| Can Console Gaming Save AMD From Collapse? |
| Articles - Opinion & Editorials | ||
| Written by Olin Coles | ||
| Tuesday, 09 April 2013 | ||
Can Console Gaming Save AMD From Collapse?The future of Advanced Micro Devices crucially depends on development partnerships for next-generation Sony PlayStation and Microsoft XBOX console systemsOne year ago Advanced Micro Devices, Inc. (NYSE:AMD) company stock was publically trading for $8.20 per share (16 March 2012), the highest point since trading at the same level a year prior (March 2010). Since then, AMD stock has sunk to as low as $1.86 per share, performing far worse than other stocks within the same industry. Analyzing the graph charted below, you can see that the Dow Jones Industrial averaged a 12% gain over that same one-year period from late March 2012-2013. Microprocessors company ARM ($18.6B current market capitalization) enjoyed a 51% growth over Dow Jones for this period, while NVIDIA ($7.8B) and Intel ($107.6B) slowly tapered off 16% and 22% respectively. AMD ($1.9B) is the standout here, dropping $4.53 Billion dollars in market capitalization value to sustain 65% under-performance below the Dow Jones. While the numbers don't lie, they don't always tell the entire story.
There are parables that illustrate how one thing can relate to another, while remaining completely unrelated. For example, I've been a fan of the Oakland Athletics (A's) baseball team since I first began following the sport as a child in 1986. They already had Mark McGwire and Jose Canseco on the team back then, joining existing legends Dave Henderson, Carney Lansford, Mickey Tettleton and even Reggie Jackson as the DH, plus the A's had just picked up pitchers Dave Stewart and Dennis Eckersley. It was a good year for the A's, later adding greats such as Walt Weiss and returning the always-amazing Ricky Henderson back to the Athletics. I loved that era in baseball, and still follow the A's, but the team hasn't been the same in almost two decades. My affection towards the Oakland A's is a fitting parable to my relationship with AMD. When I opened my business back in 2000, we built and upgraded computer systems using only AMD processors. In those early days, whenever performance wasn't the sole focus we used AMD K6-III CPU's to build price-competitive desktop computers. Then the AMD 'Thunderbird' series desktop processors arrived on scene, and easily outperformed the rival competition: Intel's Pentium-III CPU. My shop became an AMD Authorized Reseller and used that leverage to become a listed Microsoft System Builder. For the next few years I was a small business that opened each morning to build high-performance AMD computer systems. Then the Intel Pentium 4 happened, and customer demand forced my tastes to change. I've given these examples because both are near and dear to me. Each example event shared a moment of greatness, followed by a long period of struggle. It's important to get past emotional attachment, and recognize that AMD has not been doing well as a business. Regardless of how much some might disagree, and it's a certainty that fan boys will disagree in the loudest manner, AMD is a battered ship sailing in very troubled waters with (yet another) storm approaching. Intel has maintained a firm lead ahead of AMD, who has remained the distant #2 worldwide supplier of x86 microprocessors. AMD purchased graphics giant Advanced Technology Investment (ATI) in 2006, thereby buying for itself a front-row ticket into the discrete graphics market. So now, AMD also maintains an outlying second place to NVIDIA in discrete graphics sales.
AMD put itself into the game with the Radeon HD series in 2007's popularity explosion of PCI-Express, quickly securing an early head-start over NVIDIA and Intel by adopting DirectX 10 ahead of others. This market, like the others, would see many ups and downs as it suffered a global recession and was later stymied by a shrinking desktop user base. AMD debuted their Northern Islands GPU architecture in 2010, only to face disappointing worldwide sales that forced them to merely reshape and tweak their existing GPU design for a 2012 launch of Southern Islands. Now that the discreet graphics market has settled towards catering to a much smaller niche, one that produces less total revenue, fewer funds are generated for research and development. All of these things have forced AMD into the position of either perpetually trailing the industry leaders with value-priced products offering similar performance, or boldly making innovative leaps to provide the industry with fresh technology. AMD has long-resembled the former, and in 2008 choose to become the latter when they created GlobalFoundries Inc. Unfortunately, the timing for this venture forced a 10% reduction to AMD's entire global workforce of approximately 10,000 employees. A few years later another 10% reduction occurred late in 2011, followed by a 15% workforce release announced at the end of 2012. Sailing into the rough seas of 2013, this ship had taken on some water while losing 1/3 of its crew. But just when it looked like the Captain was going down with the ship, two unexpected companies tossed AMD a line: Sony and Microsoft. Not so long ago I wrote XBOX720 and PS4 Consoles To Revive PC Video Games, an article that asserted how game developers would see new demand for their products as the platform technology improved. The obvious winners were those invested in game console production and development, but it was too early to point the finger at many specific names. Sony's PlayStation and Microsoft's XBOX platform were expected to thrive well into the next generation, as would the developers and publishers working to launch fresh game titles with them, but hardware component suppliers were still an unknown... at least until now. Unofficially announced, both Sony and Microsoft have selected Advanced Micro Devices as their source for next-generation console processor needs. Details are still slowly rising to the surface, but now the question in desperate need of answering is: How much will this help to save a failing company? Enough, I hope. At least enough for AMD to enjoy smooth sailing for a little while longer, because there's plenty of room for competition. COMMENT QUESTION: Between Intel, NVIDIA, and AMD, which company do you most want to succeed and why?
