cmdrdredd
Lifer
yet millions upon millions of PS4s and Xbox1's will have x86
-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
Will x86 be here "forever"?
Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
yet millions upon millions of PS4s and Xbox1's will have x86
Arkaign
Lifer
Probably 100M+ easily just in North America.
Nothingness
Diamond Member
That's less than the number of phones sold in a quarter. What happened is that Power just died in the consumer market, we're now down to two contenders.
cmdrdredd
Lifer
Not a fair comparison because people have phones marketed to them as if theirs sucks after 6 months and they HAVE to have the new one. So of course people upgrade because they feel compelled to do so. Also by comparison phones are at throw away prices. You can get GS3s for $50 now. That's throw away money in the technology world.
For the record there will always be a segment of the market (gamers, high power server environments, rendering and research labs etc) for whom ARM power is too weak. Intel(and AMD as well) is constantly increasing performance of it's desktop CPUs and no ARM CPU comes close to the type of performance we need for gaming, video editing in a timely manner, and other professional work. There isn't a phone or tablet out there that would suffice for certain tasks. The performance is not there. When they increase performance, desktop also increases performance and remains ahead.
Last edited:
Nothingness
Diamond Member
I certainly agree with you, I would not use an ARM CPU on my desktop system. But for the rest an ARM CPU is enough and x86 would not bring me anything at all.
ShintaiDK
Lifer
ARM is not good enough. Thats why you see the rapid replacement in phones and tablets when even miniscule faster CPUs comes.
If ARM was good enough, then tablet and smartphone sales cold potentially collapse to a fraction of what it is today.
Right now ARM is riding the wave in a new market as well. But when everyone got a smartphone and tablet as well. Then sales will descrease too.
If ARM was good enough, then tablet and smartphone sales cold potentially collapse to a fraction of what it is today.
Right now ARM is riding the wave in a new market as well. But when everyone got a smartphone and tablet as well. Then sales will descrease too.
Last edited:
@galego
"The most powerful supercomputer in the world will not use x86. The goal is to build a supercomputer was 10x more powerful than the Titan supercomputer"
the article u linked says "There are a number of unknowns that may actually prevent us from building a high-density solution with them"
and the supercomputer, if it exists, will theoretically have 10x performance as Titan in 2017. at which point Intel would be at what 7nm? if haswell has TDP of 15w, whatever product in 2017 could trump ARM in both performance and power
we will see ARM creeping up in low power servers. and thats about it. even there, intel will be competing with atom. so good luck to ARM
"The most powerful supercomputer in the world will not use x86. The goal is to build a supercomputer was 10x more powerful than the Titan supercomputer"
the article u linked says "There are a number of unknowns that may actually prevent us from building a high-density solution with them"
and the supercomputer, if it exists, will theoretically have 10x performance as Titan in 2017. at which point Intel would be at what 7nm? if haswell has TDP of 15w, whatever product in 2017 could trump ARM in both performance and power
we will see ARM creeping up in low power servers. and thats about it. even there, intel will be competing with atom. so good luck to ARM
And Nvidia, Google, and others will be selling ARM consoles.
I agree. This seems most likely given what we know about the evolution of computing.
Arkaign
Lifer
HAHAHAHA
Yeah, I know you love Nvidia (sarcasm intended).
ARM is not going to challenge the traditional console for a long long time because the performance is horrible.
As far as Shield, it's going to be an absolutely staggering flop. Respectable gaming power is a constantly moving target, and will remain so for a while yet due to the future with 4K and beyond.
cmdrdredd
Lifer
LOL how is the Ouya and Shield doing for you?
Valve has a better chance competing with a mid-range PC in a box running steam than any phone hardware based gaming console does.
This is true. Eventually there will come a point where a faster CPU in your phone will mean diddly. I think they are at that point right now. For what most people use their phone for, they do not need quad core CPUs and super powerful GPU hardware. There's no software to take advantage of it. Candy crush certainly doesn't need a ton of power. There is something to be said about doing things within the OS on multiple cores and running overlays on the GPU side etc. That said, those demands are quite small compared to other tasks that people do on desktop PCs or even consoles. The problem is, people are marketed all these fancy wizbang features and what they are getting is a very slightly faster phone, with a fancier case, a bigger screen, and a camera with "moar MPs!". It doesn't help that many manufacturers drop support for a device so quickly that when the next OS version comes out you don't get to have it and you need to upgrade to have the latest. Plus people feel that they need to keep up because having the newest phone is so cool. Just watch some of the Samsung commercials with everyone oogling over that dude's GS4 in the airport or wherever and they all ask him about the fancy features etc. It's status.
What phone and tablet manufacturers need to do is work with the companies making the batteries so that they last longer. If you are watching video on your phone for a few hours the battery is eaten up way too much IMO. If they want you to do all these things with it, they need to make it last all day while using the functionality.
So having said all that the point is, the functionality of a tablet or phone doesn't necessitate huge amounts of processing power. A desktop environment does and that's where x86 obviously dominates and will continue to do so for quite some time.
Last edited:
Considering there is a lot of interest in doing ARM HPC servers, the foundation to do an competitive ARM console isn't that far off. I could see nVidia doing a Denver Android console, although that's really not happening until 2015 at the earliest. Whether they could build the tools and software to get console-like efficiency is another story.
Cerb
Elite Member
HPC is very different. HPC clusters can divide work into many small work units, and have different workloads competing at the same time, and they are typically not very latency-sensitive. So, if you can scale out to hundreds or more x86 cores, scaling out to thousands of ARM cores would be no big deal. Shared DRAM works in favor of fat x86, fat Power, BlueGene, Fujitsu's SPARC monsters, etc., but if you can work with 1-2 DIMMs/node, those dense ARM servers might have some chance (lots of less scientific work, that might managed via Python+MPI, Erlang, Hadoop, DBMS plugins, etc., would tend to work well in those scenarios).
Even with game engines able to use more threads, there's going to have a massive change in the OSes and drivers to be able to get up to 20+ threads, much less hundreds, and those threads are not going to be useful if they are each really slow.
If nVidia's Denver amounts to anything, it won't be like the slow-but-efficient HPC/"cloud"/whatever stuff coming up by every semi vendor and their dogs and their fleas.
It's a dead horse is what that is. Those tools already exist. Tools to make it easier may come about, and would be nice for a proprietary hardware's SDK, but console efficiency is basically just allowing programmers to not have to worry about compatibility with other hardware. That comes from having a fixed platform, more than anything else.
A prototype will be ready this month
https://www.bsc.es/about-bsc/press/...percomputer-combine-arm-cpus-gpu-accelerators
Sarcasm is tolerated when the main body of the post is interesting. This is not the case.
The goal of current ARM consoles is not to replace x86 consoles in raw performance. The market is different.
However, Sony and Microsoft considered ARM before taking a decision about the hardware in their next consoles. Both agreed on that ARM will surpass x86 in raw performance, but not today. That is the reason why Sony and Microsoft selected x86 today.
Ouya and Shield do not count for evaluating ARM. One is a kickstarter project, the other a Nvidia product that shows the company desperation
http://www.dailyfinance.com/2013/06/21/nvidia-drops-the-price-of-shield-50-to-299/
http://semiaccurate.com/2013/02/18/nvidias-telegraphs-tegras-woes-at-ces/
Moreover, it would be ridiculous to compare a $99 console to a $399 one and believe that the former will win in raw performance...
Again. The goal of current ARM consoles is not replace x86 consoles in raw performance. The market is different.
Future ARM consoles will compete with x86 ones in raw performance. Even Sony and Microsoft agree on this.
Last edited:
Arkaign
Lifer
It would really depend on the market not moving, in combination with an ARM-based design that targeted competitive performance. Even with the efficiencies of ARM's simplistic designs, getting the performance to challenge hardware running on a dramatically higher TDP scale is sort of a pipe dream. Basically an ARM-based design that would truly challenge that level of performance would in fact, BE a larger, more power hungry design that was connected to a hugely more robust GPU than any ARM design has ever featured. Even the $349 (?) Shield is pretty much utter garbage compared to an Xbox 360 from 2005.
Arkaign
Lifer
How fast is the fastest ARM cpu you know of? And how much faster is a 9590 or 4770 or 3960? Hell, how much faster is an i3 or A10? INSANELY faster. You can argue that ARM is efficient on a per-transistor basis and be quite correct for well matched software, but the fact of the matter is that they are simple, basic, low-performing processors compared to full fledged x86 processors. They simply have entirely different design philosophies.
ARM only exists today because it snuck into a market that the big guns were all ignoring, and because it was a stripped down product aimed at performing extremely simple workloads. As people begin to expect more and more from their mobile devices, the performance target will increase. Intel x86 products at 10nm and beyond, probably still x86 based, will give ludicrous performance at a power usage that makes everything today look like a coal-fired power plant. If the ARM group could match Intel $ for $ in a unified way, and simultaneously keep up with fabrication technology, then I don't think Intel would crack the market like they will. But with those concrete and massive advantages, it's just a matter of time.
