Slashdot Mirror


Phenom IIs, Core I7-920 Win Out In Value Analysis

An anonymous reader writes "We've all seen processor benchmarks, but how do today's enthusiast CPUs look when you account for performance per dollar? Using a smorgasbord of charts, scatter plots, and performance tests, The Tech Report attempted to single out the highest-value offerings out of 16 popular Intel and AMD processors. The results might surprise you: AMD's 45nm Phenom IIs (both triple- and quad-core) prove to be strikingly competitive with Intel's Core 2 Quads. And, on the high end, Intel's $266 Core i7-920 turns out to be a compelling step up despite the higher costs of Core i7 platforms in general."

214 comments

  1. Suprise? by AnonGCB · · Score: 5, Insightful

    Really, who doesn't know that AMD is higher performance per dollar.

    --
    http://CryoLANparty.com/ A lan I'm staff on!
    1. Re:Suprise? by nicolas.kassis · · Score: 4, Insightful

      Dollars are meaningless GIVE ME RAW SPEED!!!!

    2. Re:Suprise? by dogmatixpsych · · Score: 1

      My wife doesn't know that (she wouldn't care anyway). That's at least one person who doesn't know. ;)

    3. Re:Suprise? by alcmaeon · · Score: 0

      Are they Phenom II's or Phenom IIs's?

    4. Re:Suprise? by geekoid · · Score: 2, Insightful

      She cares because the Intel chips run cooler, and therefore the box will be quieter.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    5. Re:Suprise? by Anonymous Coward · · Score: 5, Informative

      Phenom II's is a singular possesive.
      Phenom IIs's is a misspelled plural possesive.
      Phenom IIs' is a properly spelled plural possesive.
      Phenom II is a singular.
      Phenom IIs is a plural.

    6. Re:Suprise? by diskis · · Score: 3, Informative

      2001 called, they want their AMD Thunderbird 1.4GHz back.

      Sorry, but Intel has taken the lead in the hottest CPU, 150W for the QX9775, versus 125W for the Phenom II 940.

    7. Re:Suprise? by Anonymous Coward · · Score: 0

      Actually AMD lost the price/performance crown a while back. So they have it back again at the moment, but it would be rash to assume that AMD > Intel is always true. That is what benchmarks are for.

    8. Re:Suprise? by Zashi · · Score: 4, Funny

      I love you.

      --
      Skiffy is Spiffy, but Ort is tort.
    9. Re:Suprise? by vux984 · · Score: 3, Informative

      Phenom IIs is a plural.

      Yes, yes, we all know that from grade school. But what happens when AMD launches a "Phenom IIs"

      Is the plural Phenum IIss? And even if I write Phenom IIss, most of the readers will immediately wonder if AMD has released a "Phenom IIss"...

      The reason the apostrophe is inappropriately used is because it works. When people see Phenom II's, they pronounce it correctly -and- the ambiguity is removed as to whether the "s" is part of the name. The fact that its grammatically/semantically incorrect is an acceptable (to most people) tradeoff.

      Language changes, even grammar, and 's appened to product names and abbreviations is becoming understood to mean the plural of a 'non-word'.

    10. Re:Suprise? by itschy · · Score: 4, Funny

      Phenom IIs is a plural.

      Yes, yes, we all know that from grade school. But what happens when AMD launches a "Phenom IIs"

      The reason the apostrophe is inappropriately used is because it works. When people see Phenom II's, they pronounce it correctly -and- the ambiguity is removed as to whether the "s" is part of the name.

      That is, until AMD releases the "Phenom II's"

    11. Re:Suprise? by itschy · · Score: 5, Funny

      Damn.
      We need an escape character for natural languages!

    12. Re:Suprise? by Anonymous Coward · · Score: 0

      It's not the speed of your processor that counts, it's the software that's running on it!

    13. Re:Suprise? by Anonymous Coward · · Score: 0

      The Core i7 is significantly hotter than the core2 QX9775. My Qx9775 idles at 28 degrees centigrade and load at 50- 60 on air and my Core i7 920 when overclocked to the same frequency idles around 30 and load around 70 on Water... My core i7 raises my room temperature by a couple degrees...

    14. Re:Suprise? by etymxris · · Score: 2, Insightful

      That's why you should stick to Core 2 Duos, which run at 65W, and are performance competitive with the latest AMD chips in 99% of use cases (quad core only has real value to "power" users that do things like transcoding, ray tracing, run multiple VMs, or run server apps).

