Linked by Thom Holwerda on Wed 31st Oct 2012 00:46 UTC
Windows "Taiwanese computer maker Acer is putting off the launch of tablets using Microsoft's new Windows RT operating system to give itself time to see how Microsoft's own Surface tablet fares. The world's No. 4 PC vendor by shipments initially planned to roll out Windows RT tablets based on ARM chips early next year. However, the launch of Microsoft's tablet last week and the mixed reviews it has drawn has prompted Acer to wait and see until at least the second quarter of 2013." Whatever the reason, this doesn't send a very promising message about Windows RT. Or, not entirely unlikely, Acer and other OEMs just can't measure up to Surface RT.
Thread beginning with comment 540865
To view parent comment, click here.
To read all comments associated with this story, please click here.
Alfman
Member since:
2011-01-28

bassbeast,

"As far as numbers I've simply been reading up on new chips, its something I like and since I build some HTPCs and other low power units i like to know what is coming down the pipe. Here is a good overview of what we are looking at with Haswell."

That's fine, but we cannot just assume intel will make progress while ARM will remains stuck in the mud. They're both moving targets.

"So in either case it should be interesting. You think ARM will win, I think the combo of tick tock and just plain old fashioned throwing brains and money at the problem with pull Intel ahead, in any case it'll be great for us users, its a win either way."

I'm not emotionally attached to either architecture, I just haven't seen any data putting intel in the lead on this front. I don't want to continue to spout common knowledge about ARM's superior efficiency if it's not true, which is why I ask for data showing otherwise.

Do I think intel/x86 can catch up? Well, they have the most advanced fabs in the world, should that be factored out or should that be tabulated as an advantage for x86 as an architecture? It's controversial...

From a technical standpoint, the x86 architecture has some glaring disadvantages no matter what intel does. Instruction decoding is inherently complex due to instruction being inconsistent in size and formation, ranging from 1 to 15 bytes. Prefixes, variable length opcodes, etc all imply that information contained in the instruction (such as a memory address) might be found at nearly any offset from the instruction pointer. This inherently takes more transistors to parse than a RISC instruction set, which is why x86 machine code gets reconstructed in the processor to a risk instruction set that's far easier to handle. To the extent that x86 can have dedicated transistors to do this, it may not add too much latency, but it will consume more power doing so.

General purpose register constraints on the x86 imposed such a severe bottleneck that both intel and amd needed to resort to complex out of order execution pipelines to compensate. Luckily AMD64 added another 8 registers and so intel was able to drop the out-of-order execution on Atom processors to improve the power consumption. But it is an example of the crazy lengths engineers went to in keeping the legacy x86 architecture around.

Speculative branching, although not architecture specific, is used by all x86 cores, whereas fewer ARM cores have it. Not only does it need more transistors, but it implies the core will consume energy computing values that will be thrown out on branch misses. It's a tradeoff between performance and power consumption. But my guess would be that the negatives are more pronounced in multi-core setups where a bad branch on one core might actually delay good branches on other cores due to cache/memory contention. At least in theory, cores with no branch prediction should be more energy efficient that those without since 100% of the work gets used.

So in conclusion I think that intel may have some speed advantage per core, but it's going to be very touch for them to bring down power consumption while maintaining their speed advantage. Conversely, I think that ARM chips will contain more cores while consuming no more power than x86 counterparts.

Reply Parent Score: 3

bassbeast Member since:
2007-11-11

Actually I'd say the X86 instruction set is another advantage to Intel, since ARM is having to bolt MORE instructions on, such as ARM64, whereas Intel and AMD can go the other way and begin shaving instructions off. Lets face it, no software in the past 10 years is using the ancient 386 stuff anymore, Intel can drop a LOT of that for the mobile chips and just say "Its a mobile chip, its not designed for legacy, if you have legacy code run one of our desktop chips" which, again if rumor is true, not like I'm gonna try to load Windows 386 onto my netbook to find out, is what AMD has already done with Brazos and Hondo, dropping some of the legacy stuff like 3DNow and the older instructions to make it a lighter powered chip. And if you haven't tried one they did a smashing job, I have office boxes running Brazos and its got great performance and sub 18w under load, just nuts.

