Ainol's New Choice for a Quad-Core, the "ACT-ACT7029" - Just a Marketing Move?
Ainol is known for marketing well in the US (for a Chinese Brand) and for completely dropping product lines after 3 months of production and for having incredibly poor quality control for new products. So basically they go for flash and not for quality... or support, hence the US site dropped them as a brand several months back. As they change products pretty much every 3-months, they or less never get a product completely "right". However, because they do tend to jump on platforms a little earlier than most, their products are often good "beta test" indicators of what is to come.
Enter their latest lineup of devices built around the rather unknown Quad-Core "ACT-ACT7029" processor/gpu SoC. With this launch they are also going to attempt to break into the 8", 9.7" AND 10.1" markets. Up until now, Ainol has only produced 7" tablets.
The Ainol Lineup is as follows:
-Novo 7 Flame II: 7 Inch, IPS, 1280*800 pixel, Daul Camera 5M Back and 2M Front,16GB,Bluetooth,5000 mAh battery
-Venus: 7 Inch,IPS,1280*800 pixel,Daul Camera 2M Back and 0.3M Front,8GB,GPS,Bluetooth,4000 mAh batter
-Dragon: 7 Inch,800*480 pixel,8GB,HDMI,Front Camera,3700 mAh battery
-Legend: 7 Inch,800*600 pixel,8GB,HDMI,Front Camera,3000 mAh battery
-Novo8 Dream:8 Inch, IPS,1024*768 pixel,Daul Camera 2M Back and 0.3M Front,Bluetooth,GPS,6000 mAh battery
-10 Captain:10.1 Inch,IPS,1920*1200 pixel,Front Camera 2Mega,GPS,8GB, Bluetooth, GPS, 10000 mAh battery
-Hero II:10.1 Inch,IPS,1280*800 pixel,Front Camera 2Mega,GPS, 16GB,Bluetooth,GPS,8000 mAh battery
In spite of my personal distaste for all things Ainol, they do have large customer base so it is worth keeping an eye on which direction they are moving as many other manufacturer's often follow suit.
On the face of things, the ACT-ACT7029 is an exciting processor as it is one of only a small handful of true "QUAD CORE" China-Brand processor. The chipset is so new that their aren't any performance benchmarks so we can really only speculate based on what little we know. Taking a deeper look at the processor however doesn't seem to reveal a new "performance leader" but simply more of the same mediocrity we have already seen in the Chinese quad-core space.
Low Per-Core Clockspeed
The quad-core variants of most dual-core A9 chips almost always have a lower per-core clock speed, heat most likely being the main issue. The RK3066 is one of the faster clocked Dual-Core A9 chipsets at 1.6 Ghz. We are expecting the ACT7029 to probably be around 1.0 - 1.2 Ghz per core. Ainol, of course, will more than likely market the chip as running at 4.8 Ghz... watch for it
. There are many people today who still swear up and down that the A10 runs at 1.5 Ghz which was pretty much an absolute lie unless you happened to have some liquid nitrogen on hand. Chinese companies tend to view "marketing" and "exaggerating" as one and the same thing... Regardless, we have seen pretty good performance figures for Quads in spite of lower clocks, especially in synthetic benchmarks that are made to take advantage of parallel processing. I still argue that for real-world use, "per core" performance is much more important than the performance of the whole when it comes to the processor in particular. Most games are good at taking advantage of extra GPU cores but most apps aren't good at taking advantages of extra processor cores. Hence, Apple's new A6 chip even is a dual-core, because it's what most apps are optimized for at present. Quad-Core IS a good idea, but for the time being, it isn't worth sacrificing the performance of an individual core on its own.
After some further "sleuthing" I discovered that Ainol is claiming this chip to be clocked at 1.5 Ghz... I won't believe it until I see it though as they claimed the same thing about the A10 which only ever ran at 1.0 Ghz stable on any tablet Ainol ever sold and 1.2 Ghz was the quickest it was clocked on anything beyond that.
The 10.1" Captain has a very high-res screen and GPS and a very big battery. And will come in right around were our current 10.1" devices sit pricewise. The SoC, therefore, must be fairly low cost to have all those bells in whistles in the tablet and still maintain such a low price. Also, the creators of the ACT7029 made a rather odd choice for the GPU which leads us to further believe it is a low cost part.
This really is a guess based on Ainol's device history. Generally speaking, Ainol rarely skimps on battery. It is one positive thing that can consistently be said of them. However their unit on the low end is packing only a 3000 mAh battery. If Ainol is being consistent, that device should still have fairly decent run-time which makes me guess that the ACT7029 may be fairly good at sipping power. They might just be trying to hit a price-point though in which case this observation is irrelevant.
I am very much so hoping this is a typo on Ainol's part. According to the specs for there new tablets, the ACT7029 is packing a Vivante GC1000 GPU. There are two things seemingly a bit wrong with this.
In the world of Android, most developers only really program for "mainstream" devices and they don't take China Tabs into consideration. For a long time this meant major application compatibility issues for some platforms. Although things have gotten drastically better, the rule of thumb for buying a tablet is that "if it isn't a mainstream device, it should at least be somewhat LIKE a mainstream device."
This leads to the problem of Vivante. Most Mainstream Android tablets use an Nvidia, Samsung, or Texas Instrument processors (in that order). Nvidia does its own proprietary thing with the Tegra GPU and has very good support. Texas Instrument mobile processors pretty much always use PowerVR GPU's (which is what Apple uses) and Samsung for a long time did the same, howeve the recently switched to the much more open ARM Mali-400 GPU for their Exynos 4XXX SoC. Guess who wasn't mentioned... Vivante... While I am all for more competition in the mobile space, the hard fact is that you are going to have more problems with a Vivante chipset than you are with the Mali-400(which is used most notably in the RK3066). Oh... the PowerVR party piece is that Apple uses PowerVR GPU's..so games that were originally developed for the iPad 1/2/3 all tend to work a bit better on Android devices using PowerVR GPUs. While interesting, that point is oustide of the scope of this article. So compatibility is issue one.
Issue two is the fact the ACT7029 is using a rather vanilla Vivante GPU. The Ainol website states they are using the GC1000 which is a BIG step down from Vivante's quickest GPU, the GC2000. The GC1000 is a dual-core GPU vs the Quad-Core Mali-400 used in both the Samsung Exynos 4XXX and RK3066 platforms. Lets look at some raw performance specs between the two.
For those of you who care, the "Quad" core part of the Mali-400 is the Pixel Shader units. The Mali-MP4 has 4 of them but it still only has 1 Vertex Unit. So the Triangles/s on the Mali-400 are low, however the fill-rate is fantastic. Most current Android titles are more dependent on fill-rate than on Triangles/s so the Mali-400 is weighted this way by design.
Vivante GC1000 Dual
650 M Pix/s
ARM Mali-400 Quad
Ultimately it is hard to come to a solid conclusion without any real world testing. However, recent history may hold a lesson for us here... The issue is that synthetic benchmarks can look good, but then only a handful of game will work due to compatibility issues. Case-in-point... the earlier generation Rockchip RK2918 used the GC800 GPU which had great specs on paper but in real-world performance it was outclassed by the Mali-400 MP1 in the Allwinner A10. The GC1000 is certainly quicker in some areas on paper than the Mali-400 MP4, but I suspect it will be the same story all over again when it comes to real-world use. On top of that, the RK2918 had more problems playing games than the Allwinner A10 due to the fact that the A10 had a Mali-400 MP1 GPU instead of the Vivante GC800 in the RK2918 and the Mali-400 is, as stated, more compatible. It is interesting to also note that Rockchip, for it's newest SoC (the RK3066), SWITCHED from using Vivante and went with the ARM Mali-400 MP4. Why didn't they just use the GC1000 (or GC2000) with the RK3066? I think Rockchip saw the wisdom of sticking with a more mainstream GPU.
Conclusions about the ACT-ACT7029 SoC and Ainol
The jury is most certainly still out. The ACT7029 might be a winner yet if it proves to be affordable, compatible, and power efficient. The compatibility is my biggest concern. The fact that Ainol is fully embracing it give me mixed feelings. I would like to say it is a good thing, but they fully embraced the AmLogic 8726-MX chipset line which was significantly slower than the RK3066, less power efficient, and more or less the same price. Ainol was the first to jump on the Allwinner A10 processor which was a market leader for a while. So they have a mixed history. If I had to call it though, I would place my bet on the fact that they are backing a loser primarily due to the questional GPU used in the ACT7029 SoC.
I personally believe Ainol is being true to their form. They are "jumping" on a Quad-Core part with a low price so that they can MARKET a Quad-Core part. I feel they did much the same with the Amlogic 8726 MX series as it was one of the first recent Chinese dual-cores so they could push it out more quickly.