Intel Still Trying to Make Us Believe Its Atom Chips Are Superior to ARM. Too Bad It’s Not True
Intel has been trying to get into mobile since a few years ago. The problem has not been performance (at the time), because Atom was much more powerful than the 400-600 Mhz ARM11 CPU’s we were having in smartphones back then. The problem has always been power consumption (cost, too, but that’s another issue).
They are now trying to convince us that Intel has finally cracked the code, and its chips are not just good enough for mobile, but actually superior. Well, I think that’s nowhere close to being true. Intel may have cracked the power consumption code for Atom, but only because they’ve kept the performance of Atom pretty much the same over the years, so they can work on making it more efficient. Working on increasing performance is in conflict with working on increasing efficiency, in general.
When a chip maker builds a chip, they have some choices. Do they keep the power consumption the same as last year’s chip, but increase the performance as much as they can? Do they keep the performance the same, but try to lower the power consumption as much as they can? Do they lower performance, but lower power consumption even more? Do they increase power consumption, but also increase the performance more? Or do they just increase the performance a little bit, and also reduce the power consumption a bit (a compromise)?
These can all be design goals for chips, and a chip maker has to make a decision for that specific chip, depending on what type of market they want to address. Since they wanted to address the mobile market, and since they already had a huge (not to be understated) advantage in performance with Atom, then they decided to just focus on power consumption. The problem is that during all this time, ARM chips have made incredible gains in performance, and now the latest Cortex A15 or latest Qualcomm custom ARM chips can easily beat Atom chips in performance, whic
Sentiment: Strong Sell
ARM: GPGPU Will Boost Capabilities, Lower Cost of Mobile Application Processors.
Highly-Parallel Stream Computing to Open New Doors for Mobile Chips, Reduce Their Complexity
[03/28/2013 07:14 PM]
by Anton Shilov
Nowadays application processors for mobile devices pack a lot of various hardware, including general-purpose processing cores, graphics cores, image processors, controllers and so on. End-users demand improvements in all aspects of usage, hence mobile chips are getting incredibly complex. With the rise of system-on-chips powered by modern graphics architectures that support GPGPU technology, the need for many special-purpose components will vanish, but the doors to new features will open.
Thanks to the fact that contemporary graphics processing units (GPUs) feature a relatively high number of comparatively simplistic processing units, it is possible to use them for highly-parallel general-purpose (GP) tasks. Such usage model is called GPGPU and it is on the rise on personal computer these days. Thanks to the fact that the latest ARM Mali T600, Imagination PowerVR 6 and Qualcomm Adreno 300 support GPGPU through OpenCL application programming interface, they can be used to make better photos of videos. Provided that the GPUs are not too power hungry, their usage eliminates necessity to integrate special-purpose hardware, such as image processors or video stabilizers, thus, reducing cost of the chip.
“You might be able to save some cost somewhere in the SoC by cutting out a bit of hardware that you had and run it on the GPU instead. So cost reduction and an improved capability. So maybe they will cut out some of the ISP and they will do it on the GPU because the silicon is already there, it is power efficient, it's a quicker way of doing [it], you get a cost reduction, performance goes up,” said Ian Smythe, director of marketing at ARM, in an interview
Sentiment: Strong Sell
Read between the lines here folks, this is a desperate attempt to make ARM look better by trying to make Intel look worse. For the record this is how the ARM argument has changed over the years.
#1. Intel will NEVER have have a power efficient X86 processor in the mobile space.
#2. Well Intel has something now that can work, but it's two years behind.
#3. Okay, Intel has something that works in the power envelope now but everybody knows that ARM processors are superior . Intel still can't match performance.
#4. (My favorite) Oh Intel shouldn't sell these low cost processors because it will ruin their margins. (very desperate).
