The Official Yoga 3 Pro Release Thread

Discussion in 'Yoga, IdeaPad & Essential' started by ikjadoon, Sep 9, 2014.

Thread Status:
Not open for further replies.
  1. Cellular-Decay

    Cellular-Decay Notebook Evangelist

    Reputations:
    66
    Messages:
    395
    Likes Received:
    73
    Trophy Points:
    41
    You can download the hardware maintenance manual and see there is a fan on page 39.

    y3pro_fan.png
     
    ikjadoon, gadgetrants and nim_ like this.
  2. Cellular-Decay

    Cellular-Decay Notebook Evangelist

    Reputations:
    66
    Messages:
    395
    Likes Received:
    73
    Trophy Points:
    41
    That battery is typical of Ultrabooks. They usually have most of their interior space occupied by the battery. The Yoga-2 Pro and X1 Carbon look similar inside.

    Soldered RAM is also quite common. All the Yoga 2 models have soldered RAM as well.

    And you can see the fan in that pic of the Yoga 3 Pro interior. So much for the fanless design everyone was talking about pre-release.
     
  3. MSGaldenzi

    MSGaldenzi Notebook Deity

    Reputations:
    107
    Messages:
    1,172
    Likes Received:
    59
    Trophy Points:
    66
    So being 10% better, are you implying that its on par with the i5-4300y?
     
  4. ikjadoon

    ikjadoon Notebook Deity

    Reputations:
    222
    Messages:
    1,623
    Likes Received:
    41
    Trophy Points:
    66
    Hmmm...I want to chalk this up to 1) random failure and/or 2) pre-release hiccups, but looking at the lingering issues of the Y2P...from my Google'ing, I've seen two possible workarounds: disabling hardware acceleration and disabling sound effects. Now, the actual solution? I've got no idea. :(

    We don't. Eh, hard to tell. Lenovo seemed to be the partner for Intel's reference Broadwell-U (i3/i5/i7) laptop. But, the Broadwell-U line is pretty than the Core M in terms of motherboard design. It's not a drop-in replacement. But, they included fans, too, in the Y3P which we only expected on Broadwell-U laptops. That'd be some serious sloppy design recycling, however, and I expect Lenovo to be better than that.

    Day to day usage, probably not too different between Core M and Broadwell-U. High-CPU load -> games, rendering, photo editing, etc. -> those will be on par with Haswell-U laptops like the Y2P.

    @IntelUser (I forgot to quote your post, :( )

    Oh, I see. Thanks for the clarification. And only the 5Y70, too.

    Not even pre-release. Every single article I've read about the Yoga 3 Pro, some even written just yesterday, say it is fanless.

    ZDNet: There's no fan, which makes for silent operation, and we didn't notice the system getting too warm — even when running a demanding benchmark like Cinebench R15.
    CNET: It runs so cool and efficiently that the systems do not need internal fans. With a combination of an Intel Core M processor and up to 256GB of SSD storage, the Yoga 3 Pro has essentially no moving parts inside.
    Gizmodo: Equipped with an Intel Core M-70 processor for a completely fanless PC, the Yoga 3 Pro apparently sucks so little power that it doesn't require a seperate port to charge the machine.

    Looks like they didn't actually ask Lenovo (or, as nim_ states, Lenovo's own PR people don't know) and just assumed, as we all did, that pretty much any Core M system would be fanless. Why hasn't Lenovo contacted these outlets to provide updated information? These are not small-time technology blogs, lol.

    Not to put too much negativity towards the Y3P, but this is typical Lenovo. Intel provides a specific AC-7265 WLAN for Core M geared towards low power, but Lenovo decides they know better (...like they did with the Y2P /sarcasm) and uses their own. I'm not holding my breath on its performance/compatibility. Intel provides guidelines and models for fanless 2-1 systems (detachables, even!) and Lenovo is like...nope, we're actually going to use 2 fans.
     
  5. IntelUser

    IntelUser Notebook Deity

    Reputations:
    340
    Messages:
    1,591
    Likes Received:
    21
    Trophy Points:
    56
    Very interesting. Based on the Cinebench score, it runs at 1.6-1.7GHz.

    Lenovo may have capped the clocks on purpose. It should be able to score well at least in ST if its not limited and score mediocre in MT.

    The Lenovo Helix 2 benchmarked with Core M 5Y10a gets 2064 in ST and 4249 in MT. If its throttling from its peak, the 5Y70 in the Y3P should score much higher in the first two runs. It doesn't.

    Do you mind running the benchmarks on various modes? Maybe 1) Laptop 2) Tablet 3) Stand?

    We should have known. With the Helix 2 it advertises it as fanless. With Yoga 3 Pro it doesn't. It would be a marketable feature if it was fanless.

    Umm, maybe? Though the 4300Y systems were too thermally limited to perform anywhere near the spec. I think one 4300Y system significantly underperformed 4202Y systems. Theoretically, what you are suggesting is right though.

    So what Haswell Y couldn't do is achieve <1kg Tablets without significant performance degradation. Broadwell Y at 800g can do what Haswell Y needs at nearly 1kg. Based on this: http://www.extremetech.com/wp-content/uploads/2014/09/intel-core-m-fanless-sweet-spot.png

    the super thin devices like Asus T300 Chi will need to run 3.5W cTDPdown.

    Hmm, Core M isn't as impressive as hyped.
     
    ikjadoon likes this.
  6. nim_

    nim_ Notebook Enthusiast

    Reputations:
    0
    Messages:
    20
    Likes Received:
    9
    Trophy Points:
    6
    will do these different mode benchmarks later on today. for now the hint about turning off video hardware acceleration in the browser against the crackling problem seems to have worked - at least for balanced mode. on power saver it still crackles, so I guess that's not gonna go away too soon. By the way, running Windows 8 you have to turn off the hw accel by going to "chrome://flags" and "activating" the "deactivation of video hardware acceleration" there. Am I right when guessing that battery life will take a hit using software rendering as opposed to hw-accelerated video playback?

    Update:

    Benchmark scores for Tent Mode (higher than in laptop mode?!):

    Cinebench 11.5 Open GL 17.69 fps
    CPU 2.08 pts
    CPU Single Core 1.08 pts

    Geekbench 3
    1. Single: 2504 Multi: 4290 --- going up!
    2. Single: 2231 Multi: 3821 --- throttling again...
    3. Single: 2418 Multi: 4195 --- after a little cooling down

    PCMark7
    1. 3521
    2. 3490

    3DMark06
    1. 4914
    2. 4748


    Benchmark scores for Tablet Mode (clockspeed is obviously intentionally lowered):

    Cinebench 11.5 OpenGL 14.14 fps

    Geekbench 3:
    1. Single: 2067 Multi: 3495
    2. Single: 1783 Multi: 2867 (hello throttling)
    3. Single: 1994 Multi: 3336
    4. Single: 2052 Multi: 3409
    5. Single: 2111 Multi: 3550 (after letting it cool down)

    PCMark7
    1. 3233
    2. 3246

    3DMark06
    1. 3252
    2. 3253

    that's it

    I'll try to get rid of this harmony stuff that recognizes how you fold and bend the device later on and try again, to see if I can get it to run at proper speed in laptop mode. It's odd they are presetting it to run slower in laptop mode than in tent mode.
     
    IntelUser and ikjadoon like this.
  7. Cellular-Decay

    Cellular-Decay Notebook Evangelist

    Reputations:
    66
    Messages:
    395
    Likes Received:
    73
    Trophy Points:
    41
    It's not like Lenovo kept it a closely guarded secret. One of the things I like about Lenovo is that their hardware maintenance manuals are free for anyone to download (as I did in the post before the one you quoted). And how did the people who claimed to have run benchmarks not notice the fan? Is it really that quiet?

    Anyway... I guess I'll be hanging on to my Yoga 2 Pro. I'll take the extra inches and ounces for superior performance.

    Would love to see an 11 inch version of the Y3 though. If Lenovo can fix the thermal throttling and shrink this down to an 11 inch form factor with a 1080p screen, that very well might be the "convergence" machine a lot of people have been waiting for.
     
  8. IntelUser

    IntelUser Notebook Deity

    Reputations:
    340
    Messages:
    1,591
    Likes Received:
    21
    Trophy Points:
    56
    Thank you.

    The score isn't that much better on Tent versus Laptop though. If anything I'd say perhaps it can run cooler in Tent mode if the vents are on the hinges and its in more open air then in Laptop mode? Or maybe its possible that you ran when ambient temperature was lower. :)

    It might be running at 4.5W nominal TDP in Tent/Laptop and 3.5W in Tablet. The latter is fine, but former needs to perform better.

    It does only have to cool 6W at max and probably 4.5W really. I have to say again though even 4.5W is a stretch for fanless.
     
  9. nim_

    nim_ Notebook Enthusiast

    Reputations:
    0
    Messages:
    20
    Likes Received:
    9
    Trophy Points:
    6
    Conditions where all the same.. around 20-22°C indoors.

    After removing the Harmony software the scores are all the same in all different modes - around 2500 single and 4150 multi on geekbench 3, no matter if tent, laptop or tablet, or plugged in or not.
     
  10. IntelUser

    IntelUser Notebook Deity

    Reputations:
    340
    Messages:
    1,591
    Likes Received:
    21
    Trophy Points:
    56
    I think if I understood correctly, the Core processor datasheets always indicated that the thermal boost duration can be set per application. I thought it was strange and I misunderstood it but maybe Harmony software has such settings?

    Interesting the score is practically identical to the 5Y10a used in the Helix 2. Does it throttle on Cinebench?

    *Update:* Ok, so the single thread score is higher by about 25%, corresponding to the maximum Turbo mode differences. So the reasoning for using 5Y70 would be entirely for "bursty" application performance, and responsiveness.

    Personally I would get the device if I wanted an Ultrabook(regardless of my want I have an XPS 12 with Ivy Bridge that I paid quite a lot not too long ago). Although not performing to the level hyped in Core M presentations, it has same portability, an actual keyboard(not flat touch one) and noticebly better performance. Price is a bit too high though. Why they made it *higher* than the Yoga 2 is beyond me.
     
    gadgetrants likes this.
Loading...
Thread Status:
Not open for further replies.

Share This Page