AMD: It Won't Be About 'AMD vs. Intel' Anymore

[formulav8]

The saddest thing is there are people here who wanted Intel to trample AMD. See how happy they are in a couple years.....

 

[Vesku]

Isn't that why some sports have revenue sharing? Fun to have your team keep winning the league but eventually their opponents are so terrible from lack of money that the whole endeavor loses it's appeal.

 

[cbn]

Back when the last time prevailed that Intel had a complete monopoly, they could get away with charging $1000 for 386s and comparable prices for chipsets. And there was massive performance headroom. Now, what everyone has is more than good enough and upgrading takes some incentive. I think Silverforce is right. There will be high-end CPUs but I think Intel will preserve the performance consumer segment, which is what most of us play in. Sure, I'll upgrade to Ivy (probably). But I don't really HAVE to. Photoshop already opens in two seconds. Would opening it in 1.5 seconds make my life better? Not a whole lot. Upgrading now is more for bragging rights or for those of us who have a really compelling application that justifies it, or because your current box is wheezing. Intel knows all this.

I agree about the analysis for top end Intel CPUs.

I think the major effect of a weak AMD is on how Intel positions its low end CPUs.

When AMD starts stumbling it forces Intel to institute artifical measures such as "no overclocking" on Core i3 in order to keep AMD SKUs propped up and somewhat healthy.

Unfortunately for AMD, this kind of thing can only go so far. If Intel's advances begin to slow down they might need to further compress the market. In fact, that is what I fear the most. Maybe a good example will be Ivy Bridge?<---This looks like it might be a particularly strong product with the large iGPU and trigate xtors. Unless Intel can make another large jump in process technology (or some other kind of tech) how are they going to convince Ivy Bridge users to upgrade? Lower Prices even more? Or will they decide to no longer support overclocking of any kind in order to throw some business AMD's way?

This is why I really like ARM for AMD. It gives the company options. Like I said earlier, AMD would not be competing with existing ARM players because (as I understand things) Notebook ARMv8s are a completely different design compared to a Smartphone ARM SOC---> http://forums.anandtech.com/showpost.php?p=32644265&postcount=22

 

[Arachnotronic]

Intel is competing with itself pretty vigorously. They need to develop more and more efficient parts to improve battery life and mobile performance. This is what drove SNB, and the fruits of that for desktop users are a highly efficient architecture that sips power and overclocks like mad.

SNB-E also reaps this by taking this efficient architecture, souping it up for workstations/servers/datacenters, and then also giving the high end desktops a few SKUs.

The minute Intel stagnates, the minute people start deciding what they have is "good" enough and stop upgrading/move to tablets and phones.

SNB has been on the market since January with no competition, and yet we still have the affordable and blazing fast 2500K. If you want to get a little more spendy, there's the $100 extra 2600K. If you're really burnin' for some enthusiast lovin', 3930K and 3960X are right there for you. What more do you people want?

 

[Vesku]

They don't need ARM for that though, they have a tablet bobcat chip. I think there are just two main reasons we don't see more x86 tablets. 1) The bobcat Z chip is about a watt or so too hot and Intel has let Atom stew a bit too long. 2) No Android x86 and Microsoft won't be ready for tablets until Windows 8 comes out.

This is why I really like ARM for AMD. It gives the company options. Like I said earlier, AMD would not be competing with existing ARM players because (as I understand things) Notebook ARMv8s are a completely different design compared to a Smartphone ARM SOC---> http://forums.anandtech.com/showpost.php?p=32644265&postcount=22

 

[cbn]

Intel is competing with itself pretty vigorously. They need to develop more and more efficient parts to improve battery life and mobile performance.

They might just be doing that!

http://semiaccurate.com/2011/11/22/intel’s-analog-shrink-is-for-real/

I would imagine having Wifi on "Tri-gate" would go a long way to reducing the power consumption of that radio during internet downloads.

 

[cbn]

They don't need ARM for that though, they have a tablet bobcat chip. I think there are just two main reasons we don't see more x86 tablets. 1) The bobcat Z chip is about a watt or so too hot and Intel has let Atom stew a bit too long. 2) No Android x86 and Microsoft won't be ready for tablets until Windows 8 comes out.