Related Articles:
|
||
Comments
If you understand what AMD is achieving in the APU = CPU + GPU integration area, you will understand the magnitude of AMD's vision, which is brilliantly being executed by Rory Read. No other company is integrating the CPU and the GPU in the manner that AMD is doing it. AMD will have a native fully integrated CPU/GPU where there is no difference between CPU's and GPU's but are all fully integrated.
The latter is work in progress, and the final implementations will be released in the 2014/2015 time frames. Intel and other companies does not have anything similar to it, not even in future planning. Intel does not have even a high end GPU, not even a medium end GPU.
AMD will be in the 64 bit ARM and x86 processor race. Intel will not be in the ARM race and NVIDIA will only be in the ARM side. AMD has the right profit margin models to make money in both the ARM and x86 sides, compared to Intel or NVIDIA.
AMD just got all the major video game makers consoles (PS4, XBOX, and WII) as their only graphic chip and software infrastructure provider. That is huge..!!
AMD will be releasing 64 bit x86 and ARM tablets with high end graphics capabilities in the very near future. Intel only has mediocre tablet chips and does not do any ARM chips, so Intel is handicap and NVIDIA still has to deliver and develop 64 bit CPU technology, which AMD is already is king.
AMD may have unfairly suffered under Intel's practices a decade ago, but they would have still fallen behind by the time Intel's Core architecture debuted.
Intel's server business is not that much affected by ARM because they are still 32 bit processors. Next year ARM will release their 64 bit chips: NVIDIA, AMD, QUALCOM, etc. and will be entering into the server business big time. Intel can't compete with it's current chips, and TSMC and GlobalFoundries will be releasing soon their 3d transistor based chips and 18-22 nm chips.
AMD bought SeaMicro that uses Intel and AMD chips, and will soon support ARM chips, too. SeaMicro is doing great and with AMD new ARM/x86 chips and new technology, it will get part of Intel's server market back since Intel does not have ARM solutions which are better and more economical than Intel's x86. Costumers want ARM server chips because they don't want to pay overpriced values to Intel. Once the ARM 64 bit server chips are available, Intel will lose a huge amount of profit/cash and Intel can't keep building Foundries because it takes billions in facilities and research to come up with those facilities. The latter means less profit and even loses for Intel.. Intel marketing engine is huge, and they will love you to believe that they will be doing fine over the next years, but currently, they don't have what is needed to fight in the low profit margin future world.
They are being built for performance and for servers too. They don't have the flows and baggage that x86 have.
Apple will not need Intel in their laptop/desktops/computers and Samsung, TMSC, GlobalFoundries will have 22 nm or better 3d transistor chips in around a year or less.
Even with 32 bit chips, ARM has wiped the 64 bit cpu x86 PC market. Now imagine when the 64 bit ARM chips arrive..
People love ARM because it had the right price/performance ratio. Intel business profit model can't survive selling at low price/performance ratios.. Why do you think Ottellini is jumping ship.. He knows what is coming and he doesn't want to be here when that happens...
What? How can you overtake something in a market you're not seriously competing in?
ARM will only be competitive if they can scale well power-wise. Considering how competitive Intel's chips are at similar low TDPs (i. e.
When Apple, NVIDIA, AMD, Qualcom, Samsung, Texas Instruments, and other new and old CPU designing companies release their newly designed 64 bit ARM processors, with high performance GPUs built into them, as well as discrete ones, they will displace Intels markets in the higher performance laptops, workstations, servers, computing, and other 64 bit higher performance markets.
