New AMD CEO readies strategy shift

[cbn]

I personally would rather have x86-compatible CPU in a tablet.

I am really skeptical of x86 Tablet for at least three reasons:

1. Most x86 legacy apps require keyboard and mouse/trackpad.

2. Tablet seems like a stopgap measure until smartphones work seemlessly "docked" to devices with larger screens.

3. Up and coming Lower power displays (Qualcomm Mirasol, etc) will amplify any power differences between x86 and ARM.

With those points being made, It would seem to me it is smartphone SOC for x86 or bust! (With the major selling point being able to run fullsize Window apps from the a docked device the size of palm of your hand)

Stopping at "Tablet" with development is a deadend, besides who is going to want to carry around a large Tablet and then dock it to keyboard and mouse/trackpad so they can run fullsize Windows apps? Most people will use a smartphone for this instead.

 

[podspi]

The most interesting aspect of this to me is that AMD is in a relatively unique position to bolt an x86 interpreter on an ARM based CPU, are they not? As I recall, Intel bolted one on to EPIC/Itanium.

It shouldn't be that foreign of a concept, given that their CPUs are basically RISC (mircro-ops FTW) machines behind a x86 interpreter already, right?

Wouldn't a chip that could run both ARM and x86 applications be very desirable in the market place?

Damn the performance and all that, it would be a great marketing point I would think.

This idea keeps rattling around my head, and it just sounds great. Imagine a dockable tablet that runs on ARM on battery, but is compatible with all your legacy apps at reasonable (Brazos) speeds when docked? Does this not sound awesome to anybody else?

The biggest problem is that software support for actual heterogenous computing doesn't really exist, and AMD's track record on software magic is bad.

 

[quest55720]

Stopping at "Tablet" with development is a deadend, besides who is going to want to carry around a large Tablet and then dock it to keyboard and mouse/trackpad so they can run fullsize Windows apps? Most people will use a smartphone for this instead.

I am drooling at the promise of windows 8 86 for tablets. They will be so freaking awesome for work. If MS does it right with windows 8 most apps should be fine with a tablet. The problem with tablets are they are just blown up phones. They are just toys that don't run real programs. Give me a 28nm Brazos windows 8 86 tablet with a high efficiency screen any day of the week over an Ipad or android tablet. I have read enough about people using the developer version of windows 8 on lenovo x220s. That the minute that windows 8 beta comes out I will request a new lenovo tablet to put windows 8 on it.

 

[blckgrffn]

This idea keeps rattling around my head, and it just sounds great. Imagine a dockable tablet that runs on ARM on battery, but is compatible with all your legacy apps at reasonable (Brazos) speeds when docked? Does this not sound awesome to anybody else?

The biggest problem is that software support for actual heterogenous computing doesn't really exist, and AMD's track record on software magic is bad.

You'd think that an OS would be able to sandbox those applications somehow - that way we can rely on the brilliant minds at Microsoft, Google, Apple, OSS-at-large to tackle the problem rather than AMD

Also, not sure if I really want to see this - but AMD in the black is something I wouldn't mind seeing and I don't even own any AMD stock (Unless, of course, it is sprinkled in somewhere in my 401k... in which case I think I would need to choose some different funds...)

 

[cbn]

I am drooling at the promise of windows 8 86 for tablets. They will be so freaking awesome for work. If MS does it right with windows 8 most apps should be fine with a tablet. The problem with tablets are they are just blown up phones. They are just toys that don't run real programs. Give me a 28nm Brazos windows 8 86 tablet with a high efficiency screen any day of the week over an Ipad or android tablet. I have read enough about people using the developer version of windows 8 on lenovo x220s. That the minute that windows 8 beta comes out I will request a new lenovo tablet to put windows 8 on it.

I've read that x86 apps would need to be re-written to some degree in order to work with the Touchscreen interface.

Surely this will happen with MS Office, but what about the apps beyond this?

Which begs the question:

Why choose x86 for Tablet anyway? ARM will run all the native Tablet Touchscreen apps just fine and MS will also provide Office for that ISA as well.

ARM Windows 8 Tablet vs x86 Windows 8 Tablet? Both will run MS Office, but ARM will be able to do so much more!

 

[Idontcare]

I've read that x86 apps would need to be re-written to some degree in order to work with the Touchscreen interface.

