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

The Official PC building thread - 4th Edition

Discussion in 'Building a new PC' started by ddp, Sep 13, 2010.

  1. theonejrs

    theonejrs Senior member

    Joined:
    Aug 3, 2005
    Messages:
    7,895
    Likes Received:
    1
    Trophy Points:
    116
    Sam,
    What are you and Estuansis, the Bobbsy Twins playing little sir echo! LOL!! JK! care to take a scientifically wild asses guess what market this is aimed at? It it's present incarnation, it's meant for the general public. I'm sure as time goes by, they will have more powerful APUs. I read through the review I posted as well as a bunch of others, and I just don't see the i3 whipping up on it very much. Like all comparisons between Intel and AMD, both have solid points and some not so solid points. I do a lot of video encoding, so I would stay with AMD. I won't really be interested until they have a 6 core one, or higher.

    I may even just go ahead and buy a socket AM3+ motherboard for my 6 core so I can get all the advantages out of my 6 core for now, and plug in a Bobcat or bulldozer (whatever is socket AM3+) when the time comes.
    http://www.newegg.com/Product/Product.aspx?Item=N82E16813128519

    Russ
     
  2. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    I'm with you russ. :p
     
  3. sammorris

    sammorris Senior member

    Joined:
    Mar 4, 2004
    Messages:
    33,335
    Likes Received:
    7
    Trophy Points:
    118
    As always, it depends more on which review you read, not on what program is being tested. In the bit-tech review, the Fusion CPU is plainly behind the i3 on almost every count, including video encoding, yet some sites paint the opposite picture. With a different socket required, AMD lose their main advantage for a lot of home builders, this design is pretty much solely confined to the realms of off-the-shelf PCs. A smart decision though, because it's those that make up the majority of the market share.
     
  4. Mr-Movies

    Mr-Movies Active member

    Joined:
    Nov 9, 2002
    Messages:
    1,225
    Likes Received:
    0
    Trophy Points:
    66
    Me too! Besides Intel boards are typically twice as much as AMD boards and I never really liked Intel's lack of features although they have gotten a bit better lately. Most CPU reviews are very biased much like Antivirus reviews and benchmarks just don’t cut it. I have seen 30% better performance ratings between CPU's, but when put in the real world heavy duty apps it is nowhere near that. The i3 combo is cheap, but is it cheap enough to take me from AMD platforms? I think not.

    So can you get a MB with integrated video and crossfire it with the Fusion processors? This would give you a good bump in two low GPU's and could be a benifit.
     
  5. theonejrs

    theonejrs Senior member

    Joined:
    Aug 3, 2005
    Messages:
    7,895
    Likes Received:
    1
    Trophy Points:
    116
    Sam,
    After reading several other reviews, including the one I posted, I have to ask myself what is going on in the BitTech review. There seems to be quite a difference than is shown by others, right down to the power consumption at idle. I'm not saying that CPUs can't be different, because they all are. I just openly question the tremendous amount of difference in their report. I have to question that because there shouldn't be that much difference between samples. Another thing I find questionable was the overclocking. BitTech was unable to overclock past stock, while LegitReviews was able to overclock theirs to 3.4GHz quite easily by raising the fsb to 117MHz and adding .1v to the cpu with the 29x multiplier, for a hair under 3.4GHz without issue??? Then again, LegitReviews used a Gigabyte MB and BitTech used an Asrock! Maybe the MB is to blame. Whatever it is, BitTech is way out of line with other review sites on the A-3850 test results. BitTech could have a bad APU, bad MB, or just bad luck, but their results are not representative to that of others. I don't feel that it is deliberate, but there certainly is something wrong somewhere.

    Russ
     
  6. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    I can only imagine how much better the future amd offerings will be. I just ran dvd rebuilder(CCE encoder), simultaneously with ConvertX and did not notice any performance hit on either application. My 965 continues to impress me. And it only averaged 75-80% cpu usage! If bulldozer is only 10 - 15% better, I'll love it LOL! But i'd wager it'll do better than that ;)
     
  7. sammorris

    sammorris Senior member

    Joined:
    Mar 4, 2004
    Messages:
    33,335
    Likes Received:
    7
    Trophy Points:
    118
    Surely if it's only at 80% CPU usage and not affecting your system, then the program isn't utilising your CPU properly?
     
  8. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    ConvertX, and CCE only use so much of the potential. Which leaves the processor open for multitasking :p Perhaps bulldozers instructions will allow better usage. I really don't know. CPU's are beyond my comprehension at the moment :p

    The HC encoder built into Dvd rebuilder will utilize the full processor from what I've heard, but at only marginal gains. But CCE is better in my opinion.
     
    Last edited: Jul 7, 2011
  9. sammorris

    sammorris Senior member

    Joined:
    Mar 4, 2004
    Messages:
    33,335
    Likes Received:
    7
    Trophy Points:
    118
    All the decent programs I've seen use all of your CPU, because that's the most efficient. Not sure why you'd be seeing far less unless you're I/O limited for input. With standard def DVDs I can see how that could be a problem.
     
  10. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    I gather you don't use ConvertX. Unless the coding is optimized for intel processors. That would not surprise me!!!
     
  11. Mr-Movies

    Mr-Movies Active member

    Joined:
    Nov 9, 2002
    Messages:
    1,225
    Likes Received:
    0
    Trophy Points:
    66
    What time frame did each program finish at? Where they the same amount of time to perform as they were on other computers you have or were they longer/shorter. CPU usage isn't enough on it's own I've found that when transcoding I'll use as much reasources as I can safely use so the real factor is lenght of time.
     
  12. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    I wouldn't go by time frame. The dvd that I rebuilt didn't require much compression at all. But it still utilized 50 - 60% of the processor as is normal for CCE. For me anyhow. ConvertX was a 1080P to dvd conversion, which is still running by the way. It was averaging 24Fps while the dvd was encoding, and is still at ~24Fps. Keep in mind, that each process was using it's own hard drives. VERY important factor!
     
  13. Mr-Movies

    Mr-Movies Active member

    Joined:
    Nov 9, 2002
    Messages:
    1,225
    Likes Received:
    0
    Trophy Points:
    66
    If you don't go by time frame it truly doesn't mean anything. I can run one app at 80% or three apps at 80% or higher. Unless your dropping frames or getting jitter with what you've done it doesn't really tell me anything, nor am I surprised at you revelations.
     
  14. sammorris

    sammorris Senior member

    Joined:
    Mar 4, 2004
    Messages:
    33,335
    Likes Received:
    7
    Trophy Points:
    118
    It's mainly a question of software optimisation, operating system, and the number of cores. Running 100% on a single-core CPU has more of an impact on your performance than 100% on a quad-core CPU (even though 100% on each core), but not all 100%s are equal. Some programs can 100% your CPU and you'd never notice apart from increased fan speed / temps, whereas others (usually stress tests) can bog down even the most powerful CPU like you wouldn't believe.
     
  15. omegaman7

    omegaman7 Senior member

    Joined:
    Feb 12, 2008
    Messages:
    6,955
    Likes Received:
    2
    Trophy Points:
    118
    Ehh, it tells me alot. Perhaps I'm not conveying/wording properly. Cause I understand pretty well, despite my lack of CPU knowledge.

    CPU's don't just unleash their full potential indefinitely. It depends on coding greatly. The filter used in downsizing the 1080P is what seems to determine the cpu speed. Unless I've missed something.
     
  16. theonejrs

    theonejrs Senior member

    Joined:
    Aug 3, 2005
    Messages:
    7,895
    Likes Received:
    1
    Trophy Points:
    116
    Sam,
    The reason is Cinema Craft Encoder only runs two cores. if you want the version that uses more than 2 cores, be prepared to pay more than the average decent car goes for it! The last time I looked, the price was $75,000, and it will handle up to 12 cores. Far too rich for my blood. It's not the computer or the CPU, it just doesn't put that kind of load on my 6 core at all. I average about 50% usage on the two cores. The CCE a lot of us have, started life back in the P4 days and was designed for a single core. Encodes were a 3 hour affair or more, and the program used most of the CPU. Mine has been modified for multiple cores, but it will only run 2 cores. You can set it to 6 but it still only uses 2. I imagine the new version uses a lot more of the CPU than this version does. Still the encodes average about 22-23 minutes, so I'm not complaining!

    Russ
     
  17. shaffaaf

    shaffaaf Regular member

    Joined:
    Jan 5, 2008
    Messages:
    2,572
    Likes Received:
    4
    Trophy Points:
    46
    why have a 6 core if you arnt going to take advantage of it?
     
  18. sammorris

    sammorris Senior member

    Joined:
    Mar 4, 2004
    Messages:
    33,335
    Likes Received:
    7
    Trophy Points:
    118
    +1 - I didn't realise he was using a dual-threaded program.
     
  19. FredBun

    FredBun Active member

    Joined:
    Nov 27, 2003
    Messages:
    940
    Likes Received:
    0
    Trophy Points:
    66
    I'm not sure if me and Russ have the same inards in our pc's, but I think it's close, and yes I do have a amd six core cpu with all the other goodies, whenever I use to convert my video files with convertx with my crappy hp at times it would take 2 hours, and that would be me not running anything else, if I did every thing would freeze up.

    With this new build, things are all together different, I mostly don't understand a lot of the numbers you guys are throwing around, but commin sense can go a long way, now I can convert a 1080p movie in less than a half hour, and that's while I'm on-line with people, e-mailing, surfing the net or even downloading attachments from friends, I even tried running things I shouldn't just to see what happens, the only things that slows me down is if I run a virus scan, and even than not by much.

    I am still not use to how fast things run on this rig, it still amazes me, if there are things faster out there, well so be it, this is about as fast as I need to go, if I insisted on going faster than surely I would say I got the A.D.D's or whatever the hell they call it.
     
  20. theonejrs

    theonejrs Senior member

    Joined:
    Aug 3, 2005
    Messages:
    7,895
    Likes Received:
    1
    Trophy Points:
    116
    Sam,
    I think you have got the wrong idea here. If the program only uses XX% of the processor power, it can't use 100% of it, nor does it need to! It can only work within the requirements and speed limitations of the program. The limiting factor will be the throughput of the computer because the Processor is capable of much more that the throughput will allow. You can only move information so fast through the computer. The CCE I have is old, designed for the old P4, and was modified for dual cores. I'm sure that someone that could afford the price tag of the latest CCE, which is designed for up to 12 cores, it would be significantly faster.

    You aren't old enough to remember a program called AT Slowdown. Back in the day, many early games ran faster as the speed of the CPU increased. You used AT Slowdown so the older games were playable on faster machines. Remember, the original Intel 386 was 12MHz, with a 16MHz Turbo!

    That being said, be prepared to be amazed. This is like nothing I've ever seen before!

    http://www.youtube.com/watch?v=ZboxMsSz5Aw&feature=player_embedded

    Bet you this will mess with your mind a bit! LOL!!

    Best Regards,
    Russ
     

Share This Page