- Status
cbfelterbush
Distinguished
- Oct 3, 2011
- 63
- 0
- 18,630
Underestimating Intel? I am thinking that that could lead to much soreness. Intel is an awesome company for pulling all the strings at the right time. When Intel releases its competition between now and 2014, Arm will be very aware. Despite claims that there is no worry to be had, Arm in not ignorant.
Alidan said:
"you have to remember, this is an x86 processor coming into a land scape where nothing is x86 compatible.
regardless of phone, damn near everything is arm compatible, from iphone to android, and its likely only windows will be x86, and only really in the tablet devices, because its x86 compatible it will run real programs, but it has no phone apps, so i cant see it doing well there, and probably be irrelevant for another 4-6 years. BUT at the same time, we are seeing arm getting powerful to the point where servers are looking into them as an alternative, for being cheap and power friendly.
unless a x86 phone comes out with unprecedented support (it wont) than arm has noting to worry about, because it will canablize desktop, laptop, and server cpu sales soon.
and even than, a better x86 processor for tablets, and phones, would be amds apus, graphics that intel wishes they could have, and more than enough power to get other things done, maybe not as fast as intel, but we are talking about seconds shaved off, not minutes."
Alidan, AMD APUs use FAR too much energy. Intel won't be using NVIDIA's Ion as a chipset. Ion is what consumes so much energy on that platform. ATOM as a CPU uses much less energy already than an APU and that difference will likely be much greater when ATOM gets revamped. I agree that there are no x86 apps, but the OS and kernel drivers are the bigger issue. Writing Objective-C and JAVA code for x86 based systems isn't going to be a big issue for a lot of app developers. Writing drivers that access antennas, gyroscope, devices, etc. and internal features of the chipset are big issues. Those will be solved with Intel's support much easier than people realize. What IS DIFFERENT is that Intel will have to have partners... much like ARM. Remember that Intel was ARM's BIGGEST partner in the xScale days. ATOM negated the need for xScale and I don't think Intel saw the phone as a big CPU market after the PDA died. Just poor insight on Intel's end, but they have the knowledge and the ability to quickly jump into this market. People tend to think that Intel is a company that has only made PC parts. They've made parts for all sorts of devices over the years with RISC chips, graphics chips, embedded parts, etc. They are not nearly as far behind as some would imagine. The big issue is getting partners more than making a competitive part. The last part is almost a given with a little time.
"you have to remember, this is an x86 processor coming into a land scape where nothing is x86 compatible.
regardless of phone, damn near everything is arm compatible, from iphone to android, and its likely only windows will be x86, and only really in the tablet devices, because its x86 compatible it will run real programs, but it has no phone apps, so i cant see it doing well there, and probably be irrelevant for another 4-6 years. BUT at the same time, we are seeing arm getting powerful to the point where servers are looking into them as an alternative, for being cheap and power friendly.
unless a x86 phone comes out with unprecedented support (it wont) than arm has noting to worry about, because it will canablize desktop, laptop, and server cpu sales soon.
and even than, a better x86 processor for tablets, and phones, would be amds apus, graphics that intel wishes they could have, and more than enough power to get other things done, maybe not as fast as intel, but we are talking about seconds shaved off, not minutes."
Alidan, AMD APUs use FAR too much energy. Intel won't be using NVIDIA's Ion as a chipset. Ion is what consumes so much energy on that platform. ATOM as a CPU uses much less energy already than an APU and that difference will likely be much greater when ATOM gets revamped. I agree that there are no x86 apps, but the OS and kernel drivers are the bigger issue. Writing Objective-C and JAVA code for x86 based systems isn't going to be a big issue for a lot of app developers. Writing drivers that access antennas, gyroscope, devices, etc. and internal features of the chipset are big issues. Those will be solved with Intel's support much easier than people realize. What IS DIFFERENT is that Intel will have to have partners... much like ARM. Remember that Intel was ARM's BIGGEST partner in the xScale days. ATOM negated the need for xScale and I don't think Intel saw the phone as a big CPU market after the PDA died. Just poor insight on Intel's end, but they have the knowledge and the ability to quickly jump into this market. People tend to think that Intel is a company that has only made PC parts. They've made parts for all sorts of devices over the years with RISC chips, graphics chips, embedded parts, etc. They are not nearly as far behind as some would imagine. The big issue is getting partners more than making a competitive part. The last part is almost a given with a little time.