      The i7 performance advantages just don't seem to be worth the doubling (or more) of TDP. The Phenom IIs don't fare much better.

    15. Re:Suprise? by fnj · · Score: 1

      Yep. The E6850 rocks. All around, it has never been bettered. They should have kept making the 57 Chevy BelAir, and they should have kept making the E6850.

    16. Re:Suprise? by geekoid · · Score: 1

      knee jerk much?

      I apologize for not using the proper punctuation. lets try this:

      She cares because the Intel chips run cooler, and therefore the box will be quieter~

      That said, I notice you chose the QX9775 when this is about the i7.
      If you want to play that game, then AMD is crap because there Phenom II 940 runs way hotter then Intels 286.

      And just for clarification, other factors go into how how a chip gets beside Wattage.
      Package design, heat sink, fan, placement of the bridge, etc . . .
      And from what I can tell, Intel i7 pulls a few more Wats then the Phenom II.

      While I don't buy AMDs becasue I dislike how they have treated cusomters and I prefer to buy from the local boys, as it where.

      Also, AMD sound like WMD, and I don't support terrorists~

      oh, and 1999 called, they want their joke back.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    17. Re:Suprise? by Anonymous Coward · · Score: 3, Insightful

      Dollars are meaningless? And this is rated +5 Insightful? I knew the economy was bad, but I didn't knew it was THAT bad! (That or /. is some kind of billionaire's club.)

    18. Re:Suprise? by LWATCDR · · Score: 1

      Then you should be running a Power5. Actually most people don't need the fastest CPU. If you do then get the top of the line I7 or a two socket workstation motherboard.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    19. Re:Suprise? by geekoid · · Score: 1

      Well, those features not sued today will be used a lot more in the next 18 months, or so. Since I build computers to get 5 years of gaming out of them, the top end is a better deal for me.
      That plan has always worked well for me.

      I'll be building my next computer just before Star Craft II comes out.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    20. Re:Suprise? by geekoid · · Score: 1

      At which point the market execs should be drug out of there 2 martini lunch and flogged.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    21. Re:Suprise? by Swizec · · Score: 2, Informative

      Phenom II's is a singular possesive. Phenom IIs's is a misspelled plural possesive. Phenom IIs' is a properly spelled plural possesive. Phenom II is a singular. Phenom IIs is a plural.

      You are wrong. II's is proper spelling as per the rules of English grammar.

      An apostrophe is used to form a plural for abbreviations, acronyms, and symbols where adding just s rather than 's may leave things ambiguous or inelegant

      http://en.wikipedia.org/wiki/Apostrophe#Use_in_forming_certain_plurals

    22. Re:Suprise? by jellomizer · · Score: 1

      Not always Back in 2006 Intel Core 2 Duo chip actually beat AMD.

      --
      If something is so important that you feel the need to post it on the internet... It probably isn't that important.
    23. Re:Suprise? by blind+biker · · Score: 1

      Really, who doesn't know that AMD is higher performance per dollar.

      What he says.

      Seriously: surprise? Even a moron with learning disabilities living under a rock in a country without Internet access knows that Phenoms offer the most performance per buck.

      --
      "The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
    24. Re:Suprise? by default+luser · · Score: 1

      The 45nm Core2 processors are even cooler - they have a TDP of 45w, and the average processor uses much less.

      Please note: the measurements above are of processor power consumption only, not total system consumption. As you might expect, only the top-end 45nm Core2 procesors consume anywhere near their 45w TDP, just as only the top-end 65nm Core2 processors consume near 65w.

      As for the i7, it only performs well in perfectly-multithreaded benchmarks that are completely I/O-limited (like the aggressive multi-pass .h264 encoder), otherwise it performs about as well as a Core2 Quad. And when you consider the starting price for an i7 motherboard is around $200-250, you have to be nuts to consider purchasing such a monstrosity.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

    25. Re:Suprise? by FreonTrip · · Score: 1

      It's worth noting that the recent Brisbane core Athlon X2s also have a 45W TDP. Most of what you've mentioned also applies there.

    26. Re:Suprise? by CaPn+Corelian · · Score: 1

      Shouldnt plural be Penomeae II or something?