Is Bobcat really going to save AMD's product margins?

Or is the x86 mobile market in for some real price compression?

My guess is the latter. (especially with all this talk that certain big OEMs want to leave the device maker business. If the profits don't exist in the future for those guys....I think we will begin to see the cheapest x86 CPUs get the boot!)

 

[Vesku]

Don't see how they could get better margins jumping into ARM SoCs. Next year Intel and AMD x86 hardware will be tablet ready, it's about whether the x86 OSes will be there to run on them.

 

[IntelUser2000]

This is very naive thinking. Intel already did that when it was in the lead. It will start with Ivy being late.

Ivy Bridge being late has nothing to do with AMD.

I remember how people said top line Sandy Bridge(I mean the LGA1155 chips) would cost $500 range, then they released it at $300.

They may be able to stagnate in the desktop markets, but they can't stagnate as a company. They have been increasingly focusing on mobile every generation, and will continue to do so.

You guys compare how prices of regular, non EE CPUs used to cost $1000+. I've seen Dell ads of a regular desktop costing $7k! But back then, buying the $1000 CPUs gained you something tangible, unlike today where even enthusiasts buy GTX 580 + 4.5GHz 2600K systems to watch Youtube videos of random guy farting, or twiddle their thumbs wondering what to do. A decade ago, you needed the best video cards if you wanted latest games to run at only 60 fps. Nowadays, you have to TRY to reach "lowly" 60 fps, and can easily do it on the mid-range cards.

$1000 for 2600K-like CPUs won't even work on the hardcore enthusiasts anymore. They need to be price competitive.

1) The bobcat Z chip is about a watt or so too hot and Intel has let Atom stew a bit too long. 2) No Android x86 and Microsoft won't be ready for tablets until Windows 8 comes out.

Bobcat only looks good in paper. AMD is in a completely different light than merely a year ago. Now you can buy non-Black Friday deals of non Atom/Bobcat based laptops in the $350-400 range. In Tablets its 3 watts too high(5.9W). The awkward position of "not low enough power" and "not high enough performance".

 

[piesquared]

They don't need ARM for that though, they have a tablet bobcat chip. I think there are just two main reasons we don't see more x86 tablets. 1) The bobcat Z chip is about a watt or so too hot and Intel has let Atom stew a bit too long. 2) No Android x86 and Microsoft won't be ready for tablets until Windows 8 comes out.

This link was on the last page:

http://tech.icrontic.com/news/heres...u-android-4-0-1-optimized-for-amd-brazos-apu/

Here’s a tablet strategy for you: Android 4.0.1 optimized for AMD Brazos APU

Brian Ambrozy(primesuspect) A very unassuming announcement in a Google developer's group could lead to some very big implications for AMD in the tablet space.

November 30, 2011 9:06 PM ET in News, AMD, Android, APUs, smartphones, tablets





Chih-Wei Huang, from the Android-x86 project, just released an innocuous and understated message on the Google Groups page for the project. He says,


“Hi all,
The ics-x86 branch based on Android 4.0.1 is ready.”

This is the equivalent of “Hello world” for an entirely new class of possible tablet hardware. What could this bode for 2012?


Android Ice Cream Sandwich on AMD Brazos

AMD has been angling for years to break into the mobile and low-power market, and they’ve succeeded in a small way with their Fusion initiative, which is an APU (a combined CPU + GPU on a single chip) that uses far less power than a traditional desktop processor such as Intel Core or AMD FX (read more about what an APU is here). So far, the current generation of APU (the platform is codenamed “Brazos” and includes the APU, chipset, and motherboard) has found some success in the small notebook and home theater PC space. Those under $500, small laptops you’re seeing at retail outlets like Best Buy and Micro Center by HP and others are using Brazos to provide very capable little laptops that get good battery life and deliver high-def graphics inexpensively.

Those APUs, however, use the x86 instruction set. This is an instruction set that’s been around since 1978 and underlies every single desktop PC running Windows, as well as Mac OS and many flavors of Linux as well. It is, to say the least, extremely standard.

Android, by contrast, runs traditionally on ARM architecture chips, which are used in the vast majority of smartphones and tablets. That’s why you don’t see Android running on desktop PCs or laptops, and you don’t see Windows running on smartphones.

This announcement by Huang changes the entire game by introducing the latest version of Android to the instruction set that standard computers run on—and optimizing it for AMD rather than Intel. Android 4.0.1 does run on all x86 processors, but his announcement indicates that hardware acceleration is running and optimized for AMD’s Fusion APUs, rather than ‘we’ve been working closely with Google‘ Intel.

The ghost of ATI strikes back

The key for AMD here is that their purchase of GPU company ATI in 2006 continues to pay off dividends; now Android 4.0.1 is hardware accelerated on RADEON GPUs, which makes it a no-brainer for developers to use as a reference platform for early apps and hardware. Why optimize for Intel when Intel doesn’t have an equivalent GPU/CPU hybrid, especially when the support is built in to 4.0.1 right now?

Former AMD CEO Dirk Meyer was allegedly sacked for not steering the company to a tablet and smartphone strategy. While this doesn’t necessarily mean a marketing strategy or victory on AMD’s part, Dirk’s probably laughing right now, because it just fell into their lap: Windows 8, and now Android 4, are ready to run on hardware AMD can ship in volume today.

AMD is bungling a lot of things lately, but their meal ticket may very well be their extremely strong GPU technology and its related elements, e.g. Fusion. This sort of happy accident doesn’t happen often, and it’s very likely that this wasn’t through any effort of AMD engineering or marketing. More likely is the scenario in which developers want to write for the best platform, and for now, Brazos is looking like it.

 

[Vesku]

Ivy Bridge being late has nothing to do with AMD.

I remember how people said top line Sandy Bridge(I mean the LGA1155 chips) would cost $500 range, then they released it at $300.

They may be able to stagnate in the desktop markets, but they can't stagnate as a company. They have been increasingly focusing on mobile every generation, and will continue to do so.

You guys compare how prices of regular, non EE CPUs used to cost $1000+. I've seen Dell ads of a regular desktop costing $7k! But back then, buying the $1000 CPUs gained you something tangible, unlike today where even enthusiasts buy GTX 580 + 4.5GHz 2600K systems to watch Youtube videos of random guy farting, or twiddle their thumbs wondering what to do. A decade ago, you needed the best video cards if you wanted latest games to run at only 60 fps. Nowadays, you have to TRY to reach "lowly" 60 fps, and can easily do it on the mid-range cards.

$1000 for 2600K-like CPUs won't even work on the hardcore enthusiasts anymore. They need to be price competitive.



Bobcat only looks good in paper. AMD is in a completely different light than merely a year ago. Now you can buy non-Black Friday deals of non Atom/Bobcat based laptops in the $350-400 range. In Tablets its 3 watts too high(5.9W). The awkward position of "not low enough power" and "not high enough performance".

That's at 40nm, 28nm for AMD and 32/22nm for Intel will see some good power gains. If you don't think the C-50 is high enough performance then why are you so hyped on ARM? AMD and Intel would do well for themselves to not let 28nm ARM get too much time in the limelight.

 

[Blitzvogel]

This link was on the last page:

http://tech.icrontic.com/news/heres...u-android-4-0-1-optimized-for-amd-brazos-apu/

That's great news for AMD, especially when things are looking so dark, but also great for Google, who can further extend their reach into more more devices, and even better yet, into devices that are much more powerful. Who might be fearing such a ruse is MS, who wants to probably maintain their "dominance" (if you can call it that) of the x86 tablet market. What kind of code and app portability would a user expect though with just a platform switch of the OS? I'm sure plenty of apps would have to be rewritten entirely in order to run with x86 unless there is a bloated, power sucking abstraction layer to convert code on the fly. That is why I would rather have a Windows [8] x86 Tablet in pair with my Windows computer. I can natively use anything I want on either machine without any bull****. Obviously, Android doesn't share any ubiquity on desktops but this might be the big break that the OS needs in order to become a real "full featured" OS that can compete with the likes of Windows, OS X, and other various flavors of Linux on x86 machines.

 

[IntelUser2000]

If you don't think the C-50 is high enough performance then why are you so hyped on ARM?

