Arm is having to scale their performance up while Intel is having to scale its Power consumption down and i think Toms just did a article saying the Ivy-bridge will only consume up to 77watts. That's pretty amazing. Arm is only worth about 10B and Intel is worth 120B+. And Intel has their own FABS where Arm does not. I'd say AMD, ARM, Nvidia, GF, TSMC, and All the other small guys need to put their research together and take on Intel.
- Status
X86 Android is available. You download the alpha version of BlueStacks for Windows 7 or buy ViewSonic's ViewPad.
If the Android apps you want is pure Java/Dalvik, it will already work find. If your app also uses native ARM libraries, it won't even install.
If the Android apps you want is pure Java/Dalvik, it will already work find. If your app also uses native ARM libraries, it won't even install.
shin0bi272
Distinguished
- Nov 20, 2007
- 1,103
- 0
- 19,310
since cisc concentrates on hardware and risc concentrates on software there will come a point where given enough cores a risc chip will be equal and/or superior to a cisc chip. This is evident when we look at video game benchmark FPS not changing between a dual core and quad core with HT processor. At some point the speed advantage of cisc will not manifest itself except for when running multiple vm's on a box. I mean if you had a 16 core risc processor and a wide enough bus you wouldnt notice the difference between it and a 16 core cisc chip. That is of course speculation but as we look at the differences between the two shrinking currently there should be a theoretical break even point.
Tomtompiper
Distinguished
- Jan 20, 2010
- 382
- 0
- 18,780
[citation][nom]jdwii[/nom]Arm is having to scale their performance up while Intel is having to scale its Power consumption down and i think Toms just did a article saying the Ivy-bridge will only consume up to 77watts. That's pretty amazing. Arm is only worth about 10B and Intel is worth 120B+. And Intel has their own FABS where Arm does not. I'd say AMD, ARM, Nvidia, GF, TSMC, and All the other small guys need to put their research together and take on Intel.[/citation]
What is the most successful sports brand on the planet? How many people do they employ, and how big is their manufacturing base?
What is the most successful sports brand on the planet? How many people do they employ, and how big is their manufacturing base?
technoholic
Distinguished
- Feb 27, 2008
- 800
- 0
- 19,160
IMHO intel is going to need some kind of "real" graphics horse power in their products, sooner or later. Either a new tech that gives Intel's CPU's advantage in graphic works or something else, i dunno. This is where it goes. AMD realised this earlier but gave wrong decisions in the wrong direction. (talking about BD failure not llano; although newer generation llanos will be built with BD modules?)
On the other hand, like the author mentioned: "Intel has, by far, the most advanced production process and immense capability to produce and adjust the production of microchips."
And r&d is all about who has the bigger wallet and more geniuses working for them. I think intel has both. I don't think intel is missing ANYTHING for the near future
On the other hand, like the author mentioned: "Intel has, by far, the most advanced production process and immense capability to produce and adjust the production of microchips."
And r&d is all about who has the bigger wallet and more geniuses working for them. I think intel has both. I don't think intel is missing ANYTHING for the near future
Yeah... right...
Here we have AMD, who now provides 80% to 110% of Intel's performance in everything except synthetics that were compiled with ICC, and they're "slow, uncompetitive, dissapointing", and generally considered not suitable for anybody's PC, since we all need soooooo much processing power all the time.
Now we're suggesting the real battle is smartphones and tablets, even though Intel is not remotely competitive, and never has been, we're suggesting they're somehow the incumbent, and will now have to fight off ARM for taking their non-existent piece of the pie?
This part is just classic:
""As the need for computing performance goes up, both the Intel architecture and the ARM architectures face the same fundamental physics problems, which is more performance requires more transistors"
Orly? Is that why the Atom chips for smartphones, etc... are bigger than ARM's entire SOC just for the CPU, consume 4x as much power, and are nowhere near 4x as fast? Yeah, you and ARM are both just fighting the laws of physics now, right....
Here we have AMD, who now provides 80% to 110% of Intel's performance in everything except synthetics that were compiled with ICC, and they're "slow, uncompetitive, dissapointing", and generally considered not suitable for anybody's PC, since we all need soooooo much processing power all the time.
Now we're suggesting the real battle is smartphones and tablets, even though Intel is not remotely competitive, and never has been, we're suggesting they're somehow the incumbent, and will now have to fight off ARM for taking their non-existent piece of the pie?
This part is just classic:
""As the need for computing performance goes up, both the Intel architecture and the ARM architectures face the same fundamental physics problems, which is more performance requires more transistors"
Orly? Is that why the Atom chips for smartphones, etc... are bigger than ARM's entire SOC just for the CPU, consume 4x as much power, and are nowhere near 4x as fast? Yeah, you and ARM are both just fighting the laws of physics now, right....
It still seems like a war Intel is far from even drawing blood from. Intel needs to catch up immensely in the R&D department and need to get some vendor support from current Arm users. You also need to take into account one of the biggest, Samsung, would prefer to stick with their own chipset most likely and continue their Android dominance which has an impact on Intel dramatically; not being able to cater to the biggest player will hurt.
Frankly 2 years in technology terms is a long term and if intel is still working on the catch-up game, they may have to settle with something less than spectacular. I think Intel lost huge grounds and may not ever break into this marketplace the way it could've had it really put the effort into it earlier.
Frankly 2 years in technology terms is a long term and if intel is still working on the catch-up game, they may have to settle with something less than spectacular. I think Intel lost huge grounds and may not ever break into this marketplace the way it could've had it really put the effort into it earlier.
Opinion: Intel is underestimating AMD and Nvidia on graphics...
Fact: Intel doesn't know how to do anything without having a monopoly over it, and they can't create a monopoly on anything they didn't do first. Don't look for them to take over anything, ever, because they're still holding on for dear life to what they did in the 70s and 80s.
Fact: Intel doesn't know how to do anything without having a monopoly over it, and they can't create a monopoly on anything they didn't do first. Don't look for them to take over anything, ever, because they're still holding on for dear life to what they did in the 70s and 80s.
CyberAngel
Distinguished
- Dec 11, 2008
- 113
- 0
- 18,680
AppleBlowsDonkeyBalls
Distinguished
- Sep 30, 2010
- 117
- 0
- 18,680
Nah, I don't think they're underestimating them. Not that they should be very worried, anyway. Sandy Bridge and Ivy Bridge have an absolutely enormous performance advantage in comparison to the newest chips from the different ARM semiconductors. Sandy Bridge, while it has good power consumption for laptops, for smaller devices like tablets it's simply too much of a power hog.
Given that, Intel decided to manufacture Ivy Bridge using these new 3D transistors at a 22nm process node. They have a HUGE advantage in comparison to all others when it comes to this, and even AMD that's behind was able to get 32nm products sooner to the market than ARM semiconductors, most of which are still relying on 45nm and 40nm.
ARM SoCs may have power consumption figures in the under 1W number, but its performance figures are also very small and that will not change. ARM is engineered for low power first and foremost, and it has a lot more problems than X86 when it comes to performance scaling. Intel is taking advantage of this and their manufacturing process lead and releasing Ivy Bridge, which should have around 5% higher IPC than Sandy Bridge, but more importantly much higher performance/watt. 22nm can definitely be a game changer, as in laptops more than 8 hours battery life or so with standard usage won't matter much to most consumers, but having more performance will. That's why ARM won't matter for laptops. For tablets there's now Windows 8, and by the time it comes out Ivy Bridge will probably be here. You'll give up on battery life, but you'll also get MUCH higher performance/watt.
Given that, Intel decided to manufacture Ivy Bridge using these new 3D transistors at a 22nm process node. They have a HUGE advantage in comparison to all others when it comes to this, and even AMD that's behind was able to get 32nm products sooner to the market than ARM semiconductors, most of which are still relying on 45nm and 40nm.
ARM SoCs may have power consumption figures in the under 1W number, but its performance figures are also very small and that will not change. ARM is engineered for low power first and foremost, and it has a lot more problems than X86 when it comes to performance scaling. Intel is taking advantage of this and their manufacturing process lead and releasing Ivy Bridge, which should have around 5% higher IPC than Sandy Bridge, but more importantly much higher performance/watt. 22nm can definitely be a game changer, as in laptops more than 8 hours battery life or so with standard usage won't matter much to most consumers, but having more performance will. That's why ARM won't matter for laptops. For tablets there's now Windows 8, and by the time it comes out Ivy Bridge will probably be here. You'll give up on battery life, but you'll also get MUCH higher performance/watt.
