The x86 Power Myth Busted: In-Depth Clover Trail Power Analysis
by Anand Lal Shimpi on December 24, 2012 5:00 PM ESTSunSpider 0.9.1
Now the fun stuff. Doing power profiles of our standard benchmarks gives us good insight into how well each vendor was able to balance peak performance and average power. In general it's ok to burn more power for a short amount of time as long as it means you'll get to sleep quicker. This was one of the fundamentals of the first transition to mobile from the early 2000s.
We already know that Intel completes SunSpider quicker thanks to its improved memory subsystem over the Cortex A9, but it also does so with much better average power (3.70W vs. 4.77W for this chart). A big part of the average power savings comes courtesy of what happens at the very tail end of this graph where the W510 is able to race to sleep quicker, and thus saves a good deal of power.
| JavaScript Performance | |||||||
| Time in ms (Lower is Better) | Kraken | SunSpider | RIA Bench Focus | ||||
| Acer W510 (Atom Z2760 1.8GHz) | 33220.9ms | 730.8ms | 3959ms | ||||
| Microsoft Surface (Tegra 3 1.3GHz) | 49595.5ms | 981.1ms | 5880ms | ||||
| Samsung ATIV Smart PC (Atom Z2760 1.8GHz) | 33406.0ms | 721.3ms | 3752ms | ||||
| Apple iPad 4 (A6X) | 19086.9ms | 834.7ms | - | ||||
| Google Nexus 10 (Exynos 5 Dual) | 11146.0ms | 1384.1ms | - | ||||
I also used SunSpider as an opportunity to validate the results from Intel's tablets with my own review samples. To generate this chart I measured power, every second, at the wall with both devices plugged in and with a fully charged battery. The resulting power consumption numbers include the efficiency loss at the AC adapter but the general curve shoud mimic the results above:
Note that the results do generally line up, although measuring at the battery gives more accurate results for the device and using the NI DAQ I was able to get better granularity on the power measurements.
Looking at CPU level power consumption we see a very even match between Atom and Tegra 3. Intel's advantage really comes from being able to complete the workload quicker (0.52W compared to 0.72W on average).
Once again we see a pretty significant difference in power consumption on the GPU rail between these two platforms.
Kraken
Mozilla's Kraken benchmark is a new addition to our js performance suite, and it's a beast. The test runs for much longer than SunSpider, but largely tells a similar story:
RIABench
RIABench's Focus Tests are on the other end of the spectrum, and take a matter of seconds to complete. What we get in turn is a more granular look at power consumption:
Facebook
Twitter
RSS
163 Comments
View All Comments
dc77gti - Tuesday, December 25, 2012 - link
Looking forward to 22nm Bay Trail-T. Hopefully Intel can get this out before 2014.That's half the battle. The other half lies with Google Android. We'll have to wait and see. With Intel's hardware might and Android's open source project, things will get more exciting.
kyuu - Wednesday, December 26, 2012 - link
*sigh* Everyone seems to have forgotten about AMD's Temash which should be out by mid-2013... hopefully. Hell, I'd even take a Hondo tablet over Clover Trail, if anyone actually made one.thebeastie - Tuesday, December 25, 2012 - link
I don't know about this.. all of a sudden the tables have turned against ARM?Mythbusters do a pretty thorough job on testing, I want them to do the same round of testing on these chips and see what THEY come up with.
Apples A6 still reams the Atom so its not really that great.
Anyway ARMs real kryptonite has been price. Intel might sell thee chips cheap as a last ditch stand one off or two off but they can't do it forever before what matters to them most falls apart.... and thats making a lot of money.
puppies - Wednesday, December 26, 2012 - link
I think most of us realise that intels biggest problem right now is that in the desktop environment (their main source of income from the public) a 3 or 4 year old chip is more than adequate for 99% of tasks that the average pc user wants to perform. Software just isn't being developed (or is there much need for it) that requires a quad core 3.5ghz cpu with turbo and HT when a 2.5ghz dual core will more than suffice.If however they can push the performance envelope of these ultra low voltage parts to a point where software starts being developed that can utilise those chips to their potential then ARM will not be a viable option for anyone who needs that performance.
Most tablet reviews state something along the lines of "it is ok for a few last minute corrections to a presentation but you might want something more powerful for when you aren't on the train/plain etc". If intel changes that to "This ultra portable tablet has enough grunt for all your word/powerpoint/excel creation needs" and ARM can't keep up then Intel becomes the required CPU for business users, intel really doesn't care about $200 tablet sales. They are generally bought as presents for kids and there just isn't the profit available that intel seems to desire.
Combine this with the fact that no company is going to want to deal with the headache of trying to sync workloads between X86 office pcs and ARM based ultra portables and intel suddenly has a reason to charge the big bucks again.
FunBunny2 - Friday, December 28, 2012 - link
I expect the reality is: for consumers, the need for anything much more than a 486 is, well, past. Not much computation outside of Excel. Pretty pixels, on the other hand...If M$ could write a yet more bloated OS, then the old Wintel symbiotic monopoly might return. Fact is, we're still where Xerox PARC put us 3 decades ago. The hardware isn't much different, save for touch, either. There was a time when PCs shipped with monochrome tubes, by default. And the OS was a command line ark.
war59312 - Wednesday, December 26, 2012 - link
Hey,This image as linked on page 5, the very last image appears to be broken:
http://images.anandtech.com/reviews/SoC/Intel/CTvT...
Thanks,
Will
Veteranv2 - Wednesday, December 26, 2012 - link
Reading this review, it makes me realize how websites are abused as a marketing gimmick.Keep this in mind:
- It is all teste on Win8.
- Win8 is primarely Win7 but with ARM support
- Windows has been optimized for X86 in any way, they just recently added support.
- Who says ARM support on Win8 is any good for ARM performance? It is compatible, but x86 has enjoyed +20 years of optimization
If anandtech would have wanted to do this right they would have used this:
- A6X or A15 ARMcores
What has anandtech proven:
- Win8 is bad for ARM
- Tegra3 on 40nm has worse power consumtion then a 32nm part
- It is a marketing tool for intel which is struggling in the tablet market and needs positive things like this
- That it cannot objectively make differences clear between architectures, cause this review has nothing to do with architectures...
A sad day for anandtech...
thebeastie - Wednesday, December 26, 2012 - link
I guess I could agree with this, Anandtech is my absolute favorite tech site for the truth but some times he just seems to be a little bit too much of an Intel fan. But alternatively I do see that he is ready to hand out credit where it is due, and Intel is all too often the company to beat.Ultimately I think his heading was a poor choice of words and is just as much appeared to be skewed towards headline grabbing as much has being an accurately balanced review.
I guess that will always be part of the game with "the press"
CeriseCogburn - Friday, January 25, 2013 - link
You guys missed the bucket of fudge, and forgot the nvidia hate.Thus the Tegra3. you've missed the bias boat, bros.
Anand got screwed on nVidia gpu's a long while back, and they AND Tom's have never forgotten the slight.
It was the (made notorious) "rebranded" nVidia 9800x, where Anand just reused another of the nVidia gpu's they had in house and adjusted the clocks and claimed they thus tested the "new" nVidia release.
The deep hatred has been seething here ever since, in every article, pumping up the amd fanboys, and only recently has that latter group somewhat receded, due to continual epic fails by amd.
powerarmour - Friday, December 28, 2012 - link
Agreed