One Month, Rev. C Review

Discussion in 'MacBook Air' started by ayeying, Jul 13, 2009.

  1. ayeying macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #1
    So I've had this system for approximately 1 month now (About 4 days away). I've been using it constantly as my primary machine. Don't really have much of a choice since I sold my Mac Pro for this lol. Furthermore, I am horrible at multiple systems. Keeping just my current OSX and Windows in nice working order is hard. Trying to manage 2 systems with 4 OS or more is just torture with me.

    Anyways, continuing on.

    My specs are:

    2.13GHz Intel Core 2 Duo (S)L9600 Processor, 6MB L2 Cache
    2GB Integrated DDR3 1066MHz Ram
    128GB Samsung Solid State Drive w/ 3.0Gbps SATA
    256MB Shared DDR3 nVidia GeForce 9400M Video Card (300MHz/1075MHz Core/Shader Speeds)
    40 wHr (5560mAH) Integrated Battery @ 7.2Volts, rated for 750 Cycles before 80% Health

    My current configuration for OS/Drive is:

    64GB - OSX Snow Leopard 10.6, Build 10A402a
    64GB - Windows Vista 7 Ultimate 32-bit

    Now, my current settings is perfect for me.

    *Do note, this review is based on my opinion on my system, it may differ from other people's opinions

    Good things about this revision of the MacBook Air compared to my older Rev. A:

    - It's faster, coming from a Rev. A (1.6GHz/80GB HDD), it is extremely faster in many ways even regular surfing the web and checking e-mails.
    - The screen is brighter. This is only if you got the ones with no lines
    - The battery life has improved. For me, its a pretty big improvement from my Rev. A's 3.5 Hours to current 4-4.5 Hours
    - Charging time has greatly improved. I can easily recharge the system to a good usable 80-85% within 2 hours.

    The bad things about this revision compared to the Rev. A:

    - It runs hotter. This is caused by many factors including a faster processor generating more heat, the video card generating more heat due to sharing with the northbridge, etc.
    - The volume on the integrated speaker is quieter/softer by at least a notch. I'm not sure why this is.
    - The vents are not as efficient as the Rev. A's. It seems like my heatsink is somewhat warped, pushing on the vent holes. This is not a normal occurrence so this isn't a "huge" point on the negative.

    Other notes about the Rev. C:

    The two major things I noticed is that the system is: The system will downclock itself very easily, especially the 2.13GHz model. Think of it like this:

    If CPU temperature reaches above 80-85 Deg C, the 2.13GHz speed will be downgraded to 1.86GHz or 1.60GHz to allow the fan/heatsink to cool the processor core down/manage it under 80-85 deg C temperature.

    This is also the reason why the Rev. C's 2.13GHz has repeatedly reported being "slower" then the Rev. B's 1.86GHz model, especially the Processor/Graphics scores.

    The second item on the list is the graphics core. Our graphics core is clocked at:

    2D:
    100Mhz Core
    1075MHz Shader

    3D Low Power:
    150Mhz Core
    1075Mhz Shader

    3D:
    300MHz Core Speed
    1075MHz Shader Speed

    However, the bad part about this is that on specs (or on paper), the 3D speeds should be:

    3D:
    450MHz Core Speed
    1100MHz Shader Speed

    *Note: Ram speed is excluded because the ram is shared and not dedicated.

    So in reality, our video cards are still "underclocked" compared to the Aluminum MacBook/13" MacBook Pro/15" MacBook Pro/17" MacBook Pro/iMac Base Model/Mac Mini versions. However, if we were to overclocked the 9400M to stock speeds, it is 100% possible. I was able to do it for a little while but for some reason my settings will not change anymore (not sure what happened here). At 450Mhz/1100MHz, the system is very responsive. CoD:WaW is playable at a higher resolution (1024x768 vs my current 800x600).

    *Note: These clock settings are given from Windows Vista/Windows 7. Actual settings in OSX are unknown.

    Now the funny thing is, if you were to overclock the CPU (Yes, this is possible) to 281Mhz FSB (vs 266MHz FSB), given a total clock speed of 2.26GHz, the graphics performance is the same as if you were to have a default (450/1100MHz) clock speed. (Not sure why this is, needs more testing)

    For you numbers people, my current Geekbench scores are:

    32-bit: 3372
    64-bit: 3601

    For Windows 7 scores:

    Processor 5.4
    Memory 5.5
    Graphics 5.2
    Gaming graphics 5.1
    Primary hard disk 5.9

    *Note: Geekbench scores were conducted with Coolbook running only to regulate throttling. Stock voltages were used. Windows 7 scores were conducted with no undervolting software. Updated 186.06 Graphics drivers were used with stock 2.13GHz processor speed and 300MHz/1075MHz graphics core/shader speeds.

    Overclocked GPU or CPU resulted in the same score for Windows 7. Overclocking GPU or CPU is currently not possible in OSX, if there is, please inform me so I can test this in OSX also.

    Temperatures for the MacBook Air comes to something like this:

    @ Light usage -> 50 deg C to 75 deg C, fans @ 2500RPM - 4500RPM
    @ Med " -> 60 deg C to 80 deg C, fans @ 2500RPM - 6200RPM
    @ Heavy " -> 75 deg C to 85 (90) deg C, fans @ 6200RPM

    I have redone my thermal paste using many ways. Some of the ways include:

    - One small pea sized drop
    - 1/10th the size of a pea sized drop
    - Evenly spreading out upon the chip
    - A huge blob like the picture in the service manual

    So far, my results from these 4 ways are (in order):

    - No real difference
    - Heat is transferred very easily to the heatsink, but temps hit a high of 98 deg C for GPU and 89 deg C for CPU
    - On cold boot up, temps were a cool 35-40 deg C. With 80-100% load (gaming), temps hit a high of 85 deg C
    - For some reason, this has been the best result for me. Cold boot up temps are higher, but they stay within 55 deg C to 65 deg C with fans at 2500RPM even after 2 hours of light/med usage. Under gaming conditions, temps rise but do not exceed 85 deg C for CPU or GPU.

    Last thing that comes to mind is Battery Life. My current battery life results somewhat like this:

    Normal usage includes Web surfing, Mail.app in background w/ 5 minute auto check, Adium, iTunes running on shuffle.

    @ Minimal Brightness -> 5.5 Hours
    @ Brightness 1/2 -> 4.5 Hours
    @ Brightness 3/4 -> 4 Hours
    @ Full brightness -> 3.5 Hours

    In Windows 7, I don't have "Normal usage", however, I have gaming which is far more interesting in Windows :p

    @ Brightness 1/2 -> 2 Hours
    @ Full Brightness -> 1.5 Hours, sometimes 2 depending on game.

    Games involved include: Call of Duty: World at War Multiplayer and Star Wars: Empires at War. Wifi is on.

    *Note: The battery life results have a +/- 15 minute. The times are not exact.

    My current battery status is 99% Health (5495 mAh) with 77 Cycles

    This is all I can think of so far. It's too hot to think anymore and I want my ice cream lol.

    Feel free to comment, I'll try to answer whatever I can. If there's any other tests you wish for me to run, please speak up.
     
  2. barbro macrumors newbie

    Joined:
    Dec 29, 2003
    Location:
    UK
    #2
    great review! Loads of really useful technical details. One question -your snow leopard geekbench scores are much faster than mine with the same machine. Did you format your HD when you installed SL? I did an upgrade and I think that might be why i'm getting lower scores.
     
  3. miniConvert macrumors 68040

    miniConvert

    Joined:
    Mar 4, 2006
    Location:
    Kent, UK - the 'Garden of England'.
    #3
    Very cool, thanks for all the info! My review of the new MBA compared to the original would have been "It's better", so the detail is greatly appreciated ;)

    The SSD rev A was still a great machine IMHO, but they've really perfected the concept in this latest incarnation.
     
  4. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #4
    I have had done a upgrade and clean install. My current is from a clean install. I did notice a slight jump in scores from a clean install of snow leopard compared to a migrated leopard (from Rev. A) -> snow leopard. This could have many factors since my leopard on the Rev. A had over 160 applications installed and everything
     
  5. gman901 macrumors 6502a

    Joined:
    Sep 1, 2007
    Location:
    Houston, TX
    #5
    Great review! I have a couple questions regarding clock speed for the graphic card, CPU and nvidia drivers update. I use Windows XP sp3 (however I will upgrade to Windows 7). I really do not plan on opening up the machine, so what did you use to clock the speed up for gaming and would messing with the clock speed cause my Air to malfunction? Would this also invalidate my warranty? How did you increase the speed of the graphics card? Finally, I tried to upgrade my drivers but Nvidia auto-check stated I would have to get them from the manufacturer. It said that the current drivers are set to accurately control different components of my Air. I appreciate any insight... I just want the flexibility of giving my Air a slight boost without having any real problems. My current geekbench score after a fresh install with XP is only 2355 which pales in comparison to your score on the same identical hardware. Also, does a cooling pad under the Air make a big difference in keeping the temperature down? I look forward to reading your response!
     
  6. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #6
    Messing with the GPU or CPU clock speeds won't kill your system. However, I don't recommend it running at a higher speed 24/7 and all the time. I only run it when I need the performance, otherwise I keep it at stock. At worst, you'll blue screen or freeze the system. You don't entirely need to open up your system and re-do the thermal paste. Infact, my thermal paste is back to the whole "blob" mess right now. This seems to work best in controlling the temperatures for me.

    It should not invalidate your warranty because this is a software overclock.

    I used ATiTool to overclock the video card and nTune to overclock the CPU.

    Just goto www.nvidia.com and download the GeForce 9 Series (M) video drivers. There is a chance you might have the latest since you're running Windows XP SP3.

    I personally don't use a cooling pad. Infact, I have a Speck Smoke Hard Case on all the time. This restricts airflow and cooling even further. The times I do use a cooling pad is when I'm using the system on my bed. The vents are blocked entirely if I place the system on the bed. The other time I use a cooling pad is when I do video conversions (AVI -> MP4)

    Geekbench score within Windows seems to be a lot lower for me also. I'm not sure. In windows, I was only able to achieve 2765 as best score.
     
  7. Scottsdale macrumors 601

    Scottsdale

    Joined:
    Sep 19, 2008
    Location:
    U.S.A.
    #7
    The only part of the review I have to question is your statement that the rev C MBA runs hotter than your original MBA. I say, "SAY WHAT?" The true nature of the original MBA's component makeup means it will run hotter... very hot. I mean is this your assessment based on core CPU temperature? Or the way it feels on your lap? Or what do you mean?

    While your review may be very technically accurate, I think you could add some subjective analysis in what it all means to the average user to clarify just how valuable your review is for them and their potential decision to buy a rev C MBA.

    Nice review...
     
  8. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #8
    When I say it runs hotter, I'm speaking about the actual temps compared to the Rev. A from what I've seen. The GPU temperature of the Rev. C runs about 65-90 deg C average. CPU seems to run a bit hotter then the Rev. A for me also, but not too much (+3-5 deg C).

    Yeah, I know I should clarify some of the stuff so the average user would understand. I'll try and clarify whatever I can in the upcoming days. I'm just too damned lazy right now since it's 100+ deg F Ambient temp here for the past 2 days :p
     
  9. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #9
    Okay, so yesterday I was trying to run the system under clamshell in Windows 7. Played about 4 hours of CoD: WaW and then the system shut off! (Overheated). At first I thought the battery died because I had unplugged it to run on battery.

    So because of this I thought of a brilliant (or stupid) idea. Try running the system with no thermal paste.

    I took the system apart, used a couple of dozen q-tips and wiped off the AS5 originally on there. I did not use alcohol to clean the surfaces, just to wipe the paste off. Reassembled the system and started running it.

    Tried a few games. My temps were just as normal as it with thermal paste in Windows 7. 2 Hours of gaming gave me a top temp of 86 deg C for GPU and 80 deg C for CPU according to GPU-Z and RMClock

    So I went back into OSX and just used the system normally. Temps were within acceptable limits even using it on the bed.

    So before I went to bed, I tried a less extensive test (to leave on overnight), 7 hours of iTunes + Adium + Mail.app (every 5 minute check) and Firefox on a 30 second continuous loop of YouTube.com... here's the iStat results. (See attached picture)

    My ambient room temp was 63 deg F when I went to bed at 3am. Waking up at 10am here, the room is 70 deg F.

    Note* This was all done under clamshell. The lid was not opened at any time
     

    Attached Files:

  10. gman901 macrumors 6502a

    Joined:
    Sep 1, 2007
    Location:
    Houston, TX
    #10
    ayeying, what settings did you use for nTune to adjust the clock speed? I tried RMA clock, but noticed that my cpu keeps fluctuating between 1.59 GHZ and 2.12 GHZ... when I run Geekbench, it always stays at 1.59 GHZ according to the factory settings. Also, when I tried to adjust nTune GPU or CPU clock speeds, it keeps telling me that it failed. I tried doing the nTune Tune Up option in the Nvidia control panel, but my screen started to freak out and was constantly flashing. Any ideas on the exact settings since we have the same systems? I would like to get a little better results than what I am currently getting on geekbench which is between 1800 - 2300 in Windows XP using a MacBook Air 2.13 GHZ.:confused:
     
  11. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #11
    For nTune, I just adjusted the CPU speed. I just slide the bar to "281MHz". I notice I start crashing around 283MHz so I keep it under.

    RMA Clock.. never heard of it. I'm guessing you're talking about RMClock. It is an undervolting program. I just use it to regulate throttling.

    Also, do note, just because we have the same hardware specs don't mean we'll perform the same. But I am curious why your scores are way lower then mine though.

    I have to ask, are you using Windows XP exclusively on the Air and no OSX?
     
  12. gman901 macrumors 6502a

    Joined:
    Sep 1, 2007
    Location:
    Houston, TX
    #12
    No, I primarily use the OSX side, but only boot to XP when I need to use Messenger or light gaming. I would like to boost my score closer to the 2700 range, but I have no idea why I am getting such low results, usually in the 2300-2400 range with Geekbench. So you suggest not using ntune to mess with the GPU clock, but just slide the cpu clock to 281? I wonder if that would do it... Have you tried using the nTune tuner option yet? Does that do anything worth while? I tried the RMClock which I have installed, but it seems complicated to me. I am not sure how to use it. I am really surprised that my Mac Mini at 2.0 GHZ with 2 GB of SDRAM is coming in at about 2800-2900 range compared to the Air. I really appreciate you responding to my posts because this forum seems kind of dead when you post stuff up here. I guess some of my questions are kind of lame or too specific for anyone to respond, but I appreciate it! :)
     
  13. GeekGirl* macrumors 65816

    GeekGirl*

    Joined:
    Feb 26, 2009
    Location:
    Buffalo, NY
    #13
    Nice review will stick with my A to a real update comes out .
     
  14. RichardF macrumors 6502a

    RichardF

    Joined:
    Oct 27, 2006
    Location:
    New York City
    #14
    OP, could you please comment on the iSight?

    I know in the Rev. A Apple used a lower-resolution iSight compared to the MacBook Pro.

    Have they put in a better iSight in the Rev. C?
     
  15. maroulis macrumors member

    Joined:
    Sep 27, 2006
    #15
    this was the reason I gave my Rev C/SSD to my g/f and bought an MBP 13" and installed X25-M SSD.. on my Rev A i had the core shut down which was later fixed, but NEVER had an MBA which would totally shutdown when overheating!
     
  16. GoCubsGo macrumors Nehalem

    GoCubsGo

    Joined:
    Feb 19, 2005
    #16
    ^ But you have to understand that the smaller form factor wouldn't really handle heat as well as even a 13" MBP.
     
  17. entatlrg macrumors 68040

    entatlrg

    Joined:
    Mar 2, 2009
    Location:
    Waterloo & Georgian Bay, Canada
    #17
    Thanks for taking the time to write such a detailed review.

    Reluctantly, I admit I'm in the same position as you with regards to using multiple systems. Too confusing and I'm on a mission to decide on one main machine.

    I have a Mac Pro for desktop use, a MBP 15" when I need power on the road, and my revb MacBook Air.

    The Mac Pro won't be my main machine since I need portability, for me it's down to my 15" MBP or the Air or buying a 13" MBP and selling the 15" and the Air.

    I've never been so 50-50 on a decision. It's for good reason though, I feel my Air does a terrific job and doing 'most' tasks for me, the thing I don't like is the fan coming on often when you do minor heavy lifting, I can't imagine watching a two hour movie on it, I think it would get too hot, but I should try it before I comment on it I realize but in a way I feel it's just not a machine I
    should use hard as a main machine everyday whereas my MBP I toss it in and out of a bag all day, plug / unplug usb devices often it just feels tough and it works, the Air on the other hand is more delicate...

    Screen size is a decision too, the 15" is nice, you do see more detail in photos and alike, but you're on the heavy side, the 13" MBP has the power, but if you're going to go for a 13 why not add a pound and get more screen etc etc.

    How do you like you Air as a day to day machine from a suitable for all tasks standpoint and user friendly / fun to use?
     
  18. justit macrumors 6502a

    Joined:
    Dec 1, 2007
    #18
    Fan noise is bad on all models and Apple tends to throttle the fans too early. When I know my tasks are within the norm, I force the fan speed down to 2500 rpm with an applescript http://forums.macrumors.com/showthread.php?p=8026123
     
  19. maroulis macrumors member

    Joined:
    Sep 27, 2006
    #19
    my Rev A would be scalding hot but never shut down itself is my point :)
     
  20. maroulis macrumors member

    Joined:
    Sep 27, 2006
    #20
    Went through similar decision process.. but I came from Rev A / MBA and Rev C/MBA (which now my g/f uses).. I agree 100% on the weight between MBP 13 vs 15 but the deciding factor for me wasn't so much the weight but the dimensions/portability of the 15" which is why it was ruled out immediately...

    i really can't give you advice, but I think an MBP 13" with a 24" LED ACD or other brand would work.. I tried my rev C with DEll 30" and my MBP 13" with Dell 30" and from a "styling" standpoint (not to mention speakers/webcam) I have now settler on the 24" LED ACD...
     
  21. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #21
    I came from a (in order):

    13" Black MacBook (My first mac ever)
    13" Rev A MacBook Air
    15" MacBook Pro (non unibody)
    13" Rev A MacBook Air
    15" MacBook Pro (non unibody again)
    13" Aluminum MacBook
    17" Unibody MacBook Pro
    Mac Pro 8-Core + Rev A MacBook Air (Again lol)
    current Rev C MacBook Air

    yeah, I've been there. As you can see, I keep reverting back to the MacBook Air even though I've owned way more powerful systems then this. The weight and portability (since I tend to move around everywhere) is just a huge bonus compared to all the others.

    On a day to day usage, it's the same. I see no difference between any of these systems I've previous or currently own.

    OSX is always fun to use haha.

    I tend to throw my MacBook Air into my backpack more than my previous 17" Unibody MacBook Pro because it was twice as heavy. The weight made it hard for me to "throw" it. I have an opposite feeling compared to you, where, the lighter the system, I tend to throw it/toss it into my bag rather then being slower cause its more fragile.

    For the fan, I created little applications using Automator that limit my fan speed from 2500/3600/4800/6200 RPMs. It works great (only in OSX btw) to control my fan noise, especially when I'm in class or in a meeting. At home, I have music blazing on my Audiophile speakers so the fan noise don't make any much difference. But I completely agree, the fan is a little trigger happy.
     
  22. Scepticalscribe Contributor

    Scepticalscribe

    Joined:
    Jul 29, 2008
    Location:
    The Far Horizon
    #22
    Thanks for the comprehensive and detailed review, it was very interesting and informative. In the medium term, I am thinking of getting a MBA to add to my MBP as I travel a lot, and your great review has certainly given me food for thought.

    Cheers and good luck
     
  23. muronero macrumors newbie

    Joined:
    Aug 4, 2009
    #23
    yes i know the feeling..... i still have my mbp 17 pre-unibody.... but since i bought the MBA i litteraly never opened the 17 yet after two weeks... i work with illustrator, photoshop and indesign... and i'm amazed by the performances of this little jewel... the drawback.. i'm now addicted to the ssd... how could i live without it for so many years...... the fan is there (especially when it is connected to the power adapter) but it's a lot less noisy than the 17... however... when u are carring it inside the bag... everything becomes "clear"...

    i used to have some apple notebooks in the past.... with each one of them i had a curious relation....

    powerbook G4 Ti 1st edition: a lover... it is true... it was a sexy, light computer... but with a lot of problems (never buy the first release, never)...

    powerbook G4 15 Alu: a friend, a companion of adventure... he came with me around the world, he never let me down... killed by a faulty logic board after years...

    macbook pro 17: a co-worker, a pure professional relation; maximum respect.

    macbook air 2.13: a muse, it's an inspiring/ inspired machine; a piece of art.
     
  24. ayeying thread starter macrumors 601

    ayeying

    Joined:
    Dec 5, 2007
    Location:
    Yay Area, CA
    #24
    Okay, here's an update to my review, I know I left some unanswered questions that I'll try to answer here. I apologize if I repeated my answers for no apparent reason cause I'm skipping through all my posts and reading, the users, questions answering them.

    Its been 2 months since my Rev. C, MacBook Air has arrived at my door.

    I',m Currently, I'm in the process of reinstalling my entire system with the latest 10A432 build (Aka, GM Hopefully) of Snow Leopard and Build 7600, x64 bit of Windows 7 Ultimate.

    The clock speeds listed are given from all editions of Windows I've installed and used. That includes XP SP2/SP3, Vista Business x86 and x64, and Windows 7 Build 7100 and 7600, both x86 and x64. Upgrading the graphics drivers does not seem to increase the graphics card's clock speeds at all. For some reason, I was able to overclock the video card before, but I haven't had any success anymore. For some unknown reason, the system just refuses to increase the clock speed.

    The geekbench scores in Windows seems to be a lot lower then in OS X. Why? I'm not sure. Normally the real performance comes from using the system, not benchmark scores.

    Cooling pads, I have an old one thats really loud, seems to work slightly for me.

    To adjust my CPU clock speed, I used System Tools (aka nTune) from nVidia. It's an awkward and lame program imo, but it works. On nTune, I can't seem to adjust the graphics card speeds. As a matter in fact, I can't seem to adjust the graphics card speed no matter what program I use anymore.

    iSight: Last time I checked, Apple uses a 640x480 iSight on all their systems regardless of which model/specs you have. The iSight I see on the current Rev. C looks the same as my Rev. A which looks the same as my old 17" Unibody MacBook Pro and so forth. If there's any difference, I have no idea why.

    The system randomly shutting down only happened once. This was caused by my fault. What happened was that I was using a thermal paste application where it involved putting a huge blob on the diode and squishing it with the heatsink. While this works in low load because the system has more mass to store heat, when a heavy load is induced, the heat produced by the CPU and GPU/Chipset is too much for the heatsink to dissipate. This is what caused my shut down.

    On a day to day use, the system is perfect. I do not see the reason I need to have more performance when all I'm doing is basic e-mail, ichat and a few firefox/safari windows or tabs.

    --

    My current thermal paste application is a small dab, about 1/3 the size of a rice grain thats spread evenly using a plastic bag. The layer is extremely thin but opaque. The heatsink was reinstalled than a clamp applied approx. 20 pounds of pressure to make sure the paste to evenly pressed onto the heatsink. So far, this seems to be my best method of managing heat so far. There's no cookbook under volt involved, only throttling @ 85 deg C.

    The battery is the same. I'm at 98% (5411 mAh) with 156 cycles. Times are more or less the same, give or take a few minutes.
     
  25. Jobsian macrumors 6502a

    Jobsian

    Joined:
    Jul 30, 2009
    #25
    Loving your insights ayeying, honestly it's playing a major part in my formulations and decisions. You've a slightly obsessive and perfectionist sliver which I can relate to

    Please keep posting more guys
     

Share This Page