Intel Z3735D Vs. Rockchip 3288

Thread: Pipo P1 and P8 tablets with RK3288

about new tablet’s specificantions and performance in general, I’ve seen that tablets equipped with Intel BayTrail-T Z3745D (as the new Teclast X98 3G) are even FASTER (and a lot cheaper) than Pipo P1. See 3D mark results for example.

So I do not think that RK3288 is the “CPU to go” now, also because BayTail CPU support x86 and 64bit, then can be aquipped with Windows 8.1 too.

BayTrail are built at 22nm, RK3288 at 28nm. So Intel CPU’s could probably drain less power.

For the reasons above I think Intel CPUs now are better solution than RK3288. And for the bad support of Pipo, I think that switch to another brand could be good too.

I’m just going to leave this here as one of those things that make me go, Hmmmm….

Feel free to agree or rebut in Comments, however.

Advertisement

21 Comments

Filed under Other Hardware, Reference

21 responses to “Intel Z3735D Vs. Rockchip 3288

  1. I totally agree. I was waiting for the RK3288 tablets to come, but the PIPO is very dissapointing at least: 50ºC, awful battery life with a 10A battery, less performance than expected… I’ll go with the onda v975i if their kitkat release fixes the frecuency switching issue.

    • Marcus

      What frequency switching issue? You mean the tendency to go straight up to 1.8ghz when the tablet is doing something no matter what?

      My current workaround is to dynamically underclock the tablet depending on the apps using Tasker.

      • Yup that’s what I meant. That may be a valid workaround, but I prefer it using the right frequency at the right time, just like any other tablet, phone or PC does.

  2. Matyas Farkas

    Hi Mike, I would like to congratulate to your site, it is very informative and up to date! Thank you for doing all the research, organization of information and sharing.. Having read your content I felt oblidged to contribute.. :)
    I see the point of the comment , the 28nm vs 22 nm + 64 bit seems compelling. I might want to add my doubts about the 64 bit ARM v8 instruction set towards which Android L is shifting, as the only point where the 64 bit matters. I am more concerned about the actual battery life and power consumption. RK has a history of not been very optimized for battery friendly usage in their original stage, but as time has passed reserves were tapped upon..While the Intel Bay Trail processors were delivering good performance their power management was also not perfect..

    What do you think which of the two might be less power hungry under slight-med load?
    For me this is the most important factor. A tablet reasonable on screen time (8h+) is the minimum for me.
    But I am sure they are still optimizing the RK3288, before final release. I do not care much about scores, battery life, stability and percieved experience are the decisive factors- having watched those “double speed” youtube chineese videos , you realize, either the tablets are fluid and the woman who operates has got lightning fast hands or the percieved experience is not as good as it should be…
    Until futher notice I am just waiting for the first user “true” user reviews/videos!

    • mikecane

      Well, Intel must have KitKat ready because Teclast said they’d be updating the X98 3G this week. That could change the battery-life equation so it’s too soon to even guess at whether Intel or 3288 will be better with battery.

      • So many previous comments and analysis on other products from Mike have been haphazard at best — how can you completely dismiss a product or compare processors without thoroughly testing the products discussed ? – so many assumptions here are based on partial analysis or are questionable of their accuracy and are irresponsible and unfair to readers trying to make an educated decision.

        I’ll wait for other, more reliable sites before making any clear assumptions on new products (including the several new Pipo tablets discussed here)

      • mikecane

        My Comments are based on what I read directly from China.

  3. Rockchip provided a very crappy 4.2 and an excellent 4.4. Right now the Onda v975i and the Teclast X98 are giving pitiful battery performances with 4.2, so I wouldn’t be so quick to dismiss Rockchip.

    • Marcus

      Lol. What. For battery life of the V975i. Look here.
      http://www.slatedroid.com/topic/105634-battery-life-testpoll/

      • mikecane

        There are only three tests and a very wide variation on the video playback result.

        And seriously, this?

        >>Extrapolate the time it takes to drain a 100%
        >>For example, it drained 10% in an hour’s test. This means it drains 100% in 10hours.

        When Chinese reviewers pull that, readers complain. And they’re right to do so. There’s no substitute for taking the time it requires.

      • Marcus

        I think that’s why there’s like an accuracy thing next to their measurements.

      • mikecane

        My primary use for a tablet will be reading GBooks PDFs. If I can get a full day of battery doing that, I’d be satisfied. Playing games isn’t my thing so that aspect of battery life doesn’t interest me.

      • Marcus

        Those 3 tests are the standard tests. Also, battery life that are reported by the device manufacturers are always referring to video playback time. Like iPad prides itself on having 10+ hours of battery life but that refers to video playback with specific conditions. Hence these 3 tests are sufficient to provide a general idea of how well the battery life is. Reading books and PDFs? Nobody is going to provide a benchmark of the battery life of this aspect.

      • mikecane

        >>>Reading books and PDFs? Nobody is going to provide a benchmark of the battery life of this aspect.

        Of course not. I just said that was going to be my primary use.

      • In our forum Hispatablets, I’m not seeing anyone who has been able to have that screen on for more than 5 hours. People are even reporting that the shady (read: not activated) Windows version that Teclast provides in the X98 provides better runtimes than Android 4.2. The only good thing is that unlike previous models, deep sleep is working semi-properly and the battery is not draining there.

        I don’t trust the extrapolations that are being done by using that program. Using Better Battery Stats we’ve seen that there’s a background sound process messing around, and additionally we’ve seen that the processor seems to be running at full speed all day, without any intermediate steps between deep sleep and function. Besides, I’m not quite sure how Android 4.2 is optimised for Intel devices, given that they only have been taking off the last year with 4.4.

        Kitkat is really our only hope for getting better runtimes and solving the couple of issues in these models.

      • mikecane

        Damn. I saw in a forum somewhere — won’t be able to find it til later, if I find it at all — that there’s something in the tUI in the X98 3G that keeps running and unnecessarily draining the battery.

        EDIT to add (7/14/14): I finally came across it again at XDA, applies to the Teclast X98 3G, might work with Onda V975i too:
        http://forum.xda-developers.com/showpost.php?p=53749090&postcount=72

  4. shadowcek

    So whats the conclusion? I am interested in getting new tablet. The reviews for bay trail performance seem to be great so far, but what about the app compatibility with x86? I am looking forward to play Legion of Heroes on it once released live and would be really sad to find out it doesnt work on intel cpu. As for 3288, I dont want Pipo…Teclast p90HD looks better IMO, but will the GPU/CPU be able to handle that monster resolution? Thanks a lot for help in advance.

    • mikecane

      I can’t Comment on x86 game compatibility. I have no interest in games. We have yet to see what the octa-core Allwinner A80 offers.

  5. Don Juan de Marco

    Skip Pipo’s products – unless you’re an aspiring Android dev or a hobbyist trying to prove something. Their products look good physically and on paper but never deliver on the implied performance and specs. This is typical Chinese manufacturing mentality – pretty on the outside but rotten on the inside. Sure, go ahead and get it if you enjoy flashing, hoping for a better ROM and then flashing again… ad infinitum. But for the most of us, we don’t have time to troubleshoot and be beta testers for bad firmware and hidden product/batch defects. We just want a smooth user experience with the tablet’s expected performance.

    Since Intel is “subsidizing” (giving away) its chips to Chinese makers, stick with those since they are now price-competitive with these 3rd-tier Chinese brands. Intel has something to prove in the ARM/tablet space and they will provide aggressive support and price points to win in that market. Just ask anyone using a recent Bay Trail-powered phone or tablet and you’ll know that ARM’s badmouthing of Android x86 compatibility is all FUD. I am what you would call and average user and I have never missed any of the apps that I typically use when I switched over to using an Intel tablet. Sure the Bay Trail quads get hot, but they don’t crash or exhibit lagginess and stuttering behavior unlike RK3188 and RK3288 devices.

    On the issue of battery life, my rule of thumb is simple. To get an idea of continuous intensive use/gaming, just halve the manufacturer’s claims. So when a tablet is claimed to be able to run 10 hours looping video at 50% brightness, I would think this roughly approximates to 5 hours gaming with wifi on. Ebook reading is typically able to match the claimed battery time of most manufacturers.

  6. john

    Sorry could you explain why Rk3288 performs worse? I saw antutu score of 40,000 with Rk3288, but 28,000 range for the intel chip.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s