Announcement

Collapse
No announcement yet.

AMD FX-8350 "Vishera" Linux Benchmarks

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    It's to sad to see AMD is no longer even contending in the high-end segment. AMD is even struggling to compete with Intel's mid range offerings.

    The i7-3770 is not good value at all, if you're looking for value you go for an i5-34x0 or i5-35x0, or for high-end go for an i7-3930k. The i7-3770 does not fit in very well, the i7-3820 offer better value.

    I've been comparing i7-2600 with i7-3930k for workstations, and even at low and medium loads the i7-3930k handles well idling mostly at 1.2 GHz, while the i7-2600 is jumping up in frequency a lot more, even though it's lowest clock is 1.6 GHz. I believe most power users run more applications simultaneously, and while not all applications utilize all six cores, the six cores of i7-3930k is still beneficial for running normal loads cool (e.g. running a web browser, GIMP, a music player and even virtualbox while idling).

    Comment


    • #62
      Originally posted by pingufunkybeat View Post
      Well, they're approaching parity with Ivy Bridge right now, with current tech.
      http://hardocp.com/article/2012/10/2...overclocking/1
      Clock for clock, they still have a lot of ground to cover. Granted, intel doesn't have 4GHz parts, but judging by the way the current chips overclock, I don't think they have to sweat much to get there. Oh, and intel chips aren't genuine 8 core designs, they're 4 cores with HT, which by itself puts a dent into performance.

      Comment


      • #63
        Originally posted by bug77 View Post
        And there I was thinking users were just firing up a browser, Word and Excel sometimes and a game when they had time on their hands. Silly me, apparently the average user is a prodigious engineer/artist these days.
        The average user will be playing Mp3s from a collection (probably being organised by a database backend) while browsing the web, which means 10 to 20 tabs open, which means 10-20 threads, many of them running flash or something and buffering from youtube while the user is reading something else. And they probably have an anti-virus running on the background and a chat program and a p2p client downloading 3-4 movies minimised somewhere in the taskbar. They might even sync their smartphone or MP3 player with their collection at the same time, which takes a while, so they minimise that. You don't need fluid dynamics to need good multi-threaded performance.

        The user who runs one process which then needs 300 GFLOPS of sustained single-thread performance is rather the exception. Browser, Word and Excel and most games run fine on 5-year old stuff.

        Comment


        • #64
          Originally posted by bug77 View Post
          http://hardocp.com/article/2012/10/2...overclocking/1
          Clock for clock, they still have a lot of ground to cover. Granted, intel doesn't have 4GHz parts, but judging by the way the current chips overclock, I don't think they have to sweat much to get there. Oh, and intel chips aren't genuine 8 core designs, they're 4 cores with HT, which by itself puts a dent into performance.
          AMD chips aren't genuine 8 core designs either. Both Intel's HT cores and AMD's "modules" are somewhere inbetween a single core and two separate cores. AMD's "module" is a bit closer to two separate cores than an Intel core, though.

          Comment


          • #65
            Originally posted by droidhacker View Post
            Am I the only one who noticed that virtually all of the tests where the intel had any significant lead, were tests where the software optimization would necessarily exclude one cpu to the advantage of the other?
            This cuts both ways, and in general the newer processor will always lose. It takes time to get "-march native" working for a new CPU, so of course the compiler will be better tuned for one that is already supported. There was an article a while back about "-march bdzver3", but I presume that is still for older chips than this.

            Oops, I responded to the wrong post. I meant to respond to post #55 by pingufunkybeat. It's still along the same lines as this, but a bit more specific. To amplify this post, what the heck does "-march=native" even mean for a new, unsupported CPU. At this point gcc is looking at the CPU, using some sort of information to make a guess about what it is, and picking that. In that light, it would have been really handy to tell what the real CFLAGS in use ended up being.
            Last edited by phred14; 10-23-2012, 06:47 PM.

            Comment


            • #66
              Originally posted by pingufunkybeat View Post
              AMD chips aren't genuine 8 core designs either. Both Intel's HT cores and AMD's "modules" are somewhere inbetween a single core and two separate cores. AMD's "module" is a bit closer to two separate cores than an Intel core, though.
              Yes, I find this to be a funny development.. First they started marketing "more is better" with Mhz, but then changed it so that it really had nothing to do with the performance of the chip.... and now they're doing the same thing with cores...What's next? Performance based on how many pins your CPU has? If so, Intel is screwed because they keep losing a pin or two every generation.. haha.

              Comment


              • #67
                Originally posted by phred14 View Post
                To amplify this post, what the heck does "-march=native" even mean for a new, unsupported CPU. At this point gcc is looking at the CPU, using some sort of information to make a guess about what it is, and picking that. In that light, it would have been really handy to tell what the real CFLAGS in use ended up being.
                Sure, gcc can't optimise well for these processors, but I understood his post to imply that the benchmarks were optimised for one or the other. In this particular case, they were optimised exactly for the processor they were running on, only the AMD processor (by virtue of being a completely new architecture) cannot be optimised as well atm.

                Comment


                • #68
                  Originally posted by Sidicas View Post
                  Show me your numbers... Right now....


                  Don't spread bullshit.

                  Power here in New York is much more expensive than it is in the rest of the USA.
                  The CPU uses what? 48W more power?
                  Using the PC 8 hours a day...

                  48 * 8 = 384Whr more power per day

                  30 days a month...
                  384Whr * 30 days = 11.52kWh per month

                  Cost per kWh here in New York...
                  (https://www.nationalgridus.com/niaga...our_charge.asp)
                  57 dollars for 1000kWh (1MWh).... and that's assuming I use it during the worst time of day.

                  Extra electrical cost to use the AMD CPU rather than the Intel CPU per month...
                  11.52kWh / 1000kWh = x/$56.63
                  X=$0.65

                  Whoopie, we saved 65 cents a month.


                  Price difference between CPUs...
                  AMD FX-8350 = $220 (NewEgg.com)
                  Intel 3770K = $320 (NewEgg.com)
                  Difference = $100

                  How many months is it going to take to pay for the CPU price difference?
                  $100 / $0.65 = 153 months

                  How many years is that?
                  12.75 years

                  But wait, there's more... Intel motherboards are generally more expensive than AMD boards as well.
                  Maybe an extra 50 dollars more. So we need to add another 6 years onto that....

                  Of course, I love advanced fab chips, but when companies like Intel intentionally make their chips and motherboards overpriced by so much, any kind of power savings goes flying out the window instantly.
                  now calculate again with the german price : 0,25euro per 1kw.

                  11.52kWh per month=138,24kWh per year

                  138,24*0,25=34,56

                  now your price difference: 150dollar is ~ 120 euro

                  now 120euro/34,56euro= 3,47222222222 years.

                  in fact if you use your super new PC 3,5 years the intel one is cheaper.

                  Comment


                  • #69
                    Originally posted by phred14 View Post
                    These (parent and grandparent post) are some of the more significant posts I've read. Intel is capable of excellent technical work as they have proven many times, and are currently proving with the entire Banias - through - Ivy Bridge line. But they also have a strong marketing division that frequently seems to work at odds with directions that chip-head users present here might like. It is important to realize that from a corporate decision making point of view, it appears that at Intel, marketing trumps technical design.

                    Hence not that long ago we got the "Oooh, Fast!" NetBurst designs, and the "clone proof" IA64 designs. Because the Core-X line has been hammering AMD to the edge of existence, we're now seeing "revenue maximizing" stunts like disabling on-chip features unless you've paid extra. Kick AMD all you like, but if they're gone Intel has proven multiple times that, absent meaningful competition, they wander way off-target and we the customers lose. I don't know if ARM will provide proper competition for Intel in the future, at least partly because Microsoft has managed to lock down ARM-based hardware to be Windows-only, preventing it from growing into a true general-purpose platform.
                    don't know what to do with your words as a feedback to my words: "unlocked cpu+virtualization+ECC non-reg ram."

                    my words mean if you want a pc with ALL virtualization funktions +unlocked CPU+ECC RAM you need to buy AMD because intel just don't sell products like this.

                    Comment


                    • #70
                      Threading Monster

                      Looking at the results here and elsewhere I conclude AMD has a multithreading monster on it's hands. Single thread not so great but improved over BD.

                      I had hoped that Bulldozer would be the One, and I think this CPU is about 90% of the One.

                      Keep in mind that this arch was from word go targeting server & HPC workloads, and not spotty faced overclocking boys shooting at 3D monsters.

                      Comment


                      • #71
                        Originally posted by SavageX View Post
                        No, this is measured on the 12 Volt CPU supply line. It includes only the CPU and voltage regulators, so not *everything* is CPU power consumption (the voltage regulators do not ooperate at 100% efficiency), but certainly most is.
                        A quick bit of math:
                        *CPU is really 125W
                        *12V line is 168W
                        if those are true, then the VRM eff is 74.4%. That seems a bit low, but not crazy where I dismiss it outright. Considering the temps the VRMs on my AM3 run I won't be surprised if they need to get rid of ~40 watts.

                        Heise.de does not state the actual voltage on the 12V line (it's not exactly 12V) nor do they specify the quality of measuring device, or how it was measured. Is it +/-3%? +/-5%? +/-0.001%? what sort of voltage drop is there over the ammeter? Were they accidentally over-volting the CPU? what is the 12V ripple (get out your ocilliscope to measure that) and use that to calculate the RMS voltage. Anyways, there is enough possible error in here that the real TDP could be 125W or so and we have no way of knowing for sure.

                        Comment


                        • #72
                          In defend of heise.de

                          I think the other poster that mentioned the heise.de article, did not provide all information
                          that is included in the original article.
                          (in German: http://www.heise.de/ct/meldung/AMD-A...i-1734298.html )

                          - All tests were done using Cinebench
                          - They wrote that the CPU used up to 168 Watts, while talking -indeed- about TDP
                          (but also said that the voltage (so called bucket) converters draw some power).
                          - They said that "125 Watt TDP ist pures Wunschdenken" meaning that, 125 Watt
                          TDP is plain wishful thinking on AMDs side.

                          So, at least for me, heise says that "using the 12V CPU supply connector as sample point on our
                          mainboard this particular AMD CPU used well over 125 TDP which AMD claims
                          it has". They did not mention however, what their test rig exactly is. I would
                          very much like to know what the integration time for power measurements
                          were.

                          (and by the way: Given above is correct, I agree to the OP, that AMD's CPU suck, powerwise. :-D
                          )

                          Mind you that TDP is per (Intel) definition a MAXIMUM figure. As otherwise
                          you weren't be able to properly design and dimension a cooling solution for
                          a given processor. Note: Power drawn might be higher than TDP for a (very) short
                          time if the average power is =TDP.

                          Greeting from Germany, and please forgive bad English you encounter.
                          Cheers
                          Last edited by multics; 10-23-2012, 10:02 PM.

                          Comment


                          • #73
                            Originally posted by necro-lover View Post
                            now calculate again with the german price : 0,25euro per 1kw.

                            11.52kWh per month=138,24kWh per year

                            138,24*0,25=34,56

                            now your price difference: 150dollar is ~ 120 euro

                            now 120euro/34,56euro= 3,47222222222 years.

                            in fact if you use your super new PC 3,5 years the intel one is cheaper.
                            Again, you are probably saving power going with the AMD chip because of the lower idle power usage. Very few people run their machines at 100% and then shut them off immediately when finished. Most people are running at idle 99% of the time.

                            The reason a high TDP is bad news, is because it indicates that AMD is clocking the chip up as fast as they possibly can, meaning there may not be a lot of extra headroom easily available under this design. But as far as this chip itself, that shouldn't be a problem.

                            Comment


                            • #74
                              Originally posted by cynyr View Post
                              A quick bit of math:
                              *CPU is really 125W
                              *12V line is 168W
                              if those are true, then the VRM eff is 74.4%. That seems a bit low, but not crazy where I dismiss it outright. Considering the temps the VRMs on my AM3 run I won't be surprised if they need to get rid of ~40 watts.

                              Heise.de does not state the actual voltage on the 12V line (it's not exactly 12V) nor do they specify the quality of measuring device, or how it was measured. Is it +/-3%? +/-5%? +/-0.001%? what sort of voltage drop is there over the ammeter? Were they accidentally over-volting the CPU? what is the 12V ripple (get out your ocilliscope to measure that) and use that to calculate the RMS voltage. Anyways, there is enough possible error in here that the real TDP could be 125W or so and we have no way of knowing for sure.
                              - I asked somebody from heise, in the heise forum, whether they are sure that
                              only the cpu is fed from the 12V cpu power supply. He says yes. He even
                              said that sometimes small parts of the cpu (e.g. DRAM drivers) get power from the /other/ power
                              supply connector, which means the real cpu power might even be higher than
                              this.

                              - The VRMs (in this case bucket converters) do certainly much(!) better than 74.4%.

                              - "Heise.de does not state the actual voltage on the 12V line" => they use a
                              http://www.zes.com/english/products/...zer-lmg95.html
                              for measuring. Which means they do a proper "multiply volts by amps at any given time giving watts".

                              And here is how they do it:
                              http://www.heise.de/ct/artikel/c-t-A...l-1339902.html

                              Regards
                              Multics.
                              Last edited by multics; 10-24-2012, 04:14 AM.

                              Comment


                              • #75
                                These are measurements from Anandtech:




                                Intel gets the same job done in about the same time, but the whole system consumes about half the power. Idle power consumption is also lower.

                                Comment

                                Working...
                                X