Who Controls the User Experience? AMD’s Carrizo Thoroughly Tested
by Ian Cutress on February 4, 2016 8:00 AM ESTPower Consumption: Big Improvements to Video Playback
It was teased earlier in the review, but it makes sense at this stage to talk about power consumption.
With a system as complex as a modern APU or SoC, the initial plans for this review involved getting a development system with the right shunts and hooks to measure the core and graphics power separately in a thermally unconstrained environment for both Kaveri and Carrizo, but unfortunately the parts didn’t come together at the time they were needed. Instead we had access to a Watts Up PRO, a power outlet based monitor with some recording capabilities. While the hardware was not ideal for what we wanted to test, it provided a large chunk of interesting data.
We did a number of tests with data monitoring enabled on both the HP Elitebooks. When AMD released Carrizo, a lot of fuss was made about video playback for several reasons. Firstly, Carrizo implements an adjusted playback pathway for data so instead of moving data from the decoder to the GPU to the display controller, it moves data directly from decoder to display, saving power in the process.
AMD also listed the video playback power of Carrizo (as compared to Kaveri) as significantly reduced. In the example above in the top right, the Kaveri APU is consuming nearly 5W, whereas Carrizo will consume only 1.9W for 1080p content.
The other video playback optimization in Carrizo is the Unified Video Decoder. The bandwidth and capability of the UVD is increased four-fold, allowing the system to ‘sleep’ between completed frames, saving power.
Video Playback, 1080p30 h264
For the first test, we took a 1920x1080 resolution h264 video at 30 FPS (specifically Big Buck Bunny) and recorded the power consumption for playback.
The difference here is striking. The Carrizo system in this instance has sustained power consumption lower than that of the Kaveri system. Overall the Kaveri system draws 11W over idle to play back our test video while the Carrizo system only draws 6.8W over idle for the same task. Put another way, the load power cost at the wall for watching 1080p video is about 4W lower on Carrizo as compared to Kaveri, which is close to what AMD claimed in the first slide above (and note we’re measuring at the wall, so chances are there are other chipset optimizations being done under the hood).
Video Playback, 2160p30 h264
The same video but in 4K format was also tested on both systems. It is at this point I should say that the Kaveri system was unable to play the 4K video properly (Kaveri doesn't officially support 4K decoding to begin with), and would only show about 20% of the frames. Audio was also affected.
In this case power consumption is above that of the 1080p video, and both systems require around 11 watts from idle to sustained performance. The added benefit with Carrizo though is that you can actually watch the video.
Other Power Benchmarks
We also ran power tests on a set of our regular benchmarks to see the results.
Three-Dimensional Particle Movement
In our 3DPM test, we typically script up a batch of six runs and take the average score. For this we did it to the single thread and multithreaded environments.
In single threaded mode, two interesting things occurred. First, as we expected, the Carrizo system can idle lower than the Kaveri. Second is that the Carrizo system actually goes into a higher power state at load by almost 4W. This means that the delta (Load to Idle) is 8W higher for Carrizo than Kaveri.
It is easy to take away from this that Carrizo, as an APU, uses more power. But that is not what is happening. Carrizo, unlike Kaveri, integrates the chipset onto the same die as the APU (better integration, saves power), but it also means that it is essentially shut off at idle. Part of Carrizo’s optimizations is power management, so the ability to shut something down and fire it back up again gives a larger low-to-high delta automatically. Essentially, more things are turning on. The fact that the Carrizo power numbers are higher than Kaveri during the benchmark is correlated by the performance, despite Kaveri having the higher TDP.
For the multithreaded test, both systems settle to similar power consumptions as the single threaded test, although the Carrizo system has a much more varied power profile, which also finishes the benchmark earlier than the Kaveri.
Octane and Kraken
For the web tests, we expect them to be partially threaded but because they probe a number of real-world and synthetic tests, there should be some power variation.
Octane is actually relatively flat, instigating similar power profiles to both. Again, it looks like that Carrizo expends more energy to do the same amount of work, however it is easy to forget that the Carrizo idle power state is lower due to optimizations.
With Kraken we also get a flat profile, although one could argue that we’re seeing a classic case of running quick and finishing the benchmark sooner vs. a more sedate path.
WebXPRT
This graph was shown earlier in the review, but let’s look at it again, as it is a good example of a bursty workload:
With average power numbers only a few watts above the idle numbers, both systems do a good job on overall power though again it is easy to think that the larger delta of the Carrizo numbers means that the APU is consuming more power. This is where if you try and calculate the actual energy consumed for each system, you get stupid numbers: 1208.7 joules for the Kaveri and 1932.8 joules for the Carrizo. Without starting from the same platform (or without taking numbers direct from the cores), there are obviously other things at play (such as Carrizo’s capability to control more power planes).
WinRAR
Our final power test is WinRAR, which is characterized as a variable threaded load involving lots of little compressible web files and a handful of uncompressible videos.
In this instance I was surprised to see both systems perform similarly. The HP Elitebook G2 actually has the upper hand here, as it is equipped with dual channel memory. WinRAR is a very memory bandwidth affected benchmark, so the G2 has an upper hand in performance but will also balance between drawing more power for two modules or running in a more efficient mode if there is sufficient data at the CPU.
Facebook
Twitter
RSS
175 Comments
View All Comments
karakarga - Friday, February 5, 2016 - link
Including all, AMD and nVidia both at their funeral state! They can not possibly open 22, 14, 10 etc. micron fabric.Intel spended 5 billion dollars to open their new Arizona factory, they will pass lower processes there as well. AMD and nVidia can not get, even a billion dollar profit in these years. It is impossible for them to spend that much money to a new low process factory.
Those little tweaks can not help them to survive....
testbug00 - Friday, February 5, 2016 - link
They don't build factories. TSMC and Samsung (and GloFo to a lesser extent) build factories and do R&D for these processes. Nvidia, AMD. Samsung, Qualcomm, MediaTek and many other companies design chips to the standards of TSMC/Samsung/GloFo and pay money for wafers and running the wafers through the fab.The cost for this per wafer is meant to get all that money back in a few years. And than the process keeps on running for over 10 years sometimes.
It is getting more expensive to get to smaller nodes and the performance increase and power decrease is getting smaller. And costs more to design chips and run wafers. So it is getting harder to find the funds to shrink. Which is one of the reasons Intel has delayed their 10nm process.
yannigr2 - Friday, February 5, 2016 - link
Thanks for this review. Really needed for sometime. It was missing from the internet, not just Anandtech.As for the laptops, they say as much as there is to tell. Small Chinese makers, who no one knows they exist, would built better laptops than these. HP, Toshiba and Lenovo in this case, multibillion international giants that seems have all the technicians and the R&D funds necessary, end up producing Laptops with "strange" limitations, bad choices, low quality parts and in the end put prices that, even with all those bad choices and limitations, are NOT lower than those on Intel alternatives. It's almost as if Intel makes the choices for the parts in those laptops. Maybe their is a "trololol" sticker on them somewhere hidden addressed to AMD. I guess that way those big OEM don't make Intel too angry and at the same time, if there is another legal battle between AMD and Intel in the future, they will have enough excuses to show to the judge in their defense, if accused that they supported a monopoly.
ToTTenTranz - Friday, February 5, 2016 - link
This article is what makes Anandtech great. Just keep being like this guys, your work is awesome!I'm going to spend some time clicking your ads, you deserve it :)
As for the "poll" about who's to blame, IMHO it is:
1 - AMD for letting OEMs place Carrizo in designs with terrible panels and single-channel solutions. It's just not good for the brand. "You can't put a Carrizo with single-channel cheap RAM because that's not how it was designed. You want to build bottom-of-the-barrel laptop? We have Carrizo-L for you."
I'm pretty sure Intel has this conversation regarding Core M and Atom/Pentium/Celeron solutions. I know AMD is in a worse solution to negotiate, but downplaying Carrizo like this isn't good for anyone but Intel.
In the end, what AMD needs is a guy who can properly sell their product. Someone who convince the OEMs that good SoCs need to be paired with decent everything-else.
$500 is plenty for a 12/13" IPS/VA screen (even if it's 720/800p), 128GB SSD and 4+4GB DDR3L. Why not pull a Microsoft's Surface and build a decent SKU for that price range so that other OEMs can follow? Contract one OEM to make the device they envisioned, sell it and see all others following suit.
2 - OEMs for apparently not having this ONE guy who calls the shots and knows that selling a crappy system automatically means losing customers. And this ONE other guy (or the same) for not knowing that constantly favoring Intel with their solutions is bound to make the whole company's life miserable if Intel's only competitor kicks the bucket. The consumer isn't meant to know these things, but the OEMs certainly are.
It's 2016. We're way past the age of tricking the customer to buy a terrible user experience through big numbers (like "1TB drive woot"). He/She will feel like the money just wasn't and next time will buy a mac.
Want a $300-400 price point? Get a Carrizo-L with a 128GB SSD and a 720p IPS panel. Want $500-700 Price point? Get a Carrizo with dual-channel, 256GB SSD and 900p/1080p IPS screen.
joex4444 - Friday, February 5, 2016 - link
Anything under 1080p is simply not usable. All these 1366x768 panels are just awful. I have an old netbook with one (12.1") and I've put a small SSD in there and loaded it with Ubuntu. I cannot have a Google Hangouts window open and a web browser open wide enough to view most pages. Basic web browsing + IM - 1366x768 completely fails at the task.testbug00 - Friday, February 5, 2016 - link
768p panels are fine if they are good quality, in 11" laptops.900p good up to 13", and 1080p minimum for 14+.
Honestly I wish we stayed with 8:5 14x9, 16x10, 19x12z
jabber - Saturday, February 6, 2016 - link
Indeed, 768p is fine on my 11" Samsung Chromebook but I would not tolerate it on anything bigger. IMO 1600x900 should be the minimum screen res for budget machines. 1080p for midrange and whatever you like for higher end.jjpcat@hotmail.com - Monday, February 8, 2016 - link
Resolution is not as important as the quality of the panel. I used a Lenovo X1 Carbon. It has a 14" 1080p screen. But it's a TN panel and that just makes it a pain in the ass. I am amazed that Lenovo uses such a lousy panel in its $1k+ laptop while some 10" sub-$200 tablets use IPS.testbug00 - Friday, February 5, 2016 - link
Toshiba can make a $400 chromebook with a good 1080p display. Fully agreed.1080p panel, make it thicker so you can put a larger battery and so the laptop can handle up to 35W from the APU. Do dual channel.
When plugged change APU power mad to 35W, when in battery make it 15W. Probably can be done for $500 for a 15" laptop with an A8. $50/100 upgrade to 128/256GB SSD and $50/100 upgrade to A10/FX.
Dobson123 - Friday, February 5, 2016 - link
"The APU contains integrated ‘R6’ level graphics based on GCN 1.0, for 384 streaming processors at a frequency of 533 MHz."Isn't it GCN 1.1?