Surely this will happen with MS Office, but what about the apps beyond this?

Which begs the question:

Why choose x86 for Tablet anyway? ARM will run all the native Tablet Touchscreen apps just fine and MS will also provide Office for that ISA as well.

ARM Windows 8 Tablet vs x86 Windows 8 Tablet? Both will run MS Office, but ARM will be able to do so much more!

You make a great point. Looking at what my wife does with her iphone, the access to apps she has (and such affordable apps too! $3 and you have something you can use, MSOffice 2010 cost me $300 )...it is very obvious to me that the x86 pre-existing software advantage is a small one when it comes to the consumer space.

Obviously the server and HPC markets are different, x86 apps in those markets aren't about to be tossed aside and replaced with ARM versions for $3 a pop. But in the consumer space Apple has shown just how fickle the consumer market is to the legacy of apps or hardware.

 

[Munky]

I am really skeptical of x86 Tablet for at least three reasons:

1. Most x86 legacy apps require keyboard and mouse/trackpad.

2. Tablet seems like a stopgap measure until smartphones work seemlessly "docked" to devices with larger screens.

3. Up and coming Lower power displays (Qualcomm Mirasol, etc) will amplify any power differences between x86 and ARM.

With those points being made, It would seem to me it is smartphone SOC for x86 or bust! (With the major selling point being able to run fullsize Window apps from the a docked device the size of palm of your hand)

Stopping at "Tablet" with development is a deadend, besides who is going to want to carry around a large Tablet and then dock it to keyboard and mouse/trackpad so they can run fullsize Windows apps? Most people will use a smartphone for this instead.

Apps require keyboard input and a pointer device input. As long as the OS can interpret touchscreen commands to mimic a keyboard/mouse, the apps wouldn't care.

We may see x86 on a phone eventually, but then you're even more limited by size and battery. I'm more interested in having that capability in a tablet for now.

 

[cbn]

Apps require keyboard input and a pointer device input. As long as the OS can interpret touchscreen commands to mimic a keyboard/mouse, the apps wouldn't care.

1. Will OSes realistically do what you are talking about anytime soon?

2. What about Ergonomics of Touchscreen Tablet (somehow able to mimic keyboard and a mouse with two clickable buttons) vs an actual physical Keyboard & mouse/trackpad?

 

[Munky]

1. Will OSes realistically do what you are talking about anytime soon?

2. What about Ergonomics of Touchscreen Tablet (somehow able to mimic keyboard and a mouse with two clickable buttons) vs an actual physical Keyboard & mouse/trackpad?

There will obviously be ergonomic compromises, but a tablet itself is an ergonomic compromise. No one in their right mind would chose a touch screen virtual keyboard to do any serious lengthy typing. However, even something simple as a small optical track pad and a few buttons would make things easier, no reason why it has to be a purely touch-screen interface.

 

[iCyborg]

I've read that x86 apps would need to be re-written to some degree in order to work with the Touchscreen interface.

Surely this will happen with MS Office, but what about the apps beyond this?

Which begs the question:

Why choose x86 for Tablet anyway? ARM will run all the native Tablet Touchscreen apps just fine and MS will also provide Office for that ISA as well.

ARM Windows 8 Tablet vs x86 Windows 8 Tablet? Both will run MS Office, but ARM will be able to do so much more!

Try googling for "Windows 8 Metro UI" and read a couple articles about it.
It's pretty much exclusively developed with tablets in mind. I saw a sample tablet, it was quite nice, pretty much any MS Win app was already there and we're like a year away from launch. I don't think lack of apps will be a problem.
It looked useless for desktops though...

 

[cbn]

Try googling for "Windows 8 Metro UI" and read a couple articles about it.
It's pretty much exclusively developed with tablets in mind. I saw a sample tablet, it was quite nice, pretty much any MS Win app was already there and we're like a year away from launch. I don't think lack of apps will be a problem.
It looked useless for desktops though...

Sounds like this MS OS will do very well with ARM SOCs then.

What Intel needs to do is get on MS about making a consolidated Operating system capable of running a Silvermont x86 smartphone in desktop mode. Otherwise, why even bother with x86 for Touchscreen?