I'm not hyped on ARM. I think you are confusing me with someone else. See, ARM's low performance is justified because the power consumption range is low enough to be perfect for Smartphones and Tablets. Llano and Sandy Bridge is fine because they have the performance(whether CPU or iGPU). Bobcat and Atom are both in a position where it fits neither, though the latter is closer to achieving low enough power.

According to the roadmap, the Tablet optimized Bobcat is coming in mid-2012 and its called Brazos-T: http://www.xbitlabs.com/news/cpu/di..._Tablet_Chips_to_Consume_2W_or_4W_Slides.html

That's on 40nm, meaning its somewhere in 2013 for 28nm.

 

[Silverforce11]

I foresee future smartphones running windows 8, with mini hdmi, several mini usb ports. Users can take it on the go, or when its at work or "docked" at home, plugged into it is a full size keyboard, mouse, and hdmi out to a big monitor or LED tv, running all the windows software thats compatible for work and play.

When will it happen? When bobcat gets to stretch its legs on TSMC or GF 20nm tech (i think they are skipping 22nm) and windows 8 is available. A E-450++ sipping 2W packs serious power compared to ARM devices, with x86 support thats required for so many software with a huge user base.

Heck if i had a smartphone like that which runs windows, i dont need notebooks anymore.

 

[IntelUser2000]

Pricing comparisons decade ago and today:

10 years ago:
-Pentium 4 chips cost $1500 initially along with the two 64MB RDRAM bundle. That was November 2000.
-However, pricing came down super quick. Back in those days, you knew the meaning of "obsolescence". By Spring of 2001, 1.7GHz chips cost less than $400. That's what obsolete really means.
-The strategy of bringing a $1000+ CPU then cutting the price substantially when the new higher bins arrived were common until about the Core 2 days. Leading to us enthusiasts still thinking today that new CPUs mean older ones became more affordable
-The Dell $7k desktop Ad. Pentium III days.

Now:
-Not even the specially marked "Extreme Edition" chips cost $1500. Pretty close, but even then only one SKU
-Pricing never changes. You could buy a 2600K in January for about the same $300 price range you buy now
-$7k desktop? Any company doing that would be a laughing stock of everyone. Even your mom and dad knows computers are much cheaper. Now I only see workstations above $5k.
-Amendment to the above: Pricing does change, but only for competitive reasons. You see some unknown i3-2123 chip going down in price, and even then by $10-30.

 

[aigomorla]

Uh huh. Would you like me to list the half truths and failures of intel over the years? Believe me i'd love to do it...

You may be right, it might have been 90&#37; expected. So clearly they have a strategy.

im sorry pie are you trying to support AMD here?

They just public announced that they are shooting themselves in the foot, and lost to intel, and your still supporting them in this thread?

Lets just ignore fan base for a second... AMD was ultimately there so they could regulate intel. Its simple and plan fact.. if AMD was a threat.. intel would step up.

We dont have any form of competition for intel right now in the main sector were talking about...

So can u honestly support AMD in this sector when they no longer have a sector share? im lost in your post...


Yes intel has been shaddy, but have they ever lied to us in regards to a product? Inflated it to a point where comparisons were pointless and at best colorful but held no real information.

No ALL of intel's products so far has been solid... do you see a 2600k owner who is sad? do you see a gulftown user who is sad? An i7 owner who regrets ever buying a platform? No..

I can count more then the digits on my hands and feet on the people who are utterly disappointed with bulldozer and AMD's marketing practice. :ninja:


And beyond that do you honestly think AMD has a shot against VIA and ARM? lets be serious here...
ARM market is a lot harder then the consumer desktop market.
And without the FAB's needed to separate them from VIA and ARM, they are just asking to be dropped down to the picking chain like Texas Instruments and IBM did with cyrix.

I dont want to see this happening, but some genius in AMD thinks its a better route..

In short, i dont hold any expectations of AMD lately... every news i read about them is in regards to a sinking boat...

 

[piesquared]

im sorry pie are you trying to support AMD here?

They just public announced that they are shooting themselves in the foot, and lost to intel, and your still supporting them in this thread?

Lets just ignore fan base for a second... AMD was ultimately there so they could regulate intel. Its simple and plan fact.. if AMD was a threat.. intel would step up.