Pretty sure Intel got ARM from DEC when DEC split up years ago and Intel took all of it's hardware technology. They were top dogs in the PDA and other handheld devices arena back in the last decade. Intel jumped out because it wasn't a lucrative business and PDAs were falling by the wasteside. This is about the time when the Blackberry was becoming the big thing and right before Apple jumped in with the iPhone. Once the iPhone hit it big. Intel figured the market for small form factor laptops was the big seller so they created the ATOM and thus dropped the ARM.
ATOM was a HUGE success despite what some on here are saying. They owned a single industry all by themselves for nearly 4 years until AMD just recently joined in. It made Intel a ton of money. Now the ATOM design is antiquated and needs to be revamped since things have moved from small form factor laptops to the tablet and phone arena. Who could have predicted that smart phones would be all the rage back in 2003 when Intel first concepted the ATOM. By 2005 they were committed and that is when PDAs started really falling out of favor so it looked like a good idea. I'm sure if hindsight were 20/20, they would have built an ATOM AND another variant for smart phones.
ATOM was a HUGE success despite what some on here are saying. They owned a single industry all by themselves for nearly 4 years until AMD just recently joined in. It made Intel a ton of money. Now the ATOM design is antiquated and needs to be revamped since things have moved from small form factor laptops to the tablet and phone arena. Who could have predicted that smart phones would be all the rage back in 2003 when Intel first concepted the ATOM. By 2005 they were committed and that is when PDAs started really falling out of favor so it looked like a good idea. I'm sure if hindsight were 20/20, they would have built an ATOM AND another variant for smart phones.
[citation][nom]K2N hater[/nom]I don't think selling xscale was a bad deal. Intel can fight ARM without moving away from X86. All Atom needs is to draw less power and I'm sure Intel is working hard on it.[/citation]
The fact will always be that x86 takes much more idle and load power consumption evan at 22nm especially compared to arm cortex a9 designs of sub-watt power consumption
The fact will always be that x86 takes much more idle and load power consumption evan at 22nm especially compared to arm cortex a9 designs of sub-watt power consumption
edlivian... it won't matter too much. Batteries are making steady improvement and updated battery technology is around the corner as well. If they can get it close while added a great deal more computational power, they will be extremely competitive. People will still be able to power their phones for hours. Also, I think Intel is shooting at 14nm for a competitive product. That is where they will start bumping into ARM so ARM has a few years before it has to worry. The 22nm parts will be competitive but the 14nm part will start muscling in on ARM's power territory.
You're all forgetting that the smartphone hardware market is driven by what manufacturers decide to put in the handsets, not what consumers ask to be put in them. It's not like a PC.
When was the last time you heard a typical consumer request a Cortex Ax processor? They don't, they want a 1Ghz processor or dual core or Android or an iPhone or whatever.
Samsung are a big player in the smartphone market and are somewhat unique in the regard that they are a tier 1 manufacturer. By this I mean Samsung build their own chips, memory, display panels etc whereas other companies (HTC for example, and even Apple) purchase these components elsewhere.
So, the more appropriate question is not Intel vs ARM, it's Intel vs Samsung, nvidia, Qualcomm, Texas Instruments et al who all produce hardware for the smartphone space.
My guess is that it would be a small miracle for Intel to break into this market in the same way as the PC world - because it's not consumers who choose what goes in the phones, it's the manufacturers. It would have to make a financially sound proposition for them to choose Intel over ARM (unlikely) - especially in the current margin-stricken climate with licensing costs, law suits and the overbearing high-end market dominance of Apple.
In short, Intel don't need to win over consumers to make this happen. They need to win over handset manufacturers. And at the moment, ARM have a massive advantage there.
ARM are in a far more sheltered position than AMD were...
When was the last time you heard a typical consumer request a Cortex Ax processor? They don't, they want a 1Ghz processor or dual core or Android or an iPhone or whatever.
Samsung are a big player in the smartphone market and are somewhat unique in the regard that they are a tier 1 manufacturer. By this I mean Samsung build their own chips, memory, display panels etc whereas other companies (HTC for example, and even Apple) purchase these components elsewhere.
So, the more appropriate question is not Intel vs ARM, it's Intel vs Samsung, nvidia, Qualcomm, Texas Instruments et al who all produce hardware for the smartphone space.
My guess is that it would be a small miracle for Intel to break into this market in the same way as the PC world - because it's not consumers who choose what goes in the phones, it's the manufacturers. It would have to make a financially sound proposition for them to choose Intel over ARM (unlikely) - especially in the current margin-stricken climate with licensing costs, law suits and the overbearing high-end market dominance of Apple.
In short, Intel don't need to win over consumers to make this happen. They need to win over handset manufacturers. And at the moment, ARM have a massive advantage there.
ARM are in a far more sheltered position than AMD were...
[citation][nom]LastManUp[/nom]Alidan said:"you have to remember, this is an x86 processor coming into a land scape where nothing is x86 compatible. regardless of phone, damn near everything is arm compatible, from iphone to android, and its likely only windows will be x86, and only really in the tablet devices, because its x86 compatible it will run real programs, but it has no phone apps, so i cant see it doing well there, and probably be irrelevant for another 4-6 years. BUT at the same time, we are seeing arm getting powerful to the point where servers are looking into them as an alternative, for being cheap and power friendly. unless a x86 phone comes out with unprecedented support (it wont) than arm has noting to worry about, because it will canablize desktop, laptop, and server cpu sales soon. and even than, a better x86 processor for tablets, and phones, would be amds apus, graphics that intel wishes they could have, and more than enough power to get other things done, maybe not as fast as intel, but we are talking about seconds shaved off, not minutes."Alidan, AMD APUs use FAR too much energy. Intel won't be using NVIDIA's Ion as a chipset. Ion is what consumes so much energy on that platform. ATOM as a CPU uses much less energy already than an APU and that difference will likely be much greater when ATOM gets revamped. I agree that there are no x86 apps, but the OS and kernel drivers are the bigger issue. Writing Objective-C and JAVA code for x86 based systems isn't going to be a big issue for a lot of app developers. Writing drivers that access antennas, gyroscope, devices, etc. and internal features of the chipset are big issues. Those will be solved with Intel's support much easier than people realize. What IS DIFFERENT is that Intel will have to have partners... much like ARM. Remember that Intel was ARM's BIGGEST partner in the xScale days. ATOM negated the need for xScale and I don't think Intel saw the phone as a big CPU market after the PDA died. Just poor insight on Intel's end, but they have the knowledge and the ability to quickly jump into this market. People tend to think that Intel is a company that has only made PC parts. They've made parts for all sorts of devices over the years with RISC chips, graphics chips, embedded parts, etc. They are not nearly as far behind as some would imagine. The big issue is getting partners more than making a competitive part. The last part is almost a given with a little time.[/citation]
but on a tablet, more than a cellphone, could be a more graphical thing. even if an amd apu takes more power, an intel cpu with their integrated... its abysmal, and if they use a third party, it will take more power.
if tablets or cellphones ever really do take off as a true successor to the psp or ds, the first person to do a cpu and gpu combo right will be the leader, and i also believe that x86 would be necessary to some extent, as they are more powerful than the arm solution.
im not sure if what i want to say is getting across or not.
but on a tablet, more than a cellphone, could be a more graphical thing. even if an amd apu takes more power, an intel cpu with their integrated... its abysmal, and if they use a third party, it will take more power.
if tablets or cellphones ever really do take off as a true successor to the psp or ds, the first person to do a cpu and gpu combo right will be the leader, and i also believe that x86 would be necessary to some extent, as they are more powerful than the arm solution.
im not sure if what i want to say is getting across or not.
eddieroolz
Splendid
- Sep 6, 2008
- 7,057
- 0
- 25,860
I see a lot of comments about Intel and AMD, but there is a major difference between the two.
Intel and AMD both relied on x86 processors, while ARM relies on a separate architecture.
So, in this case, Intel really might be irrelevant in the Android world.
Intel and AMD both relied on x86 processors, while ARM relies on a separate architecture.
So, in this case, Intel really might be irrelevant in the Android world.
"The fact will always be that x86 takes much more idle and load power consumption evan at 22nm especially compared to arm cortex a9 designs of sub-watt power consumption"
Except that Intel has created a Near-Threshold Voltage (NSV) processor using a Pentium architecture (it is cheaper to alter, so is often used for research CPUs). NSV uses just 500 mV of power, which is way less than any ARM CPUs. Sure, NSV will not be seen in Intel CPUs for many years, but given no one else has announced doing that yet you can bet it will take even longer for ARM, their partners, and the foundries they use to manage it.
Furthermore, you know what is not so far away? 22 nm Tri-Gate. The massive power savings this will allow Intel will make them a lot more competitive in power consumption. ARM may be close to 28 nm, but 28 nm uses more power than ordinary 22 nm let alone 22 nm Tri-Gate. And by the time TSMC and other foundries get to 20 nm Intel will be at 14 nm, again using Tri-Gate while the other foundries won't be using Tri-Gate. In fact no foundry has announced plans for Tri-Gate.
Except that Intel has created a Near-Threshold Voltage (NSV) processor using a Pentium architecture (it is cheaper to alter, so is often used for research CPUs). NSV uses just 500 mV of power, which is way less than any ARM CPUs. Sure, NSV will not be seen in Intel CPUs for many years, but given no one else has announced doing that yet you can bet it will take even longer for ARM, their partners, and the foundries they use to manage it.
Furthermore, you know what is not so far away? 22 nm Tri-Gate. The massive power savings this will allow Intel will make them a lot more competitive in power consumption. ARM may be close to 28 nm, but 28 nm uses more power than ordinary 22 nm let alone 22 nm Tri-Gate. And by the time TSMC and other foundries get to 20 nm Intel will be at 14 nm, again using Tri-Gate while the other foundries won't be using Tri-Gate. In fact no foundry has announced plans for Tri-Gate.
I don't think ARM came from DEC contrary to what is said in below post. See http://en.wikipedia.org/wiki/ARM_architecture for details
Anonymous 10/26/2011 6:26 PM
Pretty sure Intel got ARM from DEC when DEC split up years ago and Intel took all of it's hardware technology. They were top dogs in the PDA and other handheld devices arena back in the last decade. Intel jumped out because it wasn't a lucrative business and PDAs were falling by the wasteside. This is about the time when the Blackberry was becoming the big thing and right before Apple jumped in with the iPhone. Once the iPhone hit it big. Intel figured the market for small form factor laptops was the big seller so they created the ATOM and thus dropped the ARM.
Anonymous 10/26/2011 6:26 PM
Pretty sure Intel got ARM from DEC when DEC split up years ago and Intel took all of it's hardware technology. They were top dogs in the PDA and other handheld devices arena back in the last decade. Intel jumped out because it wasn't a lucrative business and PDAs were falling by the wasteside. This is about the time when the Blackberry was becoming the big thing and right before Apple jumped in with the iPhone. Once the iPhone hit it big. Intel figured the market for small form factor laptops was the big seller so they created the ATOM and thus dropped the ARM.
- Status
TRENDING THREADS
-
News AMD says Intel's 'horrible product' is causing Ryzen 9 9800X3D shortages- Started by Admin
- Replies: 83
-
-
Discussion PSU recommendations and power supply discussion thread - Tom's hardware- Started by Darkbreeze
- Replies: 2K
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
-
Facebook
Twitter
Instagram