IMHO, I think if Intel could have it would have. It's a testimony to ARMs persistence in the industry and their partnerships with manufacturers that put it at an advantage. Processors don't really sell phones, Phones sell phones and until they get their processors in one they aren't in the game. Furthermore even if they do start getting them in a couple, they're going to have to be red hot if Apple themselves don't dump their own part for Intels (and how likely is that to happen?).
It could even be argued that AMD with their low powered, decent 3d performance of their C and E series are closer to something that would fit in, if not a phone some other mobile device. Power is the major consideration for these devices after all, I haven't heard too many people screaming that they need a faster proc, and Apple has proven that balanced design is more important. It's evident that in the PC industry the race for more clocks/cores made alot of people lazier with balanced design.
It could even be argued that AMD with their low powered, decent 3d performance of their C and E series are closer to something that would fit in, if not a phone some other mobile device. Power is the major consideration for these devices after all, I haven't heard too many people screaming that they need a faster proc, and Apple has proven that balanced design is more important. It's evident that in the PC industry the race for more clocks/cores made alot of people lazier with balanced design.
kronos_cornelius
Distinguished
- Nov 4, 2009
- 365
- 1
- 18,780
Otelli is doing what he has to do. You don't expect him to come out and say "guys, we have a great challenge with ARM." PR and the business strategy are two different things. Intel's biggest problem is backwards compatibility. I have no doubt that ARM's architectures would be better if printed with Intel's technology. But Intel is having trouble matching ARM's power efficiency, and the problem, it seems, is not one that can be resolved in one year. Short of abandoning x86, I think Intel has big challenges if it want to stay attached to the past. A past that also gives it great leverage
To me, the architecture of the future is Tegra not ARM. Now, I know ARM is inside Tegra, but ARM by itself is not enough to compete with Intel. Tegra and Project Denver, however, will give Intel a very tough fight.
That is why AMD is wise to pursue ARM. Intel is not as friendly with their x86 specification as ARM is. AMD, already with its powerfull ATI branch, can imitate Nvidia. Of course they will be trailing Nvidia like they have Intel, but at least they will survive to the next techno-logic techtonic shift.
To me, the architecture of the future is Tegra not ARM. Now, I know ARM is inside Tegra, but ARM by itself is not enough to compete with Intel. Tegra and Project Denver, however, will give Intel a very tough fight.
That is why AMD is wise to pursue ARM. Intel is not as friendly with their x86 specification as ARM is. AMD, already with its powerfull ATI branch, can imitate Nvidia. Of course they will be trailing Nvidia like they have Intel, but at least they will survive to the next techno-logic techtonic shift.
The way I recall Intel's latest conference call, some combination of tablets and the lack of native 1080P support are devastating the atom netbook market. Overall Atom revenue was off 32%. On the other hand, in the embedded space Atom is on a tear with many design wins going against the ARM ecosystem.
It is hard to win when you are not in the game (tablets) and Intel really should bring 1080P to Atom (Cedar Trail). When customers care about performance, Intel does pretty good against ARM (embedded). We shall learn more as Intel expands the markets it is trying to address. However, keep in mind that Atom is only now entering these markets using the remnants of the original 5 year development cycle that started Atom. That Intel has any success with a 5 year old design suggests that this "fiercely competitive" market will be easy pickings for Atom under Tick/Tock.
It is hard to win when you are not in the game (tablets) and Intel really should bring 1080P to Atom (Cedar Trail). When customers care about performance, Intel does pretty good against ARM (embedded). We shall learn more as Intel expands the markets it is trying to address. However, keep in mind that Atom is only now entering these markets using the remnants of the original 5 year development cycle that started Atom. That Intel has any success with a 5 year old design suggests that this "fiercely competitive" market will be easy pickings for Atom under Tick/Tock.
itchyisvegeta
Distinguished
- Oct 19, 2010
- 232
- 0
- 18,680
In my opinion, the following scenario would be very interesting:
AMD and Nvidia should merge - not one buying the other but merge.
What would happen. They could speed up their products development cycle and they would be competitive on both markets - x86 + GPU and ARM + GPU.
Intel does have one very big weakness, and that is its GPU. Both AMD and Nvidia are miles ahead on this field and they should start using it in their advantage.
They would both also benefit in GPU market - even if that is not visible right now. In a keynote John Carmack said, that the software layers between OS and GPU hardware has become so complex and large, that it actually slows down the software by a significant margin - so if there were no more battle between AMD and Nvidia, they could let the developers direct access to hardware and left DirectX and other libraries go to the history dump - that would see a real graphic performance boost in x86 and ARM playground an it would left Intel exposed!
Best regards. France
AMD and Nvidia should merge - not one buying the other but merge.
What would happen. They could speed up their products development cycle and they would be competitive on both markets - x86 + GPU and ARM + GPU.
Intel does have one very big weakness, and that is its GPU. Both AMD and Nvidia are miles ahead on this field and they should start using it in their advantage.
They would both also benefit in GPU market - even if that is not visible right now. In a keynote John Carmack said, that the software layers between OS and GPU hardware has become so complex and large, that it actually slows down the software by a significant margin - so if there were no more battle between AMD and Nvidia, they could let the developers direct access to hardware and left DirectX and other libraries go to the history dump - that would see a real graphic performance boost in x86 and ARM playground an it would left Intel exposed!
Best regards. France
It is interesting that Intel thinks that have a 2 year production process advantage when real world figures shows the exact opposite. Intel is starting to lose the production process race.
Even today Glo-Fo produces processors on a 32nm SOI process node that has 37% more transistors per mm2 and 5% lower power consumption per mm2 than Intels 32nm bulk process node. And several independent fabs produce chips on a 28nm process node. Admittedly though not as advanced as Intels bulk process but it still has higher transistor density than Intels bulk process but Intel has the power consumption advantage per mm2.
Power consumption advantage per mm2 for the Intel process is rather irrelevant as it's x86 architecture requires far more transistor to achieve the same performance as the ARM architecture.
Even this week ARM succesfully demonstrated a tape-out of ARM chips on a 20nm process. This is intended for FD-SOI process that is going to be the de facto industry standard in 2-3 years. To compete Intel will have to successfully leap-frog their FinFet 14nm process node as these will have similar performance.
It is my estimation that Intel, unless it completely turns around it's business, has already lost the war on technology and design, as well as numbers. Let's not forget that every time Intel ships 1 core, ARM core manufacturers ships 20.
Even today Glo-Fo produces processors on a 32nm SOI process node that has 37% more transistors per mm2 and 5% lower power consumption per mm2 than Intels 32nm bulk process node. And several independent fabs produce chips on a 28nm process node. Admittedly though not as advanced as Intels bulk process but it still has higher transistor density than Intels bulk process but Intel has the power consumption advantage per mm2.
Power consumption advantage per mm2 for the Intel process is rather irrelevant as it's x86 architecture requires far more transistor to achieve the same performance as the ARM architecture.
Even this week ARM succesfully demonstrated a tape-out of ARM chips on a 20nm process. This is intended for FD-SOI process that is going to be the de facto industry standard in 2-3 years. To compete Intel will have to successfully leap-frog their FinFet 14nm process node as these will have similar performance.
It is my estimation that Intel, unless it completely turns around it's business, has already lost the war on technology and design, as well as numbers. Let's not forget that every time Intel ships 1 core, ARM core manufacturers ships 20.
feeddagoat
Distinguished
- Apr 11, 2010
- 329
- 0
- 18,790
Bare in mind it wasn't intel who created the netbook, it was Asus. Intel saw a chance to use old designs and make a bit of money off it. After netbooks took off intel sat up a bit but still didn't really move until recently with the ultrabooks. Why don't they strip down sandybridge and release a low power, low clocked, single core with hyperthreading version? It'll be more efficient and faster than any atoms on the market atm.
Everyone always said that intel and AMD would lock Nvidia out of the APU race, but now microsoft have thrown them a bone with windows 8 supporting ARM and x86. ARM has the advantage of to the metal programming, where as x86 is very general wide range support. X86 is arguably intels Achilles heel when it comes to low powered devices. PC's have always been "throw more power/bigger is better" where as low power requires finesse in programming. TBH there is no reason intel can't compete if they specifically design an OS around their own x86 architecture.
Everyone always said that intel and AMD would lock Nvidia out of the APU race, but now microsoft have thrown them a bone with windows 8 supporting ARM and x86. ARM has the advantage of to the metal programming, where as x86 is very general wide range support. X86 is arguably intels Achilles heel when it comes to low powered devices. PC's have always been "throw more power/bigger is better" where as low power requires finesse in programming. TBH there is no reason intel can't compete if they specifically design an OS around their own x86 architecture.
southernshark
Distinguished
- Nov 7, 2009
- 1,015
- 6
- 19,295
Intel has been behind before. Remember its 2 year tick strategy. They may end up being behind for 2 years at the tablet end of things. But when they shrink the die again... and go to whatever is lower than 22nm and also will have adopted an even more advanced architecture.... well at that point I see Intel being the only real choice. Everything else will be tier 2/3.