We dont have any form of competition for intel right now in the main sector were talking about...

So can u honestly support AMD in this sector when they no longer have a sector share? im lost in your post...


Yes intel has been shaddy, but have they ever lied to us in regards to a product? Inflated it to a point where comparisons were pointless and at best colorful but held no real information.

No ALL of intel's products so far has been solid... do you see a 2600k owner who is sad? do you see a gulftown user who is sad? An i7 owner who regrets ever buying a platform? No..

I can count more then the digits on my hands and feet on the people who are utterly disappointed with bulldozer and AMD's marketing practice. :ninja:


And beyond that do you honestly think AMD has a shot against VIA and ARM? lets be serious here...
ARM market is a lot harder then the consumer desktop market.
And without the FAB's needed to separate them from VIA and ARM, they are just asking to be dropped down to the picking chain like Texas Instruments and IBM did with cyrix.

I dont want to see this happening, but some genius in AMD thinks its a better route..

In short, i dont hold any expectations of AMD lately... every news i read about them is in regards to a sinking boat...

What a load [of well written material]

No profanity means no profanity, guys
-ViRGE

 

[piesquared]

More here:

The independent open source effort has ported the source code of Android 4.0.1 to work with tablets based on Advanced Micro Devices' low-power x86 chips code-named Brazos, which are typically used in netbooks and low-end laptops. Some AMD chips are being used in tablets such as MSI's WindPad 110W. The port means that tablets with Android 4.0 based on x86 chips could be on the horizon, but using AMD chips rather than Intel's

http://www.infoworld.com/d/mobile-t...sors-180605?mm_ref=http&#37;3A//t.co/hf3AmcO6

 

[exar333]

The saddest thing is there are people here who wanted Intel to trample AMD. See how happy they are in a couple years.....

I don't want them to fail, but why support a company that outright lies about their products, performance, and future compatibility? AMD surpassed NV this year with their 'douche-baggery' tactics.

AMD PR = Secrets, red herrings, and lies

:thumbsdown:

Edit: Strangely enough the graphics division is exactly opposite this, although they do whine a lot about NV activities instead of just being more pro-active

 

[ocre]

This is why I really like ARM for AMD. It gives the company options. Like I said earlier, AMD would not be competing with existing ARM players because (as I understand things) Notebook ARMv8s are a completely different design compared to a Smartphone ARM SOC---> http://forums.anandtech.com/showpost.php?p=32644265&postcount=22

I am telling you, AMd is not going into ARM the way you think. They value their x86 ability, it is actually something no other company can get. Their best plan at this point is to try to find a path with x86 and hope like hell intel can clear ARM out of the way. For many reasons. One being if AMD goes into ARM, they will be starting all over. The company will have to shed most of its employees if its able to stay above water at all. AMD is big, a whole lot bigger than Nvidia. Nvidia is just now gaining some in ARM and they have been trying for a good long while. AMD cannot afford to go into ARM without a huge slashing and hacking. Only a tiny piece of the company would remain. They will keep most of their chips in x86 and pray for a miracle. The short term out look for AMD is bad, bad or worse. ARM will not save them, it would be the very last resort like floating away on a raft after the titanic sinks. The only reason for them to even work on an ARM chip is for that raft. For now they must act bold and speak very high about x86. They must publicly be 100% behind x86 and not show any doubt in this architecture. This is the only route the have,the only one that keeps most of the employes in jobs as long as possible.

Anyway, i found more information, which is from AMDs Mike Silverman, the same guy who said AMD is no longer competing with intel. After all the reports twisted up his statement he wanted to clear up the confusion. This is what he has to say:

"AMD is a leader in x86 microprocessor design, and we remain committed to the x86 market,” Silverman told us. “Our strategy is to accelerate our growth by taking advantage of our design capabilities to deliver a breadth of products that best align with broader industry shifts toward low power, emerging markets and the cloud."

The articlle goes on to explain pretty muc exactly what i have been saying. AMD going to arm would be disastrous at this point and they have to put their cips in x86 right now, no matter if they fail or not.

