Results 1 to 8 of 8

Thread: Is there a "Power Myth" about ARM and Tegra?

  1. #1
    Seething Cauldron of Hatred TheAnimus's Avatar
    Join Date
    Aug 2005
    Posts
    17,147
    Thanks
    798
    Thanked
    2,151 times in 1,407 posts

    Is there a "Power Myth" about ARM and Tegra?

    http://www.anandtech.com/print/6529

    Very interesting bit.
    throw new ArgumentException (String, String, Exception)

  2. #2
    root Member DanceswithUnix's Avatar
    Join Date
    Jan 2006
    Location
    In the middle of a core dump
    Posts
    10,113
    Thanks
    503
    Thanked
    1,043 times in 887 posts
    • DanceswithUnix's system
      • Motherboard:
      • Asus X470-PRO
      • CPU:
      • 3700X
      • Memory:
      • 16GB 3200MHz
      • Storage:
      • 1TB Linux, 1TB Games (Win 10)
      • Graphics card(s):
      • Asus Strix RX Vega 56
      • PSU:
      • 650W Corsair TX
      • Case:
      • Antec 300
      • Operating System:
      • Fedora 30 + Win 10 Pro 64 (yuk)
      • Monitor(s):
      • Benq XL2730Z 1440p + Samsung 2343BW 2048x1152
      • Internet:
      • Zen 80Mb/20Mb VDSL

    Re: Is there a "Power Myth" about ARM and Tegra?

    My Tegra3 phone is wonderful in this cold weather at keeping your hands warm.

    I am not sure ARM is the best place to start for running mobile applications, though x86 sure as hell is the wrong place.

  3. #3
    Seething Cauldron of Hatred TheAnimus's Avatar
    Join Date
    Aug 2005
    Posts
    17,147
    Thanks
    798
    Thanked
    2,151 times in 1,407 posts

    Re: Is there a "Power Myth" about ARM and Tegra?

    Quote Originally Posted by DanceswithUnix View Post
    I am not sure ARM is the best place to start for running mobile applications, though x86 sure as hell is the wrong place.
    Did you read the article?

    It shows that the power useage of the Intel CPU is roughly the same as the ARM one, thats really quite impressive and, I admit, not what I would have thought.

    However when you consider the advanced pipeling used in x86, it is I suppose, possible to be rather efficent, if your able to shut down and start up the appropriate bits of silicon. My understanding is that with the Cortex designs, nothing really gets shut down.
    throw new ArgumentException (String, String, Exception)

  4. #4
    root Member DanceswithUnix's Avatar
    Join Date
    Jan 2006
    Location
    In the middle of a core dump
    Posts
    10,113
    Thanks
    503
    Thanked
    1,043 times in 887 posts
    • DanceswithUnix's system
      • Motherboard:
      • Asus X470-PRO
      • CPU:
      • 3700X
      • Memory:
      • 16GB 3200MHz
      • Storage:
      • 1TB Linux, 1TB Games (Win 10)
      • Graphics card(s):
      • Asus Strix RX Vega 56
      • PSU:
      • 650W Corsair TX
      • Case:
      • Antec 300
      • Operating System:
      • Fedora 30 + Win 10 Pro 64 (yuk)
      • Monitor(s):
      • Benq XL2730Z 1440p + Samsung 2343BW 2048x1152
      • Internet:
      • Zen 80Mb/20Mb VDSL

    Re: Is there a "Power Myth" about ARM and Tegra?

    Quote Originally Posted by TheAnimus View Post
    Did you read the article?

    It shows that the power useage of the Intel CPU is roughly the same as the ARM one, thats really quite impressive and, I admit, not what I would have thought.

    However when you consider the advanced pipeling used in x86, it is I suppose, possible to be rather efficent, if your able to shut down and start up the appropriate bits of silicon. My understanding is that with the Cortex designs, nothing really gets shut down.
    I read it when it came out.

    Tegra-3 isn't a good low power implementation of ARM.
    Atom isn't a good implementation of anything. I certainly don't buy it having advanced pipelining, the fact the static scheduled pipelines suck is what makes it such a good target for hyperthreading.

    Intel are trying to spin up the Atom as being competitive. It is still a poor choice for a PC where it actually has software support, and a worse choice for mobile where it has awful drivers on top of a lack of native support for things like Android.

    The fact that it sucks for power draw no harder than Nvidia's wonky 40nm silicon doesn't do it for me, sorry.

    Kind of feels like comparing the donkey and llama entries in a motorbike race. I don't care which smells least, I'll take the Ducati please

    Or maybe I'm just being grumpy after a long week

  5. #5
    Bows out! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Hopefully somewhere less backstabby
    Posts
    28,789
    Thanks
    3,203
    Thanked
    4,456 times in 3,442 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Is there a "Power Myth" about ARM and Tegra?

    Quote Originally Posted by TheAnimus View Post
    The companion core seems to be not active under Windows 8 IIRC. Moreover,you are comparing X86 Windows which has had years of optimisation work against the ARM version which has not. Maybe,Anandtech should check the same hardware under Android too??

    Another aspect is the ancient IGP on the Intel SOC,which also is not helpful to the overall user experience.