And while I don't honestly CARE which "side" wins, as my business is not heavily invested in mobile products, speaking strictly as an outsider and looking at the trends the combo of tick tock, the fabs, and the truly insane levels of IPC are gonna be hard to beat. Just look at the new games companies like Epic are working on for mobile, they are wanting Crysis levels of physics and graphics, you have the TV manufacturers getting ready to start pushing 4K sets, the demands in mobile are simply going nowhere but up. Can ARM possibly find a way to scale without blowing the power budget? Sure anything is possible but so far Nvidia and the other companies haven't found it. Even the ARM Holdings group has been talking about "dark silicon" for its new designs because flipping all the chip on at once would just kill your battery dead.

So in the end its gonna be those insane levels of IPC that decide this thing, and Intel is frankly so damned far ahead of the rest it isn't even funny. Again wish it weren't so, I think having AMD and ARM and maybe even MIPS and PPC in the race would be better for innovation but you look at the benches on the Atom and CULV Core chips and its just insane the amount of performance Intel is getting with less and less power. People want their mobiles to be "gamer rigs in their pockets" and that kind of performance is right up Intel's alley, that is what they do best. Again they could probably shave a good 40%+ of the old instructions off for those mobile chips and not effect any software written in the last 5 years and nobody is gonna be running legacy on a tablet anyway.

Reply Parent Score: 1

Alfman Member since:
2011-01-28

bassbeast,

Like I said, I don't have head to head benchmarks for the latest Atom vs latest ARM processors, but I don't feel like letting you get away with such blatant bias. So, with the help of some hand waving and numerical conversions, here's my best attempt at working with the data we have.


Find the most recent benchmarks comparing Atom and ARM:
http://www.anandtech.com/show/6422/samsung-chromebook-xe303-review-...
I'm very disappointed the benchmark compares Atom N570 (which is 2011Q1) with Exynos 5 (which is 2012Q3).
Never the less the benchmarks show the ARM processor beating out the Atom processor by 40-60% (atom is far worse with graphics benchmarks, but that's out of scope).

http://en.wikipedia.org/wiki/Exynos
http://ark.intel.com/products/55637/Intel-Atom-Processor-N570-1M-Ca...


Now we can attempt to get a performance ratio between the 2011 atom chip in the benchmark and the very latest atom processors in intel's database:

http://ark.intel.com/search/advanced?FamilyText=Intel%C2%AE...

Exynos5 dual core 1.7GHz 2012Q3
Atom N570 dual core 1.66GHz 1MB cache 8.5W 2011Q1
vs
Atom D2550 dual core 1.86GHz 1MB cache 10W 2012Q1
Atom D2560 dual core 2GHz 1MB cache 10W 2012Q4

Find benchmarks for the x86 CPUs (these benchmarks are linear)
http://www.cpubenchmark.net/cpu_list.php

N570 - 639
vs
D2550 - 689 (7.8% faster than N570)
D2560 - n/a, lets approximate by multiplying the clock ratio -> 741 (16.0% faster than N570)


Now, with a bit of handwaving, we can say that even the latest Atom processor would still be beaten by the latest Exynos 5 ARM processor by 20-40%.

"So in the end its gonna be those insane levels of IPC that decide this thing, and Intel is frankly so damned far ahead of the rest it isn't even funny."

I'm a bit surprised myself actually, I expected intel to pull out ahead slightly. But even if you'd like to attribute the difference to benchmark conversion anomolies, I think it shows that you are wrong to *assume* that intel has such a huge lead. Note that the ARM had better performance even though the Atom's clock rate is actually faster, so where ever you heard those claims about Atom's superior IPC, it's probably baloney.

I'm willing to look at other numbers, any other numbers at all...give me *something* to back your view because from where I'm sitting it looks like pure prejudice.

Reply Parent Score: 2