    27. Re:Suprise? by Anonymous Coward · · Score: 0
    28. Re:Suprise? by Just+Some+Guy · · Score: 1

      But what happens when AMD launches a "Phenom IIs" Is the plural Phenum IIss?

      No. It's Phenom IIses.

      Language changes, even grammar, and 's appened to product names and abbreviations is becoming understood to mean the plural of a 'non-word'.

      However, it's still considered a sign of illiteracy among educated people.

      --
      Dewey, what part of this looks like authorities should be involved?
    29. Re:Suprise? by Anonymous Coward · · Score: 0

      The correct spelling is "possessive" - with twice two s, or two s's or two ss...

    30. Re:Suprise? by Anpheus · · Score: 1

      I disagree, purchasing an i7 now as opposed to a Core 2 Quad makes a great deal more sense as long as you're the sort of person who upgrades their computers themselves. If you don't ever swap a CPU, if you don't care about memory latency, or a whole mess of other things, then, really it doesn't matter what you buy. Buy a Phenom II rig, they're certainly Good Enough(tm) for whatever you want to do.

      On the other hand, if you want to be able to upgrade your computer and you do want that extra performance over the competition, buying i7 now is the only thing that makes sense. The LGA-1366 is the new "workstation" class Intel socket. There will be a lower pin count variant for desktops, and there will be a higher pin count variant for servers and multiple-processor systems. So if you need higher performance now, and you want to be able to upgrade, you either get an AM3 board (Phenom II) or you get an LGA-1366 board (i7.)

    31. Re:Suprise? by Anonymous Coward · · Score: 0

      While you're here can you explain the correct way to write nVidia at the start of a sentence.

    32. Re:Suprise? by Anonymous Coward · · Score: 0

      Sorry, but no, you are wrong. There exists no lowercase Roman characters, therefore it is not ambiguous or inelegant. Also, I would argue that your wikimedia quote is wrong; as I have never seen a case where writing something like "I have two PC's." is acceptable. In fact, between 3 different spell checkers it tells me that using the apostrophe there is wrong. Correcting it to "I have two PCs." yields no error on checking.

    33. Re:Suprise? by alexo · · Score: 1

      The reason the apostrophe is inappropriately used is because it works. When people see Phenom II's, they pronounce it correctly -and- the ambiguity is removed as to whether the "s" is part of the name. The fact that its grammatically/semantically incorrect is an acceptable (to most people) tradeoff.

      What's wrong with qute marks?
      "Phenom II"s seems to work.

    34. Re:Suprise? by vux984 · · Score: 1

      No. It's Phenom IIses.

      Touche.

      But how do I differentiate the plural of IIs from the plural of the IIse?

      And its not a hypothetical question either...

      There is a Porsche 911, plural 911s
      There is also an actual 911S. The plural form of that would be 911Ses
      But even worse, there is a 911SE, which in plural is also 911SEs
      For completeness there is also a 911E, plural 911Es.
      As far as I know there is no 911ES... yet.

      Personally I think most people agree its practically impossible to keep the 911S separate from all the 911s, or all the 911Ses separate from the 911SEs. Especially since some products don't capitalize the product letters like is usual with Porsche.

      Like the Apple IIe for example. If someone talked about the Apple IIes at least half of us would think it was a new 'es' model.

      However, it's still considered a sign of illiteracy among educated people.

      I'd think intelligent people would be the first to recognize the inadequacy and ambiguity perpetrated by traditional plural forms to abbreviations and product code names. Oh... you said 'educated' not 'intelligent'. Your probably right about 'educated' people.

    35. Re:Suprise? by vux984 · · Score: 1

      "Phenom II"s seems to work.

      Nothing is wrong with that. How about 'Phenom II's in single quotes instead of double? That would work too. Of course why bother with the leading single quote? A single quote in the one place its acutally needed to delineate the end of the word from the plural construction is shorter and works just as well.

      And besides, the 'apostrophe' serves multiple functions in the english language. That of marking a possessive. And that of marking omissions. (You used it in this form when you wrote "What's" instead of "What is".)

      It is also used to help sort out awkward inflections. KO'd for example is a recognized way of writing the past tense of 'Knock Out' appears in the Oxford english dictionary as such.

      Its use in something like "Phenom II's" is, in my opinion, entirely consistent with both the latter 2 uses I mentioned for the apostrophe. It assists in resolving the ambiguity that arises when attaching a plural inflection to an abbreviation or product code. It also reflects an omission in that something like the plural of a hyptothetical Phenom IIs that could be expressed as Phenom IIses, but as Phenom IIs's the apostrophe serves both to assist with the inflection, and covers the omission of the e as well.

    36. Re:Suprise? by Anonymous Coward · · Score: 0

      Is the plural Phenum IIss?

      No, it's Phenom IIses.

    37. Re:Suprise? by Anonymous Coward · · Score: 0

      Then what's the plural possesive? Phenom II's'? Phenom II's's? Or is it just ambiguous as to whether Phenom II's is possesive or not?

    38. Re:Suprise? by StikyPad · · Score: 1

      The use of apostrophes to pluralize nouns is ambigious at best -- there are conflicting authoritative opinions on whether or not they are proper. Basically, unless you're writing a term paper and your prof. has an opinion one way or the other, it doesn't matter.

    39. Re:Suprise? by caitsith01 · · Score: 1

      My god, you're citing Wikipedia about grammar???

      An apostrophe is used by some writers to form a plural for abbreviations, acronyms, and symbols

      And those writers would be totally, utterly wrong, as all of the examples cited in that article demonstrate.

      --
      Read Pynchon.
    40. Re:Suprise? by Anonymous Coward · · Score: 0

      The TDP ratings are not interchangeable. Intel's CPUs have always been closer to the theoretical maximums than AMD's. With P4's it got so bad that many people considered them "averages" rather than "maximums".

      That 125w Phenom II probably only consumes 45 watts when in a game that uses two cores. To hit 125w you really have to slam every core with heavy number crunching, like you might see in scientific calculations.

      Intel's chipsets have also tended to use more watts - in part because of components like the memory controller, which weren't on-die before.

      The result of the skew in measurements was a 100w TDP Intel CPU consuming 75-150 watts depending on the task, and a 100w TDP AMD CPU consuming 50-100 watts depending on the task. Significantly lower, once you factor in the chipset.

    41. Re:Suprise? by UnknownSoldier · · Score: 1

      And in the real world, not everyone has an infinite budget, nor care to waste money for negligible gains.

    42. Re:Suprise? by leromarinvit · · Score: 1

      Is the "you" singular or plural?

      --
      Proud member of the Ferengi Socialist Party.
    43. Re:Suprise? by Tubal-Cain · · Score: 1

      Well, those features not sued today...

      Been reading too many RIAA articles lately?

    44. Re:Suprise? by Anonymous Coward · · Score: 0

      AMD hasn't launched a "Phenom IIs", so your entire post is stupid and irrelevant.

    45. Re:Suprise? by Anonymous Coward · · Score: 0

      shhh they are in fanboy trance.

    46. Re:Suprise? by gfody · · Score: 1

      what about shanghai opterons which can be had for $179 and work in dual socket mobos? I wonder where a pair of shanghai's would fall on their scatter chart..

      --

      bite my glorious golden ass.
    47. Re:Suprise? by Anonymous Coward · · Score: 0

      Perhaps you mean possessive?

    48. Re:Suprise? by Anonymous Coward · · Score: 0

      Several companies are happy to sell you raw power. The SGI Altix 4000 systems and the IBM POWER6 based servers are two examples.

      A SGI Altix Linux server can be delivered with up to 131072 GB RAM and 4096 CPU cores in one physical server.

      There are several other companies that also sells you some interesting solutions if you can pay for it.

    49. Re:Suprise? by heson · · Score: 1

      I've designed so many systems with the possibility to upgrade in the future. Future comes, and all cpus more than marginally better are not supported by the mobo (due to speed or slot), or the old type of ram supported by the mobo is now rare and very expensive. Maybe Im unlucky or unskilled.

    50. Re:Suprise? by owlstead · · Score: 1

      And they can have it! I've still got one lying around. That said: that TDP is becoming the norm nowadays. Gosh, that was certainly the loudest processor fan I've ever used. Got it traded in for a x64 as soon as it became affordable.