Right now the problem with ARM is not lacking performance(they are faster than my PIII 600Mhz) ,but compatibility between devices and price.I have seen full blown laptop with dvd and 15' screen with athlon processor for less than pretty much every tablet on market ,being ipad or android. For ARM Windows 8 will really be milestone and a step forward, but they will face though competition from intel's and AMD's yesterdays products.In the x86 corner there is HUGE array of free and good programs.
I think AMD is actually competing with Intel a lot better than the general perception is out there.
For instance, a 8 core processor will be far superior to a 4 core once the applications are multi core capable.
Perhaps AMD rushed to the future too soon, since currently performance per core beats more cores (in games this is especially true).
Myself, id like to see games that would use 8 cores + GPU core + Integrated GPU for mainwain for faster alt-tab.
That would be sick quality
.
For instance, a 8 core processor will be far superior to a 4 core once the applications are multi core capable.
Perhaps AMD rushed to the future too soon, since currently performance per core beats more cores (in games this is especially true).
Myself, id like to see games that would use 8 cores + GPU core + Integrated GPU for mainwain for faster alt-tab.
That would be sick quality
.sonofliberty08
Distinguished
- Mar 17, 2009
- 658
- 0
- 18,980
[citation][nom]copfrance[/nom]In my opinion, the following scenario would be very interesting:AMD and Nvidia should merge - not one buying the other but merge.What would happen. They could speed up their products development cycle and they would be competitive on both markets - x86 + GPU and ARM + GPU.Intel does have one very big weakness, and that is its GPU. Both AMD and Nvidia are miles ahead on this field and they should start using it in their advantage.They would both also benefit in GPU market - even if that is not visible right now. In a keynote John Carmack said, that the software layers between OS and GPU hardware has become so complex and large, that it actually slows down the software by a significant margin - so if there were no more battle between AMD and Nvidia, they could let the developers direct access to hardware and left DirectX and other libraries go to the history dump - that would see a real graphic performance boost in x86 and ARM playground an it would left Intel exposed!Best regards. France[/citation]
AMD was once going to merge with nvidia b4 acquiring ATI, they are well partner on the Athlon64 + nforce days......
AMD was once going to merge with nvidia b4 acquiring ATI, they are well partner on the Athlon64 + nforce days......
eddieroolz
Splendid
- Sep 6, 2008
- 7,057
- 0
- 25,860
In terms of performance per watt, it's hard to beat ARM especially in the mobile sector. I fear Intel's foray might be too little, too late.
someoneelse
Distinguished
- Feb 20, 2009
- 126
- 0
- 18,680
Intel can deal with arm any time it wants to one way or the other.
Intel market cap 2010 $128 billion
Intel net income 2010 $11 billion
ARM market cap 2010 £8 billion
ARM Net income 2010 £0.086 billion
£1 = $1.5 ( roughly)
I don't think intel is that worried.
personnally I recon the unofficial deal of we do laptops, desktops and servers and you do phones and tablets keeps both parties happy. The only business of intels Arm can move into is laptops and netbooks. I think intel are just preparing for a retaliatory strike should arm try and muscle in on their turf.
Intel market cap 2010 $128 billion
Intel net income 2010 $11 billion
ARM market cap 2010 £8 billion
ARM Net income 2010 £0.086 billion
£1 = $1.5 ( roughly)
I don't think intel is that worried.
personnally I recon the unofficial deal of we do laptops, desktops and servers and you do phones and tablets keeps both parties happy. The only business of intels Arm can move into is laptops and netbooks. I think intel are just preparing for a retaliatory strike should arm try and muscle in on their turf.
- Status
TRENDING THREADS
-
-
-
-
Question Display no signal when 1440P or higher full screen gaming- Started by Wouter_U
- Replies: 3
-
Question Download speed is slow on only my PC - please help!- Started by Ograsmium
- Replies: 14
Latest posts
-
Question Buying a used card, what should I look for before purchase?- Latest: jordanbuilds1
-
-
-
-
-
Facebook
Twitter
Instagram