Even in the top 100 supercomputers, it was the use of NVIDIA and AMD's GPU/Tesla/CUDA/etc. computing/processing units what makes the systems the fastest computing devices in the world since all the graphics are vector/scalar and highly parallel. CPU's are not highly parallel devices and don't handle vector processing, which is the basis for all 3d graphics, 2d graphics, and computational processing. The graphics processing will explode in the next couple of years with games, fancy 3d graphics OS's, data processing (voice, tracking technology, gestures, etc) so GPU/APU processing will become key to success. Discrete cards will not be feasible for smaller devices, so Intel still have to prove a lot in the next couple of years since it lacks high performance GPU's and all the satellite dispaly/graphics/processing technology.
Which comes to another key point. AMD and NVIDIA are creating fusion/hybrid processors, where the CPU and GPU are one and the same, and there is no difference on how they process vector or non-vector data, which means graphical data are going to be screaming on those hibrid processors. Intel does not have any hybrid processors, and it will take it years to come with one, even more, not having high performance GPU technology in house.
Why are hybrid processors important? because those processors are going to be able to run with GDDR5 memory, which is fast memory used by the GPU cards, but you need a hybrid design to exploit the benefits. They are a lot lower cost and a lot faster than the older CPU/GPU technology. AMD will have those processors this or next year, and most likely in their x86 and ARM systems, while NVIDIA will most likely have them in their 64 bit ARM processors.
And heterogeneous processors have only been used for low-to-mid-power parts so far.
And ARM taking over Intel? Not as long as most of the major software exists in x86 form.
AMD's Rory Read and all the new managers and new technical gurus are making an incredible job in concentrating AMD resources in the key areas that will enable them to grow in the future in a huge way.
AMD is going to be a formidable company with huge future wins. Being ARM and x86 and being the first ones to implement a pure hybrid APU (CPU+GPU) is a perfect winning combination. In this case, past performance is not an indication of future performance for AMD and for those that understand the future business AMD has in front of them
I know this rant makes me sound like an Intel lunatic. I promise I am not, I merely follow the performance leaders. And in that sense I, more than ANYTHING, want AMD to get the performance title back, because nothing is going to move us forward better than healthy competition. Life will suck if Intel becomes our only CPU manufacturer. But raving like a lunatic that AMD is right around the corner of success is painfully detached. They are hurting right now, and it doesn't look good.
Sony and Microsoft more than likely gave AMD the business because AMD was willing to sell much lower than Intel because Intel frankly doesn't need it. Here's to hoping this lifeline will give AMD R&D money and get back into the game.
They don't understand, for example, that without NVIDIA, Intel high end profitable workstations would be worthless, in performance, since Intel does not have any higher end GPU's and workstations need GPU performance for graphics and and to accelerate computations.
Intel is a great company with huge money reserves and technology, Intel have been accumulating it's money by deeply gouging on all the people and companies who bought and currently buy x86 CPU's, which is why ARM usage and market exploded in the last couple of years. Not to mention that Intel has never been a nice partner to work with. Intel always have backstabbed it's partners and threatened companies selling it's products if they did not accept Intel competition/politics terms.
Why is Intel going to be losing business? One single answer that makes the whole difference. Because Intel does not have the right low-profit product margin, and it can't have it. Why? Because if it changes to a low-profit model, it will immediately begin to lose all it's money reserves, it will have to fire a lot of employees and close a lot of facilities, it will have to stop building new foundries and stop researching new chip processes, it will have to be nice with other companies and partners, etc..
Intel is currently trying to conquer ARM advances by massively spending billions of dollars in trying to close the gap with ARM current and future technology. Even if it closes the technology gap, it will not be able to close the profit-margin gap, which means checkmate if Intel does not modify it's business profit model, which would mean Intel would go through a company crisis. Intel's competition? APPLE, NVIDIA, AMD, QUALCOM, Samsung, Texas Instruments, etc.. All of these companies will be hacking at the remaining of Intels computer market.
Intel knows it and is trying to find new types of business, but that takes years and if it does not work, then it will be a harder fall.
And that "temporary" lead is the type many companies would die for.
And it's not like all AMD chips (i. e. workstations) are APUs. If Intel workstations are worthless without GPUs, so are those with AMD.
You're also extrapolating some of their anticompetitive practices.
With that in mind I see a couple glaring technical issues. First is that AMD must catch Intel in IPC. This might not seem like a big deal to proponents of the APU. This is true to some extent. Sure, the GPU can help make up for that but we are too far away from an APU that can house a mid-high end GPU that will attract the more serious hardware nuts. Also, Intel has the money for R&D to drop into their graphics tech so it's only a matter of time until Intel can match the APU for graphics power. Another major issue is the difference in power usage. This goes beyond the difference in process. For comparable performing processors, Intel can offer 77 watt parts but AMD is at best 95 watts and most likely 125 watts.
