fires up their "you require win9/dx12 to play games at all or you'll just miss out on the world and die as your head explodes from not running win9" campaign that is surely coming at some point. :)

Ignoring reality doesn't make it go away...LOL. Any money AMD spends on Mantle from here forward is wasted R&D and chasing the dream will just cost more checks they can't afford to pay to get it used. This is something Apple could pull off as they could bleed forever until they win the api war, but this is a loser for a company nearly bankrupt. If MS actually gets threatened by Mantle at some point (say if steam sells 30mil consoles with AMD cards inside...Ok NEVER happen), they would just pay every dev a few million to use DX12. They could easily afford to blow a billion on doing this and laugh, which even at 10mil/game dev that would be 100 devs with the cost of their whole game paid for in most cases and surely they could do better than that. Considering they've lost a few billion over xbox360/xbox cycles clearly they are willing to bleed for a decade or two just to play the console game (much like NV with the first 4 revs of tegra just to get to desktop gpus in SOCS with K1 fianlly).

MS bled on IE also until netscape died. etc etc...This is a tactic they will use if needed and steamos I suspect all will go opengl just because it's portable everywhere else (mobile, ps3, ps4, windows etc). Valve chose NV and they'll want OpenGL used as they are smart enough to know mantle can't dethrone DirectX so any effort for them there is wasted. They will force people to use OpenGL because it can win as android+linux will=more machines running capable opengl gpus than dx at some point most likely with steamos in there and valve+google both pushing opengl. 350mil pc's sold yearly vs. 1.2B mobile already, but mostly weak gpus except the top end. This year the bulk get pretty capable at 20nm in the back end and especially next year. That will be a lot of OpenGL capable devices to aim at that can put up xbox360/ps3 like games on gamepads (why would you go mantle?) for devs on smaller budgets that will aim at a common denominator first always. Too many fanboys here who ignore the basic facts staring them in the face.
 

irish_adam

Distinguished
Mar 30, 2010
217
3
18,685
0



epic wall of txt bro

You say that any money AMD wastes on mantle from here on in is a waste right? Well you are aware that both AMD and Nvidia both help game developers customise their games to work well with their hardware right? Well what AMD is not that stupid, they created a platform to make that easier. Sure the first few releases took time and probably a lot of effort while they sorted out problems with the API but once they have these things sorted they have a solid blueprint that they can apply to any game or studio under the gaming evolved brand.

You say that AMD is on the verge of bankruptcy yet their last quarter they actually posted a profit and even though they made a loss on the year it was tiny compared to the year before. AMD is actually turning itself around, they were selling more GPUs than they could make, they have their APUs in both the xbox one and the PS4. This all from a company that is a tiny fraction of the size of intel.

You say we will get 20nm chips in phones this year but i really doubt it, all current info points to late 2014 for mass production of 20nm which is why AMD and Nvidia have pretty much said there wont be any 20nm GPUs this year. Hell no one has even announced a 20nm mobile processor yet and it takes 6-12months normally after that before you actually see it in a device.

Also you say that valve will move steam to opengl but you dont seem to understand how steam works, how many games do valve actually make? all those games that it has are mostly from other publishers, they have no say in what API they use at all and valve isnt going to recode every game it has to work with another API. Of course steamOS and the games that will run on it will use opengl thats pretty obvious considering that its linux based and therefore they have no choice as microsoft will never let them use directx on it and mantle wont support the back catalogue.

Once again you consider yourself far more intelligent and capable than the people that work for these companies. They are not stupid and as you pointed out AMD hardly has the resources to be pouring money into something for no reason, they and the game publishers seem to think its worth using and developing. I trust them to be right far more than you. Personally i have been wanting a wider adoption for opengl for years, for me even if mantle doesnt last long and even for people with Nvidia cards its been a positive influence on the gaming community even if its just because microsoft and Khronos are scrambling to push out updates that compete
 

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20


My post was 474 words (go ahead paste it into word it will tell you). Your post is 482 words.