ARM will continue to exist, and future iterations will continue to improve. However, the R&D and fab $ isn't there to keep up with Intel at the top sectors for very long at all. This will mean ARM will be the basis of perhaps even the majority of Smartphones and entry-level tablets, while the premium market will demand the best, and that simply won't be ARM products. Cut to 2016, you'll almost certainly see all of the flagship phones/tablets with Intel inside. Meanwhile, you'll see decent cheap smartphones in the $49-$99 range with ARM chips that would be considered very good today. 🙂
cmdrdredd
Lifer
No...for every step arm takes, you think Intel sits on their ass?
A Cortex-A9 has about the same IPC as a Pentium 4. Cortex-A15 will likely be faster, but still far below even Bulldozer's IPC. Any "ARM-based" supercomputer will undoubtedly be using GPUs for computation, not the CPUs.
This whole ARM craziness is stemming from the simple fact:
Somewhere between late 90s and today, PC platform has drastically expanded as non-technical people started using PCs for online activities.
Thing is, these people never actually needed the whole PC. However, the technology of early 2000s simply did not offer anything better - ARMs at that point were simply not powerful enough to run "home PC" tasks.
Then, everything changed with powerful tablets such as iPad - and masses of "ordinary people" are simply moving to these devices, as they never needed anything else to begin with.
So, now everybody is in the panic 'PC is dead' mode - which is simply not true, PC is just going to scale back to its pre-2000s days as many "ordinary people" found a better gadget for their daily does of YouTube and Facebook.
ARM found a great market in this - since ARM was, from the day one, producing chips with extremely low idle-power requirements.
However, I am yet to see ARM competing in high-end desktop or server market. I will believe it when I see it, as all of this hype today has no actual substance, but it is just based on vapourware.
Intel is also not siting idle, and I am sure by the time of Skylake, there will be no significant difference in idle power draw of PC-class Intel CPU and mobile chips from ARM. However, on the HPC/server field, I am not seeing ARM gaining experience that fast to counter Intel's steady improvement in IPC and adding more cores.
Problem with running gazillions of small cores such as ARM on the server is that what you save in CPU TDP, you lose in having to add interconnection fabric. This might work for few applications where individual compute units do not require significant CPU load, but for anything which can use CPU cycles, having a large cache and lots of cores glued together is simply the best approach due to latencies.
This is why we are not seeing too many Atom-based servers - simply, it is not efficient for tasks that need to be broken down then to many CPUs, due to interconnection penalties. I believe I read somewhere that Google experimented with this setup years ago and found that the latencies are simply too high.
So, unless ARM figures out how to avoid this problem OR starts producing cores comparable to Ivy Bridge or Haswell Xeons, I doubt they will be able to gain significant foothold in the server market, except few specialized areas.
Somewhere between late 90s and today, PC platform has drastically expanded as non-technical people started using PCs for online activities.
Thing is, these people never actually needed the whole PC. However, the technology of early 2000s simply did not offer anything better - ARMs at that point were simply not powerful enough to run "home PC" tasks.
Then, everything changed with powerful tablets such as iPad - and masses of "ordinary people" are simply moving to these devices, as they never needed anything else to begin with.
So, now everybody is in the panic 'PC is dead' mode - which is simply not true, PC is just going to scale back to its pre-2000s days as many "ordinary people" found a better gadget for their daily does of YouTube and Facebook.
ARM found a great market in this - since ARM was, from the day one, producing chips with extremely low idle-power requirements.
However, I am yet to see ARM competing in high-end desktop or server market. I will believe it when I see it, as all of this hype today has no actual substance, but it is just based on vapourware.
Intel is also not siting idle, and I am sure by the time of Skylake, there will be no significant difference in idle power draw of PC-class Intel CPU and mobile chips from ARM. However, on the HPC/server field, I am not seeing ARM gaining experience that fast to counter Intel's steady improvement in IPC and adding more cores.
Problem with running gazillions of small cores such as ARM on the server is that what you save in CPU TDP, you lose in having to add interconnection fabric. This might work for few applications where individual compute units do not require significant CPU load, but for anything which can use CPU cycles, having a large cache and lots of cores glued together is simply the best approach due to latencies.
This is why we are not seeing too many Atom-based servers - simply, it is not efficient for tasks that need to be broken down then to many CPUs, due to interconnection penalties. I believe I read somewhere that Google experimented with this setup years ago and found that the latencies are simply too high.
So, unless ARM figures out how to avoid this problem OR starts producing cores comparable to Ivy Bridge or Haswell Xeons, I doubt they will be able to gain significant foothold in the server market, except few specialized areas.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