As for business sense, Rory Read is likely one of the few people that could turn them around. I think he has put into motion a lot of good changes and I think why people still see them as dying is because they are comparing AMD now to the AMD that was. It appears to me to be so far different from the company it was in 2005 no one would have seen it coming. I think if people look at AMD not as "little Intel" but as it's own entirely different culture and approach that just so happens to make the same product as Intel the story changes a lot. Even in this purview they aren't doing perfect but I think now is when years of just making it work are catching up with them on paper.
In the end, I think AMD's future is marginally more up in the air than the other big tech firms. The landscape is changing so much that they will all have to evolve in a big way and if AMD can match or beat the curve I think they could mostly reverse the situation they find themselves in. It will be tough but I think it's a little soon to be hearing a flatline.
As long as this keeps going on, Intel will have no real motive to crank up the clock speeds/core counts.
If you knew the bigger picture you would know that AMD is not failing and does not need saving!
Next you will be writing that Pc World in the UK is the best computer store to ever have been created..
Lets face it your articles are getting worse, you sure you dont write for The Sun?
Hey, I'm an AMD fanboy from way back. I still run my custom water-cooled, overclocked Piledriver system with AMD Black Edition memory and twin 79xx video cards. I'd love for them to produce competitive high-end CPUs.
The executives that left were part of the problem. Rory Read and the new comers are a new generation of management with actualized incredible knowledge and vision on products and direction (SOC's, etc).
So executives leaving is a no brainer, sales failing (we are in a recession and soon Intel is going to continue to slide down faster and with bigger loses), stock price is a no brainer either.
What matters is the new technology that continues to come out. What matters is the new fully fused APU's coming next year, what matters is the higher performance GPU's that AMD have compared to Intel which had no high end GPU's. Intel's Larrabbee GPU effort was a total disaster that even Intel's huge marketing system could not hide under the rugs.
Intel just have a momentary foundry advantage and a very expensive higher performance CPU. That is not much and AMD knows it is closing the gaps with it's new and future APU's and leaving Intel in the dust with the new high performance discrete GPU's.
At this time, I will invest in AMD, since it already hit bottom, and not on Intel which is starting to dive slowly but surely since Intel can only lose the business that it already had. If Intel does not find new source of revenues, then Intel is done. Hopefully Intel's cable business will help him in a couple of years, if it succeeds.
Definitely, if you don't understand what Rory Read, Lisa Su, Mark Papermaster, and others are doing in the business and technology sides, then you are missing the whole present and future aspects of the company. I do agree with you that AMD past management really screw up big opportunities, but that is all in the past.
AMD,s incoming APUs using fast memory GDDR5 instead of DDRR is huge and the SeaMicro venture is doing great and it has a lot more coming that will bring a lot of profit to AMD.
Not to mention their iGPUs are making significant strides. They may not catch up anytime soon, but they are closing the gap.
Where these companies make most of their business is in business workstations, not water cooled 16 core quad SLI toy-boxes. And you know what big-business cares about? Power consumption, and price. Why? Because every 5-7 years they'll be buying 10,000 computers and they want to save money on their electric bill. Guess what, a fast quad sli graphics card, not even a value priced mid range card GPU means squat in that market.
You think people care about how many FPS they get working on an excel spreadsheet?
FACT: PC Gaming is a novelty/nitch market for chip-makers, not what they gauge the value of their company on. The best thing they really get out of it is advertising/brand recognition.
CLIFFS: High end gaming stations don't make these companies, low power well priced products do.
Think about it.
Just do the math. 10k computers with a 90 watt APU at 100$ each or 10k computers with a 200$ 75 watt cpu and a 20$ onboard 50 watt GPU, times 10 years of power consumption.
Hmm i wonder what chip-maker is going to be filling up all those skyscrapers on Wall-Street.
Intel's top mainstream chips run between $200-300 and smoke AMD's chips in most tasks.
You must be thinking about the -Extreme end, which goes for $600 and up.
/images/articles/editorial/NYSE-Chart.png
All of them. Imagine a world without AMD.
Or imagine without Nvida, Intel!
No competition = we are doomed!
I wonder if they also think the most successful auto-maker is the one that can build the fastest sports-car.
What about the kiddies who say that he who has the strongest iGPU wins?
It's single-threaded performance. Most everyday apps rely on it, silly. AMD is lagging behind, big time.