Of course, In order to do that I'd imagine Intel would need to be sending MS some pretty convincing Atom Smartphone SOCs. Will Silvermont be powerful, energy efficient enough, and delivered within the right time frame for this to happen?

 

[cbn]

You make a great point. Looking at what my wife does with her iphone, the access to apps she has (and such affordable apps too! $3 and you have something you can use, MSOffice 2010 cost me $300 )...it is very obvious to me that the x86 pre-existing software advantage is a small one when it comes to the consumer space.

The MS software you are mentioning is expensive.

With Lenovo aiming at the mass market in China I just wonder what type of plans Mr. Rory Read is gearing up for AMD?

My guess would be that the AMD ARM license is on the way. But if we are talking the massive China market wouldn't that be very capital intensive?

 

[3DVagabond]

You make a great point. Looking at what my wife does with her iphone, the access to apps she has (and such affordable apps too! $3 and you have something you can use, MSOffice 2010 cost me $300 )...it is very obvious to me that the x86 pre-existing software advantage is a small one when it comes to the consumer space.

Obviously the server and HPC markets are different, x86 apps in those markets aren't about to be tossed aside and replaced with ARM versions for $3 a pop. But in the consumer space Apple has shown just how fickle the consumer market is to the legacy of apps or hardware.

Open office.

 

[cbn]

More information on Lenovo's mass market push into China with smartphones.

http://www.consumerelectronicsnet.c...ion-of-Chinas-3G-TD-SCDMA-Smartphones-1734624

Marvell, Lenovo and China Mobile Team Up To Drive Mass Market Adoption of China's 3G TD-SCDMA Smartphones

Marvell continues its leadership in China's 3G TD-SCDMA standard with introduction of Lenovo's LePhone A66t powered by industry's first commercially available single-chip TD-SCDMA solution


BEIJING, Oct. 26, 2011 /PRNewswire/ -- Marvell (NASDAQ: MRVL), a worldwide leader in integrated silicon solutions, today announced the launch of Lenovo's LePhone A66t, a 3.5-inch high-definition screen smartphone customized for China Mobile's 3G TD-SCDMA market and priced in the 1,000 RMB range. The LePhone A66t is powered by the Marvell PXA918 platform, the first commercially available single-chip solution that integrates a high-performance, low-power application processor with an advanced multimode modem. Marvell's advanced low-power TD single-chip solution is designed to enable exceptional user experience for watching mobile TV, navigation, video conferencing, social networking, and other popular mobile applications. It also features the Marvell Avastar- 88W8787 wireless solution, which is Bluetooth 3.0 and FM enabled, offering exceptional Wi-Fi range with beamforming technology, robust 802.11n connectivity and crystal clear audio quality.

(Logo: http://photos.prnewswire.com/prnh/20100719/SF36559LOGO-b)

"China is in a period of epic mobile growth, fueled by consumers' desires to leverage their devices for activities that are integral to their daily lives, from social networking to watching live contents. I believe Marvell's collaboration with Lenovo and China Mobile will help drive the connected lifestyle for hundreds of millions of consumers in China," said Weili Dai, Marvell's Co-Founder. "We're proud to see the leadership of Lenovo's Mobile Internet Digital Home Group in developing seamlessly connected lifestyle out of consumers' hands and into their homes."

You just have to wonder how AMD and Rory Read factor into what is happening here!

Rory Read would of course have accurate knowledge on the component pricing required to build a smartphone. This in turn would help AMD strategize some type of SOC entry into the market.

Furthermore, there is a desire stated by ARM's to "Flank Intel" with a lowest common denominator SOC. Originally they wanted to protect their flanks with Cortex A5 (that didn't work), but now I just wonder if they would be willing to give AMD a great deal on the Cortex A7 "Kingfisher".

 

[cbn]

More background information on Rory Read and Lenovo's smartphone entry. This article is from 2010.

http://www.fiercewireless.com/story...phone-uncorks-smartphone-ambitions/2010-04-20

Lenovo releases Android phone, uncorks smartphone ambitions
April 20, 2010 — 11:05am ET | By Phil Goldstein

Lenovo said it will begin selling its first smartphone, dubbed LePhone, in China next month--and the computer maker outlined its plans to push deeper into phones and mobile Internet services in general.