The article says, "AMD has neither the time nor the cash flow to commit to an equivalent long-term ramp, particularly when it has proven technology already on the table."

its a pretty good read, and really based on reality.

http://www.extremetech.com/mobile/107133-amd-to-attempt-self-resurrection-but-not-with-arm

For whatever happens to AMD, they are stuck in x86 for some time. ARM can only be a life boat if the company tanks, and i doubt they would even try then.

 

[piesquared]

I am telling you, AMd is not going into ARM the way you think. They value their x86 ability, it is actually something no other company can get. Their best plan at this point is to try to find a path with x86 and hope like hell intel can clear ARM out of the way. For many reasons. One being if AMD goes into ARM, they will be starting all over. The company will have to shed most of its employees if its able to stay above water at all. AMD is big, a whole lot bigger than Nvidia. Nvidia is just now gaining some in ARM and they have been trying for a good long while. AMD cannot afford to go into ARM without a huge slashing and hacking. Only a tiny piece of the company would remain. They will keep most of their chips in x86 and pray for a miracle. The short term out look for AMD is bad, bad or worse. ARM will not save them, it would be the very last resort like floating away on a raft after the titanic sinks. The only reason for them to even work on an ARM chip is for that raft. For now they must act bold and speak very high about x86. They must publicly be 100% behind x86 and not show any doubt in this architecture. This is the only route the have,the only one that keeps most of the employes in jobs as long as possible.

Anyway, i found more information, which is from AMDs Mike Silverman, the same guy who said AMD is no longer competing with intel. After all the reports twisted up his statement he wanted to clear up the confusion. This is what he has to say:

"AMD is a leader in x86 microprocessor design, and we remain committed to the x86 market,” Silverman told us. “Our strategy is to accelerate our growth by taking advantage of our design capabilities to deliver a breadth of products that best align with broader industry shifts toward low power, emerging markets and the cloud."

The articlle goes on to explain pretty muc exactly what i have been saying. AMD going to arm would be disastrous at this point and they have to put their cips in x86 right now, no matter if they fail or not.

The article says, "AMD has neither the time nor the cash flow to commit to an equivalent long-term ramp, particularly when it has proven technology already on the table."

its a pretty good read, and really based on reality.

http://www.extremetech.com/mobile/107133-amd-to-attempt-self-resurrection-but-not-with-arm

For whatever happens to AMD, they are stuck in x86 for some time. ARM can only be a life boat if the company tanks, and i doubt they would even try then.

You are forgetting Open CL and GPGPU/ heterogenous computing for which both ARM and AMD agree is where the future is, and AMD leads in this space.

 

[exar333]

You are forgetting Open CL and GPGPU/ heterogenous computing for which both ARM and AMD agree is where the future is, and AMD leads in this space.

Your blind faith in AMD is dizzying. How exactly do they lead in GPGPU? NV is much more involved and is currently making money from GPGPU operations. AMD hardly supports it. More misinformation from you, after reading marketing slides.

 

[piesquared]

Your blind faith in AMD is dizzying. How exactly do they lead in GPGPU? NV is much more involved and is currently making money from GPGPU operations. AMD hardly supports it. More misinformation from you, after reading marketing slides.

uhhhh because the post is in the context of AMD and intel?

 

[cbn]

its a pretty good read, and really based on reality.

http://www.extremetech.com/mobile/107133-amd-to-attempt-self-resurrection-but-not-with-arm

Furthermore, the numbers don&#8217;t add up. In late October, Nvidia CEO Jen-Hsun Huang told the AsiaD conference that his company had invested $2 billion into Tegra in the past five years. Nvidia&#8217;s most recent quarterly filings reveal that its consumer products business (CPB) earned $481.4 million in the first nine months of 2011. Tegra 2 is doing extremely well &#8212; but it hasn&#8217;t yet made back its startup costs. AMD has neither the time nor the cash flow to commit to an equivalent long-term ramp, particularly when it has proven technology already on the table.

Why does this article even bring up Tegra? <----That is a smartphone ARM SOC!

How does "Smartphone SOC" have anything to do with "ARMv8 Notebook chip"? As far as I understand these two designs are on a completely different level.

 

[exar333]

uhhhh because the post is in the context of AMD and intel?

More lack of substance responses.

Can you say anything besides 'AMD is the bomb yo!' :whiste: