1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Blogs Don’t be fooled by Laptop CPUs

Discussion in 'Article Discussion' started by Meanmotion, 25 May 2013.

  1. Shirty

    Shirty W*nker! Super Moderator

    Joined:
    18 Apr 1982
    Posts:
    12,446
    Likes Received:
    1,657
    Both I think, but if you stop taking him even slightly seriously he can be highly amusing :hehe:
     
  2. cdb

    cdb No comment

    Joined:
    25 Apr 2009
    Posts:
    476
    Likes Received:
    4
    Was the image editing done using the cpu graphics? If so I still can't see how a 2 core can beat a 4 core cpu.

    As a comparison I've just benchmarked my desktop (I7 860@3.8GHz and gtx460super clocked) vs laptop (I7 3610 standard and GT650m) and the results from the CPC benchmark were

    Desktop Laptop
    I 1399 ....1445
    V 2786 .....2784
    m 1363 .....1235
    0 1849 ....1821

    As you can see, my quad core laptop is as near as damit as fast as my overclocked quad core desktop. I bet at equal speeds, the laptop would win.

    (sorry, I think the I, V, M and O were , Image editing? Video, multi tasking and over all, but forgot to check before I closed the window)

    PS I checked a few months back and my laptop uses about 45W and my desktop 200W+ to achive the same tasks.
     
    Last edited: 25 May 2013
  3. Combatus

    Combatus Bit-tech Modding + hardware reviews Staff Super Moderator

    Joined:
    16 Feb 2009
    Posts:
    2,756
    Likes Received:
    69
    Well for starters your laptop has a modern Ivy Bridge CPU I'm guessing bought in the last 6 months, whereas your PC has a four year old CPU that we actually reviewed in 2009 - not really a fair comparison ;) Yes they're both Core i7's but that's just a name - the newer CPU in your laptop has a vastly different, superior architecture.

    The fact of the matter is it's not just about the number of cores, it's about the raw power of the CPU. Laptop CPUs have to operate at lower voltages so are generally less powerful. Lots of applications still don't make proper use of more than two cores as well. The exception is video editing, where the laptop CPU we tested was faster than the Core i3 CPU. However, the Core i5 desktop processor was still much faster than the Core i7 CPU in the laptop, as it can run at much a much higher frequency.
     
  4. barrkel

    barrkel New Member

    Joined:
    31 Jan 2007
    Posts:
    82
    Likes Received:
    1
    cdb - you're comparing a desktop CPU from 2009 with a laptop CPU from 2012, and overclocked frequency doesn't matter as much as you think it does across different architectures.

    Most of the gains in CPU power in the past decade have been from architecture improvements rather than frequency bumps. I was shocked at how much faster my 3770K was than my old 920, when clocked to the same frequency - twice as fast at some tasks.
     
  5. Combatus

    Combatus Bit-tech Modding + hardware reviews Staff Super Moderator

    Joined:
    16 Feb 2009
    Posts:
    2,756
    Likes Received:
    69
    Have a look at the Core i7 3770K cpu at the top of the graph - it uses the same architecture as the one in your laptop so is a better comparison. It managed a score nearly 500 points higher, and that's at its stock speed...
     
  6. cdb

    cdb No comment

    Joined:
    25 Apr 2009
    Posts:
    476
    Likes Received:
    4
    Sorry, must have missed the bit where it said you had to compare architecture of the same generation.:duh:
    The 3770K is running at 3.5Ghz, so still not really a fair comparison unless you underclock it.

    So all I've proved is my 4 year old oveclocked cpu is as fast as my 1year old laptop?

    Maybe by the end of the year I'll have a Haswell desktop and laptop so I can try again.
     
  7. true_gamer

    true_gamer Well-Known Member

    Joined:
    4 May 2009
    Posts:
    6,622
    Likes Received:
    1,049
    I'm sorry, but this is crap...

    I did many test with my 3920XM in my Alienware laptop, and it destroy a desktop 2600K easily. It was on par with a desktop 3770K.

    The proof is in the pudding...

    My thread

     
  8. Valinor

    Valinor New Member

    Joined:
    30 Oct 2010
    Posts:
    90
    Likes Received:
    0
    I think this shows more that geekbench isn't a good way of comparing performance than anything else...
     
  9. TC93

    TC93 New Member

    Joined:
    11 Nov 2011
    Posts:
    15
    Likes Received:
    0
    In America the shopping channels on TV like to spread lies about how fast a laptop cpu is. Its no wonder people get confused. The TV hosts will outright lie about a product they are selling.

    Yet recently on the news all I hear is how desktops are dead. Yeah right... not.

    I will never give up my desktop computer for any notebook, or tablet computer. Any desktop will kill a tablet in every way. Yet we have the media fools running around claiming the desktop is dead and the tablet is the new computer.

    I would also stay away from any all in one monitor type "desktops". They are just proprietary crap that can't be upgrade, and don't have the speed of a real tower desktop computer. A powerful computer requires space for a real graphics card, and a large cpu cooler for the cpu.
     
  10. cdb

    cdb No comment

    Joined:
    25 Apr 2009
    Posts:
    476
    Likes Received:
    4
    Why not? Surely running exactly the same tests on two items is a very good way of comparing performance?:confused:
     
  11. Yslen

    Yslen Lord of the Twenty-Seventh Circle

    Joined:
    3 Mar 2010
    Posts:
    1,966
    Likes Received:
    48
    Geekbench is really inaccurate, in my opinion.

    Aside from anything else, look at the time it takes to run. The bit-tech suite takes 10x longer despite being only 3 tests. That's because its a real benchmark.

    Geekbench averages scores from loads of tests but if you look at the individual scores they're often totally screwed and make little sense, like a 2009 AMD CPU being quicker than a 3770k, not just fractionally, but by an order of magnitude.
     
  12. thom804

    thom804 Member

    Joined:
    22 Oct 2009
    Posts:
    714
    Likes Received:
    6
    If you are taking benchmarks like that as gospel......well.

    Try the same tests with the CPC bench. 6 years old it may be, but it's still the best measure of a cpu i've ever used.
     
  13. true_gamer

    true_gamer Well-Known Member

    Joined:
    4 May 2009
    Posts:
    6,622
    Likes Received:
    1,049
    CPC bench is old, outdated, and doesn't utilize all cores and threads. Geekbench does... and it also is a good way of comparing cpu's to see how much gain you get from the next gen.

    Also, if you check my thread, I have many other benchmarks I have used. :)
     
  14. cdb

    cdb No comment

    Joined:
    25 Apr 2009
    Posts:
    476
    Likes Received:
    4

    Just looked at that graph again now I'm sober. The 3770K is tested at 4.8Ghz according to that graph. The 3610 is stock at 2.3Ghz, so less than half the speed, but gets 75% of the score?!?!

    Even the I3 3220 runs at 3.3Ghz stock, so again not a massively fair comparison.

    Run all cpus at the same speed on the same number of threads on the same test and see how they compare, that's the only way I can see that would be fair.


    (And for what it's worth my 860 runs superpi1.5 at 10.43s and my 3610 at 11.38s.)
     
  15. ChromeX

    ChromeX Active Member

    Joined:
    12 Aug 2004
    Posts:
    1,606
    Likes Received:
    22
    Absolutely wrong. I travel a lot due to my job as an Engineering Officer for the merchant navy, so I need a laptop that can handle gaming as well as doing my day to day work. My laptop at the minute has a 3630M i7 processor and a nvidia 675M graphics card, and I max out everything at 1920x1080 full options and the only game i've noticed any slow down is metro 2033 where I get around 40 fps
     
  16. true_gamer

    true_gamer Well-Known Member

    Joined:
    4 May 2009
    Posts:
    6,622
    Likes Received:
    1,049
    I 100% agree! I was getting 100+ FPS in BF3 on ultra @1080p with my 2 HD7970M CFX. They maxed out every game with ease... So that guy is talking utter.... :)
     
  17. rollo

    rollo Well-Known Member

    Joined:
    16 May 2008
    Posts:
    7,792
    Likes Received:
    117
    Most people's laptop experience is a £300-£500 cheap as chips laptop. Those who have spent big cash on SLI laptops Ect know exactly what they are capable off.

    Anandtechs CPU bench is better than bits at the moment as its more highly threaded and actually has been updated a bit. Also don't see why the odd game was not included plenty of CPU bound games out there that could of been run.
     
  18. Yslen

    Yslen Lord of the Twenty-Seventh Circle

    Joined:
    3 Mar 2010
    Posts:
    1,966
    Likes Received:
    48
    Gaming on the vast majority of laptops won't be anything like that level of experience. The real dissapointments are those priced roughly the same as decent gaming desktops, with similarly named components. The customer is meant to think they're equivalent, but they're nowhere near as good.

    Your laptop is (or was? you sold it, right?) awesome, certainly, but it's proving the point of this article quite nicely whether you like it or not. It's much, much more expensive than a desktop with an i7-3770k, the CPU it's roughly equivalent to. You spent over £3k on that thing, unless I'm mistaken, which probably double the cost of a desktop of equivalent power.

    From the price and the name, the uninformed consumer would think that it would match an i7-3930 or similar, but it won't. That's the point this article is making, and I think it's a perfectly valid one.
     
  19. Boogle

    Boogle New Member

    Joined:
    8 Mar 2002
    Posts:
    282
    Likes Received:
    6
    I agree with true_gamer here. This single benchmark is odd, especially compared to what's out there and true_gamers numerous benchies. I've informally tested my Macbook (i7-3740QM) against my (PC) desktop (i5-2500) and they're pretty damn close. It's hardly a surprising result since other than a small architecture change (SB-IB) and clock delta, the CPUs are almost identical (even turboing to the same speed):

    http://ark.intel.com/products/52209/Intel-Core-i5-2500-Processor-6M-Cache-up-to-3_70-GHz
    http://ark.intel.com/products/70847/Intel-Core-i7-3740QM-Processor-6M-Cache-up-to-3_70-GHz

    One cherry-picked benchmark is definitely not enough to say a desktop i3 is faster than a mobile i7. The i3 is even 2 cores down, and it's not like the i3 has IB cores and the i7m has Atom cores.

    I just wish the GT650 was as close to it's desktop counterpart namesake as the Intel CPUs are.
     
  20. rollo

    rollo Well-Known Member

    Joined:
    16 May 2008
    Posts:
    7,792
    Likes Received:
    117
    The fact a gaming laptop can do that level of performance means its out there despite some people saying its impossible. All true gamer and others have done is prove if money is no object you can buy a decent spec laptop.

    Desktop gaming pcs don't really exist for most of the general public its whatever pc world and company sell. Dells cheapest gaming pc is pretty expensive for what it is as well.

    Overclockers and scan both sell pcs but I wonder how many people in the general population even know either site exists.

    The biggest seller of pcs in the world now is Lenovo and they don't sell a single desktop gaming pc.
     
Tags: Add Tags

Share This Page