The company first unveiled the device at the Consumer Electronics Show this year, and will release it next month running on the WCDMA standard (which is used by China Unicom). Lenovo said it is also working on versions of the phone for China's other 3G standards, CDMA2000 and TD-SCDMA.

The smartphone runs a tweaked version of Android 2.1 and includes support for push email and other content services. The Chinese search engine Baidu replaces Google as the phone's primary search engine. The LePhone has a 1 GHz Qualcomm Snapdragon processor, 3.7-inch AMOLED screen and dual 3-megapixel cameras.

Lenovo plans on "winning in China first" before selling the phone in other markets, company COO Rory Read said, according to the Financial Times. Read said Lenovo plans to score between 10 percent and 20 percent of its revenue from mobile Internet services in five years. Interestingly, that estimate sharply contrasts with what Lenovo CEO Yang Yuanqing said in March; in an interview with the Associated Press, the Lenovo chief said the company expects mobile Internet products to be 70 percent to 80 percent of sales within three to five years.

Lenovo's Read said the company will have a leg up on Research In Motion's BlackBerry products in the Chinese market, and will aim its phones at consumers and enterprise users alike. He said Lenovo aims to sell "millions" of its phones over the next five years and "tens of millions" beyond that. However, the company will have to contend with other Android phones as well as Apple, which muscled into the Chinese market last fall when China Unicom began selling the iPhone.

Lenovo is one of a number of new entrants to the smartphone market. Dell, Garmin and others also recently jumped into the game in hopes of cashing in on what many expect will be dramatic, global growth in smartphones.

 

[cbn]

And ARM SOCs will stay in the same power usage/performance ??

Dont you thing that ARM Cortex A15 at 28nm will be much better in power usage/performance than current A9 at 40nm ??

http://www.anandtech.com/show/4991/...alcore-more-power-efficient-highend-devices/2

The plan is to drop the power usage of Cortex A15 by using Cortex A7 in a configuration called big.LITTLE.

I know this will drop idle power consumption, but I wonder how much will it affect power consumption with sub maximal loads?

Obviously there will be certain heavy tasks where turning on the Cortex A15 will be more efficient...the bigger CPU will eat more power but will get the job done quicker (resulting in less total energy consumed)

But how many loads will be better served with the Cortex A7?

 

[Predicament]

Can someone explain what x86 is?

 

[Idontcare]

Can someone explain what x86 is?

What does x86 mean

 

[Idontcare]

Rory Read would of course have accurate knowledge on the component pricing required to build a smartphone. This in turn would help AMD strategize some type of SOC entry into the market.

Rory Read would know a GREAT deal of info about not only Lenovo's business plans (and the motivations behind them) but also the business strategies of every potential chip supplier negotiating being a part of those business plans.

In terms of gaining insight and industry knowledge, hiring Rory was a stroke of genius for AMD. What remains to be seen is whether or not he's CEO level leadership.

Consider that hiring Dirk was also a stroke of genius for AMD (see K7), but he didn't quite turn out to be pro-league material once he was drafted to be quarterback on the exec team.

 

[cbn]

http://www.marvell.com/communication-processors/pxa918/assets/Marvell_PXA918_Platform-001.pdf

This is the SOC that will power Lenovo's Mass Market Chinese smartphone. (See Oct 26 Press release from Post #114)

624 Mhz ARMv5 Single Core CPU.

Process technology: not listed.

The Copyright on the PDF says 2011, but this design certainly looks like a old one. ARMv5 is no spring chicken!

My guess is that this part is very low cost....to go along with other low cost components comprising the 3.5" LCD smartphone.

 

[cbn]

Apparently Urban China likes the established smartphone brands (popular in the US), but Rural China still has lower penetration of smartphones.

I wonder how much of this has to do with power efficiency of the low end Smart phone devices targeted at Rural China? (often built with SOCs using larger process technology from the past)

My *guess* is that access to power outlets will be a factor in a lot of these cases. If that is true then maybe some of these old SOCs are holding down adoption of the more modern Smartphone OSes in these Asian Countries.

With that being said....maybe Rory Read has a unique perspective on how AMD can get funding, enter the Smartphone SOC business and produce something faster with better energy efficiency. AMD is a unique position to adopt tech like Cortex A7 as a standalone because they have no conflict of interest with an existing low end smartphone SOC lineup.