Epic wall of text bro...blah blah...Who cares if it's chock full of good info? I never understand someone complaining about DATA. I can type 60wpm so it's not hard to type a "wall". I will welcome a BOOK post from you if it informs me of info I do not know (opinions backed by data). I put out "walls" at times just to test the data or theories to see what comes back. If nobody can argue with the data or point out stuff I'm missing I'm more confident in my stock trades. I see a lot of opinion in your post but no data. How much does a company have to get wrong before you'll say they are stupid? How much do they have to lose? If you've never made a dime over the life of your company did you screw up? If you're the #2 supplier of gpus and cpus and you can add up ALL of your years and you've lost money are you making massive mistakes in your opinion? Let me tell you why I think they're "stupid". Ready for another wall? :) Feel free to move along but if you own either stock I'd at least give the data some thought. Great employees at AMD mind you, but management? UGH.

I'm aware of what both companies do with gaming and helping them out. I was a partner of both companies for 8yrs. I still have an AMD shirt, and Nvidia/AMD swag/bags etc from years ago from going to the conferences. AMD created mantle to help them gain market share from NV and defend against Intel while having weaker cpus and probably prepping for ALWAYS having weaker cpu/gpu going forward as their resources have no way of keeping up in R&D. But OpenGL has already had the same crap as mantle basically for 3yrs. NV shows with their driver that topples AMD in mantle that there is no need for it even with DX11. They beat mantle with DirectX11 for star swarm which shows you don't need MANTLE to get DX11 to work just as fast. If you can't beat NV using DX11 in your SHOWCASE scenario for Mantle by swarming objects on a screen like no game before it, then you have a problem. OpenGL will start being used in the same manner now that NV covered exactly how to do it (and showed results) on stage. You might be interested to know people skipped the VR speeches for the NV OpenGL speech. The speakers were quite shocked and even admitted they'd be in the other room were it not for having to give the OpenGL speech. So I think its pretty impressive people skipped the VR stuff for OpenGL. AMD created something that had no need to exist. First few releases took time? Every one is taking MONEY too. There is no financial reason for a dev to code for Mantle if AMD isn't writing checks. Cuda would not be a standard if Nvidia hadn't pushed it with billions and 7yrs of ground work. How can AMD do this for Mantle? Why would I code for it at all when a DX11 update from NV drivers catches it in the best case scenario without ANY work from me? Whatever I thought I'd gain from coding for mantle just showed up via a driver updated by Nvidia? OpenGL already can do this too? DX12 coming? Everyone just stopped wasting time on mantle if not paid to keep doing it. It's just common sense.

Name a company using mantle that hasn't been paid to do so please. I don't know of anyone using it that isn't being paid. IE, 8mil just to EA for using it. Note also that when a game is in trouble at launch immediately mantle goes to back burner even WITH AMD paying a check. Nvidia isn't paying companies to test grid. Perf without massive changes or costs for the company you're trying to sell that perf to, actually sells itself. Mantle is a tough sell from a weak position. NV can't even push physx and they have no $$ troubles (same with tressfx, etc). You'll get a few mantle titles a year if that and as with those, ONLY when paid by AMD/NV. AMD has no pricing power either in any product segment they are in, which also makes things far more difficult.

RE: Their financial status -
http://www.macroaxis.com/invest/ratio/AMD--Probability-Of-Bankruptcy
50% chance of bankruptcy in 2yrs. Far better than a year or two ago where it was almost a given. NV is less than 15%. They have to earn 200mil+ just to pay interest on their 2B+ debts. They have 1B in the bank (over leveraged, and credit rating shows this, junk status). NV has 3.5B+ and basically no debt (shows a billion+ higher but I'll explain why I think that is). They took a long term debt of ~1.5B but it shows as mostly on the balance sheet still with an extra ~1.5B in both assets and debt. I think they took this for the icera purchase and most (I'm guessing why it's still here, showing 1.1B in extra short term assets so icera probably got some of it) is still on the sheet because NV wisely put off their HQ upgrade. It's a bad economy, so good choice to delay that. I hope they invest it in their own stock.

AMD is barely making money only quarterly and over the year NOT as you already stated. It's a waste of money seeking proprietary stuff when your drivers are in phase 3 getting you a bad rap, retail cards flopped on 290x, had to speed up the fan via drivers for 290 and ask reviewers to re-bench everything, gave up the cpu race with Intel etc etc. You can't come up with proprietary stuff that is very hard to push as the weakest guy in the market, while ignoring your core products. Microsoft can't even pull big changes to an OS off with win8 and they have major muscle/cash. AMD lost $6B+ in the last 10yrs. NV made ~3.4B during the same time. Are you not understanding the difference between these two companies? At the same time check AMD's stock vs. NV's.

http://investing.money.msn.com/investments/financial-statements?symbol=AMD
AMD 10yr summary. Note the SHARES OUTSTANDING. Double what it was 10yrs ago, showing MASSIVE dilution of the stock. Stock people know this and it's BAD.
http://investing.money.msn.com/investments/financial-statements?symbol=US%3aNVDA
NVDA 10yr shows buying back shares and almost back to 10yrs ago. They will be this year with another Billion to spend on buybacks/dividends (most earmarked for buybacks they say).

Now take a look at ASSETS. AMD had $13.15B in assets in the last 10yrs, but now down to 4.34B. So if you lost 2/3 of everything you owned over the last 10yrs, you'd say you're up a creek correct? At best, in terrible shape compared to even 5yrs ago at 9B assets. Go ahead and take a look at NVDA. Increasing from 1.63 to 7.25Billion in assets over the same time and only went down slightly in 2009 due to bumpgate payouts etc. But up all other years each year during the 10yrs. We could keep going but I think you get the point. IF you don't just read those sheets some more. These two companies are not even on the same playing field financially. That's not AMD hate, it's reality. Having said that if I see AMD has things that I think can make them money I'll be buying the stock that day. I don't hate the company I hate the way management has been running it. Breaking even won't get you to #1 especially after 6B in losses over 10yrs on top of huge asset losses, shares diluted etc. Did I mention NV has a 2.2% dividend yield? So all the stuff I just said going up has happened while giving share holders some cash back. WOW. Yes, I like my quarterly check while I wait for the stock to explode.

"You say we will get 20nm chips in phones this year but i really doubt it"
OK, lets look at that:
http://www.digitimes.com/news/a20140311PD203.html
TSMC moving AHEAD of schedule with A8 20nm and NO shipment will be delayed. Samsung is starting production in april (so even they should have a A8 in a phone for apple by dec). TSMC ahead of samsung says you're wrong correct?
"TSMC expects to continue advancing its manufacturing efficiency over the next quarters and have 20nm products make up to 20% of its total 2014 revenue."

You can't make 20% of your revenue this year on something you're not selling right? It takes 4.5-6 months to get products out the door from the time a SOC arrives on your doorstep (google did nexus in 4.5 the record so far AFAIK). NV is said to be pulling in their chip to try to get into Q4 but I will say even I doubt this means in a product though it is possible but I think apple gets it all for xmas. They'll be shipping I think, just not in a device. Even if I'm off by a quarter on 20nm it won't change what I said is happening, it's just 1-3 months later. I said "This year the bulk get pretty capable at 20nm in the back end and especially next year." It seems I was pretty clear the REAL damage happens next year (when everyone has them in devices). Iphone 6 comes in Aug, and 20nm so clearly someone can get in by 4-5 months later for xmas sales. Apple had to switch fabs too, so people are not that far behind. You are comparing a 550mm2 gpu to a 40-120mm2 SOC. SOCs come first as they are less risky than tossing HUGE chips in the trash. Note TSMC says yields are up on 20nm. Everyone will be shipping probably this year, but we'll have to see how many make it into xmas devices. Apple will be this year. Qcom announced 810/808 both 20nm but no word on xmas or not but I doubt it, as said I think apple gets almost everything but only because of the almighty $$. They had a lot to do with getting the TSMC 20nm process fixed I'm sure (probably many people on site helping, they own a fab too so not fab ignorant today).

NO 20nm gpus (agreed), but that wasn't what I said was coming. SOCS first, I said mobile.

I think you don't understand the power of owning the #1 digital distribution point by FAR. It isn't just about Valve's games (not at all). If you want to sell your game on a PC (mac now too I guess, hence the massive jump in active users over the last year), you pretty much do whatever you can to get your game on steam. Everything else is 2nd or worse by a long shot. You do not ignore 75-80mil ACTIVE users on steam. It's kind of like Intel telling you if you support AMD cpus on your boards you probably won't get our next chipset at the same time as everyone else when our next gen hits the shelf. I was selling Asus boards that came with NO label and in a white box. Valve has a lot to say about your NEXT games on steam, which prompts you to negotiate especially if valve or nvidia etc is footing the bill (not coding assets here, just compatibility). You don't have to recode every game, just the most popular few hundred (half-life 2, portal, etc etc, most of valves stuff already works, I'm talking others) until more new IP comes to ARM & Linux/steamos. NV has shown it takes 2 weeks or less. Serious Sam 2 and Trine 2 both took two weeks with most time spent on mapping buttons not coding. Portal and Half-life 2 came in the same week. This isn't that difficult and they already have experience as OSX games are popping out right and left and they are running OpenGL (pc ports). Gog sells both for many games. Many games launching on PC, android, ios, osx, linux now. PS3/PS4 already run it too, so that is an easy port also. The wave of the future is portability, and that is most easily done on OpenGL.

I never said I was smarter than anyone, but I do have an IQ over 130 ;) So that puts me in the top 2% of the world last I checked. I'd join mensa but for $70 a year? I don't see the point other than bragging rights and who would I tell that cares? LOL. That said, I would not have bought ATI for 3x what it was worth and I said so at the time and many times since. That purchase price cost them the fabs and has been killing them since in fines to GF. If they hadn't wasted the $6 Billion in the last 10yrs (half on ATI write down) we wouldn't be talking about bankruptcy and AMD would still be in the CPU race and would be on AMD SOC5 like Nvidia, and about to take over just like them with a gpu from desktops and 20yrs of game dev/driver experience nobody else has. They would have a much better stance in Grid stuff too. Instead they are behind by far on everything. NV already has 600 companies testing grid, up from 200 last Q, 100 Q before (massive interest, when testing is over this will add to the bottom line). AMD went console which are dying instead of mobile etc. Xbox1 is already down to 115K a month in US (down from 311K last month) and those are not SELL THROUGH numbers just like the 5mil sold so it's even worse (also MS says supply is outstripping demand, so sitting on shelves). They sold 3.9mil Q4 at launch but only 1.2mil in twice as long (only 1/2nov+dec for 2013)? That's a tanking. PS4 already under a million and both dropping fast. At this rate sony may barely crack 10mil, and MS likely won't. 1.2M sold in Q1 is 400K world wide per month and again not SELL THROUGH. 400K x 9 months is only 3.6M which is 8.6m NON sell through, so far shy of 10mil and dropping connect won't help much for more than a month or two. AMD is low double digits for margins ($12 or less per chip on $100 chip) so these can barely pay the interest on their debt and they will be fined by GF for take or pay shortly. They delayed that 200mil payment of Dec31, or you would have not been able to claim AMD made anything. Even 15mil consoles this year total would only be 180mil profit or so for AMD. That won't help the bottom line much with fines to GF, 200mil in interest etc. Kinect won't spur many sales either as removing it kills most of what MS has marketed the device on (voice, gestures etc). Sony console sales were 16.5mil globally in 2012 BEFORE ps4 launched. Last year 14.6mil were sold total. Sony just reported a 1.26B loss for 2013 and expects another 490mil loss this year. I would not have sold my mobile gpu division to Qcom in 2009 either which became adreno we see today. They did that for $65mil. WOW. Stupid right? This crap happens when you make one stupid decision worth billions (ATI over priced acquisition) breaking you down until you are selling everything in sight at a discount (fabs, mobile division, digital TV division etc) to survive vs. competition that didn't do this crap.

Companies make mistakes. Qcom just missed the 64bit boat and had to go off the shelf. Rimm went from kings to worthless in 5yrs. People run companies, which are not always ruled by smart decisions, but often ruled by politics (internal or external) etc. You act like AMD only makes smart decisions (or any company), when the $6B lost in 10yrs says it wouldn't be too hard to beat their management in the last 10yrs. They gave away the mobile division that they are now trying to get back into...LOL. I've pretty much laid out their failure here financially and I think I've made the case for STUPID management decisions. The numbers don't lie. They have 5.3B+ revenue and lost 83mil last year while NV made 500mil on 4.1B while giving a dividend and buying back shares and growing cash and moving into new markets solidly. I fail to see a product AMD has coming that will make them a billion in profits or even half that. Where will they get the R&D to keep up with Qcom (6.5B/yr), NV (500mil), Samsung (35-40Bil), Apple (40B+), Intel (10B+) etc. You can't win when you're barely breaking even. This isn't rocket science. I'm stating the obvious, so not sure how you get to me thinking I'm Einstein. If you can read a balance sheet you know most of what I'm saying. You're already saying some of it yourself (made money barely for a Q, but NOT for the year!). Couple that info with reading tech sites and this isn't difficult stuff to predict.

Attack the data not the person. Sorry if the data doesn't fit your narrative or make you happy, but as a person who trades stocks (with my future retirement...LOL) I don't have the luxury of ignoring the data. AMD is doing better at not wasting money. But that isn't the same as saying they're rolling around in dough like their competition and I didn't even list them all. That is merely saying they're not completely failing like they were before. That isn't what I'd call success.

Now that's a real wall of text ;) But backed by data, common sense and no love for anyone but the numbers. It's not just an opinion as the data clearly shows, so good luck tearing it all down. See if you can do it without landing in the top 4 of this list:
http://www.paulgraham.com/disagree.html
IF you like charts instead that say the same stuff, stay out of the bottom 4 of this chart:
https://commons.wikimedia.org/wiki/File:Graham%27s_Hierarchy_of_Disagreement.svg
Same story but pretty with less explanation :) IF you can refute my central point (or any of it with data), I'll be happy to read it. I don't care who wins as long as I'm in the stock before they do. So bring on the data wall if you have one.
 

somebodyspecial

Honorable
Sep 20, 2012
1,459
0
11,310
20
A few more pieces of data:
http://jonpeddie.com/press-releases/details/intel-gains-nvidia-flat-and-amd-loses-graphics-market-share-in-q1/

AMD losing share Q1. Everyone is coming for low end now (Intel from top down, ARM from bottom up). AMD shipments are way down (more than seasonal). Nvidia was flat, and as you can see from the chart Intel is taking it from the low end as APU's are down. Intel used to not be interested in the low end, but then along came ARM taking 21% of notebooks. AMD is getting squeezed even as they run from high-end to low-end (gave up cpu race with Intel remember?). It will get worse as ARM moves up the cpu/gpu ladder and Intel moves down with an already great cpu and repeated improvements on gpu now (broadwell all about gpu again, small cpu increase as transistors are dedicated to gpu now).

http://www.labaton.com/en/cases/Hatamian-v-Advanced-Micro-Devices-Inc.cfm

AMD will have to settle or lose at some point probably. Pumping llano numbers, then proven lies, then stock drop of 72% etc. It's tough to get out of claiming one thing to shareholders about Llano sales, then writing off 100mil + a 72% stock price drop solely due (in AMD's own words) to Llano overstock. Either your stupid and don't know what is going on in your company, or you lied. Either way hard to get out of some sort of mismanagement suit. But maybe I'm wrong. :)

oops a few more: AMD has announced 20nm also:
http://www.enterprisetech.com/2014/05/05/amd-unify-x86-arm-systems-skybridge/

http://www.anandtech.com/show/7925/qualcomms-snapdragon-808810-20nm-highend-64bit-socs-with-lte-category-67-support-in-2015
Qcom announced 810 20nm in april but modems were announced in 2013 on 20nm.

http://vr-zone.com/articles/samsung-project-kq-leaked-galaxy-s5-qhd-screen-20-nm-exynos-5430-launch-year/76410.html?utm_source=rss&utm_medium=rss&utm_campaign=samsung-project-kq-leaked-galaxy-s5-qhd-screen-20-nm-exynos-5430-launch-year
20nm 5430 leaked all over the web in April. Due this year. This from April.

http://www.gizmochina.com/2014/05/18/meizu-mx4-prepared-exynos-5430-64-bit-cpu-same-as-samsung-galaxy-s5-prime/
Same chip going into meizu also. So samsung, Apple, Qcom, AMD. No 20nm announcements myth dead now right? Nearly everyone has announced something. As I said, even if I'm off a quarter, it changes nothing. They are coming soon.
 
Status
Not open for further replies.

ASK THE COMMUNITY

TRENDING THREADS