    Also,one of the main performance benchmarks Anandtech uses ,like SYSMark has Intel involvement:

    http://www.anandtech.com/show/6529/b...wer-analysis/5

    Look at this post:

    http://www.xtremesystems.org/forums/...=1#post5150663

    The new Intel multimedia benchmark used by AnandTech: TouchXPRT 2013 is another brainchild
    of dr Who's boss Shervin Kheradpir, General Manager of Intel's Performance Benchmarking and Analysis
    Group and founding President of Bapco (via HDXPRT/Principled Technologies)
    Bapco is behind Sysmark.

    BTW,the chap who started that thread is an Intel employee.


    Those despicable Elk,stealing the pond weed!

  6. #6
    Not a good person scaryjim's Avatar
    Join Date
    Jan 2009
    Location
    Manchester
    Posts
    15,044
    Thanks
    1,194
    Thanked
    2,246 times in 1,847 posts
    • scaryjim's system
      • Motherboard:
      • Dell Inspiron
      • CPU:
      • Core i5 8250U
      • Memory:
      • 1x 8GB DDR4 2400
      • Storage:
      • 128GB M.2 SSD + 1TB HDD
      • Graphics card(s):
      • Radeon R5 230
      • PSU:
      • Battery/Dell brick
      • Case:
      • Dell Inspiron 5570
      • Operating System:
      • Windows 10
      • Monitor(s):
      • 15" 1080p laptop panel

    Re: Is there a "Power Myth" about ARM and Tegra?

    I thought it was fairly widely accepted that Tegra, in each of its incarnations so far, has been a power-sucking abomination of an ARM SoC. Just look at the graphs in the AT article and it's fairly clear that the bulk of T3's power deficit comes from the GPU, not the CPU. Take a shoddy ARM implementation and slap a power hungry GPU on it, it's going to look bad. That'd be a worthwhile article if it compared a range of ARM SoCs to Atom, but it doesn't. It picks one shoddy one, and goes "look, Atom's better than this bad ARM SoC, so it must be way better than all ARM SoCs". Sorry, but basic rigor fail.

    Plus, there's several language fails in the article too, so it's not even a particularly good piece of writing....

  7. #7
    Senior Member
    Join Date
    Jul 2009
    Location
    West Sussex
    Posts
    1,075
    Thanks
    73
    Thanked
    132 times in 124 posts
    • kompukare's system
      • Motherboard:
      • Asus P8Z77-V LX
      • CPU:
      • Intel i5-3570K
      • Memory:
      • 2 x 8GB Crucial Ballistix Elite PC3-14900
      • Storage:
      • Crucial MX200 | Sandisk Extreme 120GB SSD | WDC 1TB Green | Samsung 1Tb Spinpoint
      • Graphics card(s):
      • Sapphire R9 290 VaporX 7950
      • PSU:
      • Antec 650 Gold TruePower (Seasonic) or Seasonic SII-330
      • Case:
      • Aerocool DS 200 (silenced, 53.6 litres)l)
      • Operating System:
      • Windows 10-64
      • Monitor(s):
      • 2 x Dell P2414H

    Re: Is there a "Power Myth" about ARM and Tegra?

    Yes, but no techsite is going to write an article about Tegra being JHH without any clothes one... Sites which previously wrote bad stuff about Nvidia have been blacklisted by them. Intel is hardly that different either. Certainly a site like SA - which admittedly does seem to be mainly an excuse for Charlie to rant - which has a go at Intel, Nvida, Microsoft etc. is hardly going to get much advertising revenue. These days even Guy Kewney (never mind Mike Magee) would probably be dropped by any big run magazine (not that there any left!) after complaints from advertisers and industry heavyweights.

    Nvidia get treated with gloves most places anyhow. Certainly bumbgate did not get the coverage it deserved at the time or still does (lots of desktop 8800GT's still dropping like flies). But with their ARM strategy, I imagine they'll eventually get a decent design out (Project Denver or whatever it's called) but Qualcomm needn't worry yet.

    As for Intel, well Atom until recently has been a half-hearted effort. They seem to be aware that a low-end CPU was required but in the netbook days they just used it as excuse to churn out poor chipsets (so as to keep old fabs going). Intel are very successful and make huge margins on their mainstream and server CPUs but people do tend forget that they make lots of mistakes too (P4, Itanium, Larrabee etc.).

    Difference is that because they have such huge margins elsewhere they can afford to make mistakes. Bulldozer may end up almost killing AMD while Larrabee while expensive won't do Intel much harm (and there's always the chance that Knight's Corner will make money). Being cash-rich Intel can afford the best fabs which has saved them in the past (P4 vs Athlon64) and may do so in the future (Atom 22nm may be a so-so design but if it's one node ahead of ARM it may still be impressive).

  8. #8
    Bows out! CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Hopefully somewhere less backstabby
    Posts
    28,789
    Thanks
    3,203
    Thanked
    4,456 times in 3,442 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Is there a "Power Myth" about ARM and Tegra?

    If the companion core is not working under Windows 8,they are comparing X86 windows and ARM windows ,and the benchmark is made by a company formed by the General Manager of Intel's Performance Benchmarking and Analysis Group,then Intel might just do a bit OK!!

    Then add an old GPU into the mix,and I wonder what SOC has a better user experience??


    Those despicable Elk,stealing the pond weed!

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •