Motorola shifts to using Intel Atom mobile processors...

Matthew Kane

Supreme [H]ardness
Joined
Dec 1, 2007
Messages
4,233
Was just browsing gsmarena as it's that time of the year where you feel like upgrading or purchasing another phone to play around with, as a current Atrix 2 user what grabbed my attention was the upcoming Moto Razr i android phone that will be using a default clocked 2GHZ Intel Atom mobile processor.

Not much info besides the fact that Anandtech has stated that Moto will be shifting towards this direction of using the Atom line for their CPU's and the Razr i will be the first on the phone market to utilize this.

Their is no info on the expected performance figures the Atom's will be able to produce but it will be interesting to see it compare with the likes of the A6 chipset and the current average 1.5ghz clocked dual and quad cores android phones.

What do you guys think? Nothing besides speculation at this point, but it is interesting....

Source: http://www.anandtech.com/show/6307/motorola-hits-2ghz-with-intel-powered-razr-i

(Do hope that Motorola realizes that their software updates and locked bootloaders suck!).
 
Wow... you're a bit behind. ;)

There are plenty of reviews already. The performance is comparable to the RAZR M. Basically, it's a little slower in all of Engadget's favorite benchmarks except SunSpider... where I'm sure Intel enjoys a huge compiler advantage.

Power consumption appears to be almost the same as is weight and thickness.

Motorola won't be strictly Intel... This is just Google throwing them a bone I believe. The other RAZR products they're simultaneously launching are ARM.

I would be concerned with non-Java app compatibility though. I know Intel says they have an ARM emulation layer or whatever... but that kind of thing doesn't come for free (in terms of performance). Here I'm primarily referring to games and video and such.
 
hmm guess I am behind, don't follow on the latest tech news no more.

But if that's the case, then I'd rather hold off and see what other phones are on the market later on.
 
Next year's Intel chips are going to be a force to be reckoned with.
 
Compared to Nvidia and Samsung, does anyone know if Intel posts good documentation for their atom chips for ROM developers?
 
Will these Intel chips be capable of running in an RT form factor with full Windows 8 compatibility?
 
Will these Intel chips be capable of running in an RT form factor with full Windows 8 compatibility?

good question, I imagine they would be fine since there are Atom Processors that run Vista fine, I imagine these could run Win8 fine too.
 
Next year's Intel chips are going to be a force to be reckoned with.

You must work for Intel. They've been saying that every year for the past 4 about their mobile chips.

To date, they have what, 2 phones ever released?
 
I have to agree with chang on this one. Intel isn't stupid and they didn't grow to this obnoxious size by lack of business sense or smart people. They didn't know if the mobile shift away from PCs was real; now they do. Haswell will be the first architecture that Intel actually prioritizes mobile performance, and I fully expect it to blow everything else in the mobile space out of the water.
 
You must work for Intel. They've been saying that every year for the past 4 about their mobile chips.

To date, they have what, 2 phones ever released?
I wish I did!

The current roadmap shows Intel phone SoC's to go to 22 nm with 3d transistors next year, and 14 nm Broadwell in 2014. Current Haswell preview shows a upwards of 20x power reduction in idle mode. That technology should carry over to the tablets if not phones. A Haswell SoC would run circles around an ARM based tablet. When current upcoming Windows 8 tablets running a Core i3 Ivy Bridge lasts 10 hrs, a haswell tablet could run more than that (but manufacturers would probably rather reduce battery size for a lighter-weight tablet...)

When Broadwell comes out, it's pretty much the beginning of the end to ARM.
 
Last edited:
I have to agree with chang on this one. Intel isn't stupid and they didn't grow to this obnoxious size by lack of business sense or smart people. They didn't know if the mobile shift away from PCs was real; now they do. Haswell will be the first architecture that Intel actually prioritizes mobile performance, and I fully expect it to blow everything else in the mobile space out of the water.

Intel was fully aware of the huge market opening with the phone/tablet market (maybe not tablet) but that is the reason for the push on the Atom. They had been working on the Atom platform for a while especially when they had zero need on the desktop side since Celeron was very cheap and beating the competition with AMD's low end.

People were saying Intel was dumb for staying with a x86 CPU when an ARM based CPU was supposed to be unbeatable at performance/power at that level. It looks like Intel may be ready to destroy at that level.

The big problem with this is the GPU performance, the mOS's take advantage of GPU performance and I'm interested to see if Intel can maintain there too.

Sadly I don't think AMD can compete at this level, they have never been good at performance/power consmption and that is trully epic at this level.
 
Sadly I don't think AMD can compete at this level, they have never been good at performance/power consmption and that is trully epic at this level.
never? never? never? AMD K, Athlon, and the Athlon 64's architecture when they came out was way better in power efficiency and consumption compared to the Intel counterpart. It is not until Intel's Core architecture came out on their laptop platform did Intel catch up and they did in a big way. It was essentially AMD built by Intel with their superior manufacturing process. Intel keep the lead due in combination of mismanagement on AMD's part and mostly due to Intel's monopolistic practices in forcing OEM's to choose Intel or pay a much higher price.

Aside from historical facts, AMD has not been able to catch up again since Core. And I really doubt AMD would again unless AMD's combined CPU/GPU into one for reals. (AMD is supposedly working on it.)

Unto the Intel Broadwell hype:
14 nm with the Broadwell might be able to bring 64-bit processing down to mobile levels too. Mobile Broadwell could also support Intel HD5000 graphics. Ridiculous?!?

ARM: The Next AMD?
The only thing ARM can do now is hope that their SoC's would be 12-cores by that time. Broadwell will most likely be quad-cores with hyper-threading. So much of how AMD is still able to stay afloat is having more real cores clinching on their multi-threading lead. ARM will basically have to follow that model. ARM and AMD needs to merge sooner or later.

Will it be called ARM'D?
To be continued 2014...
 
lol @ the comments about ARM being buried by Intel

Intel is barely competitive in the embedded space as it is. It isn't going to get any easier in the coming years, especially as ARM software becomes more entrenched. And with companies like Apple and Samsung rolling their own chips, I don't see any reason they're going to buy them from Intel.

One area where Intel will need to compete however is price... and that means participating in a sector where there are less margins than they're used to.
 
Essentially what you're saying ARM will survive off embedded devices, like AMD will survive off graphics.

If I were to buy a $650 phone at a $200 subsidized price, I'll buy the Intel Broadwell phone over whatever ARM has...
 
Essentially what you're saying ARM will survive off embedded devices, like AMD will survive off graphics.

If I were to buy a $650 phone at a $200 subsidized price, I'll buy the Intel Broadwell phone over whatever ARM has...

Broadwell is at least two years away. Who knows what the situation could be by then? If NVIDIA's ridiculous press releases are to be believed, their chip will be "75 times faster than Tegra 2". (Ok, I laughed. I think they left a decimal off somewhere.)

Maybe by then Windows phones are big and Intel serves a purpose but otherwise the mobile software ecosystem is built around ARM and why buy a chip that has a performance penalty for hardware emulation (of certain native apps)?

And there's only so much you can get out of a process node shrink, and even there ARM isn't much behind.

I can see some people buying Intel-specific devices... but as it is now, the market is quite dominated by Apple and Samsung for the average folk. So I think ARM is here to stay for awhile.
 
Broadwell is at least two years away. Who knows what the situation could be by then? If NVIDIA's ridiculous press releases are to be believed, their chip will be "75 times faster than Tegra 2". (Ok, I laughed. I think they left a decimal off somewhere.)

Maybe by then Windows phones are big and Intel serves a purpose but otherwise the mobile software ecosystem is built around ARM and why buy a chip that has a performance penalty for hardware emulation (of certain native apps)?

And there's only so much you can get out of a process node shrink, and even there ARM isn't much behind.

I can see some people buying Intel-specific devices... but as it is now, the market is quite dominated by Apple and Samsung for the average folk. So I think ARM is here to stay for awhile.
yet how long ago did ARM come up with the Cortex A15 design before it reached the market? Somehow I don't think the next architecture ARM comes up with will arrive in time for the Broadwell onslaught.

The only reason why a person wouldn't want a phone with true desktop level performance while lasting a full day or more is that whoever manufactures the ARM processor may make a 12-core processor that beats Intel in few multi-threaded benchmarks and it will last 2 days or more.
 
yet how long ago did ARM come up with the Cortex A15 design before it reached the market? Somehow I don't think the next architecture ARM comes up with will arrive in time for the Broadwell onslaught.

The only reason why a person wouldn't want a phone with true desktop level performance while lasting a full day or more is that whoever manufactures the ARM processor may make a 12-core processor that beats Intel in few multi-threaded benchmarks and it will last 2 days or more.

Broadwell is not going to bring "desktop level performance" to a phone. There's a huge difference between their mobile and desktop chips, and once you have to work in a certain power envelope, performance comes down dramatically.

Not only is Intel incapable of doing that, it would be devastating to their business. Because phone SoCs sell for $25 and Intel won't even sell a low-end Core i3 for less than $125.
 
never? never? never? AMD K, Athlon, and the Athlon 64's architecture when they came out was way better in power efficiency and consumption compared to the Intel counterpart. It is not until Intel's Core architecture came out on their laptop platform did Intel catch up and they did in a big way. It was essentially AMD built by Intel with their superior manufacturing process. Intel keep the lead due in combination of mismanagement on AMD's part and mostly due to Intel's monopolistic practices in forcing OEM's to choose Intel or pay a much higher price.

AMD's offering that was doing better then Intel due to netburst failure and its not like AMD was offering stellar performance/watt then either. It looked that way when compared to the hog-netburst.

Core architecture came from the original Core architecture, which originated from Pentium M (which was an amazing laptop CPU) which had a lot of Pentium 3 technology. AMD didn't have much to do with the core architecture, hell Intel didn't put the memory controller on the CPU till after the core feature (which was one of the reasons for AMD's advantages over netburst).

Intels dominance comes from developing a majority of the x86 technologies and implementing them faster then the competition. They essentially destroy Cyrixx with it over night. Intels dominance with OEM's was mainly because they could supply the amount of CPU's needed, yes they still strong armed but they could actually supply a product too.

Either way this is off subject. AMD's current ability to get price/watt down is looking extremely slim especially their current line of increasing pipelines and core counts just to compete somewhere. The best thing for AMD is to strip their CPU and push performance through GPU acceleration. They have a great GPU architecture and just try to save on the CPU side!
 
Apparently you missed the part where Intel was found guilty...

Airmont based off Broadwell will reach us in 2014. 14 nm. Its performance is being advertised as better than last generation AMD's. And that's before they decided to redo something and moving Airmont back a year to 2015 and then have an accelerated roadmap back to 2014.
 
Apparently you missed the part where Intel was found guilty...

Airmont based off Broadwell will reach us in 2014. 14 nm. Its performance is being advertised as better than last generation AMD's. And that's before they decided to redo something and moving Airmont back a year to 2015 and then have an accelerated roadmap back to 2014.

Will this be in future smartphones?
 
Intel is definitely behind the curve when it comes to a mobile SoC solution for smartphones and tablets. That fact and the slow sale the desktop market segment has sent their stock price into a tailspin. Intel's stock (INTC) hit a new 52-week of $21.70 on Oct 10, 2012.

It's nice to hear that at their Atom CPU is being used in 2 cellphones, but make no mistake, Intel has an uphill battle to make a niche for itself in the mobile market. The one thing that Intel has is $$$$ which it can use to help accelerate R&D in this market segment if it wanted to. Based on current market conditions, I think Intel needs to come out swinging in this market segment so that it can continue to grow as a company.

While Intel's 14nm Airmont SoC is expected to be released in 2014, that is still a decent amount of time away from now. At least it is a step in the right direction. Perhaps Intel's Silvermont CPU will offer good enough performance in 2013 to be at about on par with the other major players in the mobile CPU business.

Once Intel's stock price stabilizes I just might pick up some shares. At least the $0.92 per share dividend will offer some return as the stock slowly recover from the sell off. It would be nice if pick up INTC for $16 per share, but I really doubt it will drop that far without a general market correction (i.e. all stocks drop simply because people are selling out of fear and / or to lock in gains).
 
Unfortunately I don't believe Silvermont at 22nm would be competitive with the top of the line cortex A15 quads. Silvermont is just a die shrink. Intel needs two/three years to heavily integrate its desktop/tablet design to the mobile front. Haswell will already be a killer in tablets.
 
Back
Top