Are we really supposed to get our investment advice from this guy? I mean Merlot_1 pretty much spelled out the flaws in his argument. But I love how ARM fans are now preaching the gospel a bit louder as the market tries to make sense of the mixed signals coming out of the ARM camp. Quadcore smartphones that may or may not be leaking power at 28nm are what consumers are demanding? Windows RT sales are 'Really Terrible'. The dream of taking over the Wintel empire has died on the vine. The ARM64 is still not a reality. To top it off Warren East unceremoniously leaves ARMH one afternoon like a villain leaving town with the loot.
Across the board people are wondering if ARM's fabless Cinderella story has the flaws that Mr. Weaver has been preaching for years. So, ARM is in it's desperate mode of reaching the hearts and minds of nervous investors who aren't liking what they are reading on the ARMH boards. Intel on the other hand will say this.
"We have the best technology in the industry, how much of our newly added capacity will you like to sign up for?" As the money flows and flows in, the stock will go up You can't argue with success.
For the record, I took a snapshot of Intel, ARMH and Nvidia on one day last week and we'll see how right this guy is over time. Strong sell? Yeah buy ARMH! LOL!
There are a few fundimental fkaws in ARM's argument.
1. X86 vs RISC: There is no significant difference between instruction set code with respect to power consumption.
2. ARM traditionally produced designs for a mobile device using much lower core frequancies(approx 600mhz), with the resultant much lower power consunsumption.
3. By Apples genius play they realised that an ARM powered phone could fundimentally also power a limited capability large ipone without phone capability, ie a tablet. This led to the realisation that a "just good enough" device could create a niche market in the portable space.
4. Esentially, Intel in the past have had a desktop part (Atom) with far higher performance (1.2ghz) though with much higher power consumption.
5. ARM's A15 (28nm Process) supposidly delivers ATOM like performance at moderate power consumption (1.9W).
6.Intel's latest Clovertrail+ Atom (32nm Process) performance at moderate power consumption (1.7W).
Simply look at the facts
- ATOM Clovertrail+ comparable in power consumption to ARM A15 even though Intel is using 32nm 5 year old process to produce the product. Add in capability to ONLY run RISC coded applications.
- ATOM Clovertrail+ superior in performance to ARM A15 even though Intel is using 32nm 5 year old process to produce the product.
- ATOM next generation will feature NEW architecture + shrink in process node to 22nm. Add in capability to also run x86 software, with most if not all RISC applications also enabled.
ARM camp has to face up to the facts that in a maturing (tablet/Smartphone) market the notion of "just good enough" will not be good enough.Performance and feature will prevail. As history has shown us time and time before.
" this is how the ARM argument has changed over the years."
The INtel argument has always been the same... for years & years...
INtel is coming out with something next year... the competition is thru...
Yet ARM revenues keep growing... and adoption is increasing...
" As the money flows and flows in"
The money has been flowing into ARM R&D by the world's companies...
latest benchmarks Tegra 4 vs Z2760... well how can that be... and Tegra 4 is only 80 sq mm...
Well... INtel is coming out with something next year...
Somehow, you think it's fair game to compare what INtel will have next year
against the current competitors products... Standard MO for this board...
"ARM is in it's desperate mode of reaching the hearts and minds of nervous investors "
put down the pipe...
ARMH is near its 52-week high...
INTC is near its 52-week low...
The author conveniently omits several critical points; 1). As the smartphone and tablet markets have matured the demand for greater performance has increased beyond ARM's original power efficient design, 2). As ARM attempts to match performance of Intel chips ARM's power efficiency declines. ARM chips are now made that are too power hungry for use in smartphones, and 3). For the tablet market and above vendors don't just choose between ARM and Intel's Atom, they choose between ARM and the entire x86 product line including the much faster Core architecture that scales all the way to supercomputers.
And the last line about A15 performance advantage against Atom offers zero data to back up the claim. It's amusing how ARM supporters now resort to comparing its chips using the power envelope of a tablet against Atom with a power envelope of a smartphone. Vendors won't be tricked by this.
No one believes Warren East's rhetoric about ARM power efficiency superiority any more.