    51. Re:Suprise? by toddestan · · Score: 1

      Actually, it's probably a bad idea to buy anything Intel at the moment if you care about upgradability. Intel has already announced plans to create a new, lower pin socket for the desktop version of the Core i7 processors, and yet another new socket with more pins for the server-class hardware. There really isn't room for three sockets here, and being in the middle, I see the role that LGA-1366 plays being squeezed at both ends, with the result that it will get absorbed into the other sockets. As for LGA-775, I don't see much new coming out for it on the high end, though I suppose you could alway start out with a budget processor today knowing that you could drop in a Core 2 Quad at a later date. If you can wait, I would wait for the lower-pin desktop version of Core i7 (Core i5?) and buy one of those as it will probably be around a while.

    52. Re:Suprise? by Anpheus · · Score: 1

      LGA-1366 is promised to continue at least on to the 6-core, 32nm "Gulftown." At the very least, Gulftown will be at least another 50% improvement for highly threaded applications.

    53. Re:Suprise? by PCAdept · · Score: 1

      UUUHHH... Those wattage numbers refer to the maximum amount of power the cooling system for a CPU is required to dissipate, and have absolutely nothing to do with a CPU's clock speeds or ability to do calculations. o.O http://en.wikipedia.org/wiki/Thermal_Design_Power

    54. Re:Suprise? by default+luser · · Score: 1

      The last time I upgraded a processor was with my Celeron 300a (450 overclocked) to a Celeron 533A (800 overclocked), and that was a special point in history:

      At this time, computers were still "slow" for most tasks, so people were always hungry for more processing power. The BX chipset brought a %50 increase in bandwidth, and the on-die cache of the celeron made chipset bandwidth even less of a factor than usual. ATA/33 was fast enough for most hard drives for several years after release, and add-in cards were cheap. The move from slot to socket meant that there was a huge demand for slotkets, so building-in support for the newer flip-chips was an easy add-on. Finally, the i820 chipset meant to replace the BX chipset had no performance advantage, and required expensive RDRAM.

      This bore us a market ripe for upgrades. This market does not exist today.

      In my current machine purchased a year ago, I bought a midrange Core2, with the capability to upgrade to Core2 Quad in the future. So far, I'm not feeling the desire - very few games support quad-cores, and all of these games run smoothly on my Core2. I don't do much video transcoding, so there's really no drive for me to double my cores.

      --

      Man is the animal that laughs.
      And occasionally whores for Karma.

  2. One question by Jurily · · Score: 4, Funny

    which one is more more secure?

  3. Mistake in TFS by Evanisincontrol · · Score: 2, Insightful

    And, on the high end, Intel's $266 Core i7-920 turns out to be a compelling step up despite the higher costs of Core i7 platforms in general.

    TFA says that the Core i7-920 is $284; the chip below it (The Core 2 Quad Q9550) is $266. It's still up there on the performance/price scale, though.

    1. Re:Mistake in TFS by Anonymous Coward · · Score: 1, Informative

      There are other costs involved in a Core i& system. You'll end up paying $150 extra on a motherboard that supports i7, for example.

    2. Re:Mistake in TFS by jgtg32a · · Score: 1

      And RAM

    3. Re:Mistake in TFS by Anonymous Coward · · Score: 0

      ddr3 + x58 board kills any deal on a 920 until there are mobos in the 150$ price range and ddr3 in the 100-150$ for 6gig.

    4. Re:Mistake in TFS by Timothy+Brownawell · · Score: 1

      ddr3 + x58 board kills any deal on a 920 until there are mobos in the 150$ price range and ddr3 in the 100-150$ for 6gig.

      6GB DDR3-1066 @ $81
      6GB DDR3-1333 @ $94 (on sale for $85)
      6GB DDR3-1600 @ $100 (on sale for $90; needs 1.65v instead of 1.5v)

      But it looks like i7 mobos currently start at $200 ($190 cheapest on sale), so maybe you're half right.

    5. Re:Mistake in TFS by Anonymous Coward · · Score: 0

      the x58 chipset for core i7's are really expensive motherboards in the 300$ price range. which is 6x as much as a cheap 50$ option for phenom IIs.
       
      plus, don't forget your only option for these motherboards for ram is ddr3, which is at a mininum 2x as much.
       
          your basic setup is going to be around 300$ for a phenom II , mobo + 4gb of ram. vs 600->700$ for a i7, mobo and ram.

  4. Bottom LIne by cybrthng · · Score: 4, Insightful

    Its about your investment.. For me Phenom II was a no brainer because of AM2+ compatibility. Once newegg put those suckers at 200 bucks i jumped. Its like i have an entire new PC and that was upgrading from the 9600 quad core.

    Oddly enough i didn't have complaints about the performance of the 9600.. i just figured encoding times and processing times would be reduced enough that it would accelerate my work and well, for 200 bucks its done so and more so than i expected.

    i7 is a nice platform but i'm penny pinching right now and looking for better ROI vs bragging rights.

    1. Re:Bottom LIne by gad_zuki! · · Score: 4, Insightful

      >i7 is a nice platform but i'm penny pinching right now and looking for better ROI vs bragging rights.

      I wonder if penny-pinching will be more common with the economic downturn. If AMD can price itself lower then ROI will be very tempting, even if the Intel product is faster.

    2. Re:Bottom LIne by __aatmkh9910 · · Score: 2, Insightful

      I agree, me(and most other consumers) will buy whichever chip gives them the most GHz for their buck, it doesn't matter who makes it.

    3. Re:Bottom LIne by Anonymous Coward · · Score: 1, Insightful

      I have an AMD 3800 X2 processor machine. I *never* had a problem with the processor's speed being my limiting factor in term of work getting done. *Never*.

      Vast majority of consumers (including businesses) of computers are no longer limited by processor speed. It is not 1990s anymore. Any cheap processor will run your shit adequately provided you have 4GB RAM or more.

    4. Re:Bottom LIne by IorDMUX · · Score: 1

      I just upgraded to a Phenom II 940 last night (I still haven't had time for benchmarks). One of the best features, for me, is that I was able to drop it right into my 1.5 year old AM2+/DDR1066 motherboard, upgrade the BIOS, and be on my way.

      Though ROI and convenience weighed heavily, for me, there is one more bit of philosophy that affected my decision:
      If Intel and AMD were to offer me equal and balanced options, I would purchase AMD, so that next time I go looking for a CPU, I still have the choice.

      --
      >> Standing on head makes smile of frown, but rest of face also upside down.
    5. Re:Bottom LIne by INowRegretThesePosts · · Score: 1

      I agree, me(and most other consumers) will buy whichever chip gives them the most GHz for their buck, it doesn't matter who
      makes it.

      First, equating clock frequency with performance makes me cry (I'm an Electronic Engineer with an interest in computer architecture).

      Second, you should consider value per buck and not just bang per buck.

      For example, suppose that I could get 5% better capacity/dollar with a 400GB HD than with a 200GB HD. Since 200 GB is enough for me, buying the 400GB would basically mean paying 90% without enough benefit. Buying the 400GB is more bang per buck but less value per buck.

      And the justification that I should buy the 400GB because "I will need it in the future" is just a shallow excuse for consumerism. When/if I need the additional 200GB, it will be cheaper, faster, and more reliable (due to being less used) than now.

      Third, I also give importance to company behavior and not just to the final price. I prefer AMD not only because its processors usually offer better performance/dollar and because motherboards for AMD processors are often cheaper, but also because of the anti-competitive behavior of Intel.

    6. Re:Bottom LIne by __aatmkh9910 · · Score: 1

      First, equating clock frequency with performance makes me cry (I'm an Electronic Engineer with an interest in computer architecture).

      Second, you should consider value per buck and not just bang per buck.

      For example, suppose that I could get 5% better capacity/dollar with a 400GB HD than with a 200GB HD. Since 200 GB is enough for me, buying the 400GB would basically mean paying 90% without enough benefit. Buying the 400GB is more bang per buck but less value per buck.

      And the justification that I should buy the 400GB because "I will need it in the future" is just a shallow excuse for consumerism. When/if I need the additional 200GB, it will be cheaper, faster, and more reliable (due to being less used) than now.

      I know GHz != FLOPS, and that Hz/FLOPS in CISC processors can vary a lot, but I find that lot of the time, hertz are still a decent measurement of power.
      Also, with respect to bang for your buck, I'm talking within reasonable bounds(a few hundred MHz difference)

    7. Re:Bottom LIne by Anonymous Coward · · Score: 0

      You know it,
      I priced out a pretty sweet gaming rig the first was an Intel, Nvidia rig and the second was an AMD/ATI rig.. after at least a month comparing the hardware, reading reviews, and testing demos at local gaming stores I chose the AMD/ATI format and saved almost $500. Add to that the motherboard I chose is setup to support the 6/8/12 core CPUs of the future.. and this baby should last for at least five to six years.. (last gaming rig lasted 5 years)

      Here is the Intel System
      ------- Intel/Nvidia PC ----------
      COOLER MASTER HAF 932 RC-932-KKN1-GP Black Steel ATX Full Tower Computer Case - Retail Model
      $149.99

      ASUS P6T Deluxe LGA 1366 Intel X58 ATX Intel Motherboard - Retail Model
      $289.99

      EVGA 896-P3-1262-AR GeForce GTX 260 Superclocked Edition 896MB 448-bit GDDR3 PCI Express 2.0 x16 HDCP Ready SLI Supported Video Card
      $189.99

      SILVERSTONE ST1200 1200W ATX12V / EPS12V SLI Ready CrossFire Ready 80 PLUS Certified Modular Active PFC Power Supply - Retail Model
      $299.99

      Intel Core i7 920 Nehalem 2.66GHz LGA 1366 130W Quad-Core Processor Model BX80601920 - Retail Model
      $279.99

      mushkin 3GB (3 x 1GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Triple Channel Kit Desktop Memory Model
      $219.98

      Western Digital Caviar Black WD5001AALS 500GB 7200 RPM SATA 3.0Gb/s Hard Drive - OEM Model #:WD5001AALS
      $139.98 (2)

      LG 22X DVD±R DVD Burner Black SATA Model GH22NS30 - OEM Model
      $24.99

      Arctic Silver 5 Thermal Compound - OEM Model
      $8.99

      XIGMATEK ACK-I7363 CPU Cooler -
      $9.99

      XIGMATEK Dark Knight-S1283V 120mm Long Life Bearing CPU Cooler
      $39.99

      Saitek PZ30AU Black USB Wired Standard Eclipse Keyboard - Retail
      $34.99

      RAZER Copperhead Tempest Blue 1 x Wheel USB Wired Laser Gaming Mouse - Retail
      $54.99

      Microsoft Windows Vista Ultimate SP1 64-bit for System Builders - OEM
      $179.99

      Total: $1,923.84

      --------- AMD/ATI PC -------
      COOLER MASTER HAF 932 RC-932-KKN1-GP Black Steel ATX Full Tower Computer Case -
      $149.99

      ASUS M4A79T Deluxe AM3 DDR3 AMD 790FX ATX AMD Motherboard - Retail Model
      $189.99

      SAPPHIRE 100270SR Radeon HD 4850 X2 2GB 512-bit (256-bit x 2) GDDR3 PCI Express 2.0 x16 HDCP Ready CrossFire Supported Video Card
      $279.99

      SILVERSTONE ST75F 750W ATX 12V 2.2 & EPS 12V SLI Ready Modular Active PFC Power Supply -
      $144.99

      AMD Phenom II X4 810 2.6GHz Socket AM3 95W Quad-Core Processor Model HDX810WFGIBOX - Retail Model #:HDX810WFGIBOX
      $175.00

      CORSAIR XMS3 DHX 4GB (2 x 2GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Dual Channel Kit Desktop Memory Model TW3X4G1600C9DHX
      $98.00

      Western Digital Caviar Black WD5001AALS 500GB 7200 RPM SATA 3.0Gb/s Hard Drive - OEM Model #:WD5001AALS
      $139.98 (2)

      LG 22X DVD±R DVD Burner Black SATA Model GH22NS30 - OEM Model
      $24.99

      COOLER MASTER Hyper Z600 RR-600-LBU1-GP CPU Cooler - Retail Model #:RR-600-LBU1-GP Item
      $59.99

      XIGMATEK Dark Knight-S1283V 120mm Long Life Bearing CPU Cooler
      $39.99

      Saitek PZ30AU Black USB Wired Standard Eclipse Keyboard - Retail
      $34.99

      RAZER Copperhead Tempest Blue 1 x Wheel USB Wired Laser Gaming Mouse - Retail
      $54.99

      Microsoft Windows Vista Ultimate SP1 64-bit for System Builders - OEM
      $179.99

      Subtotal: $1,537.89

      Conclusion:

      I bought the AMD/ATI solution and had enough left over to buy a brand new widescreen 22-inch monitor..

      So I bought a used copy of Crysis and BOOM baby plays the game with everything maxed!

  5. Value overload by Anonymous Coward · · Score: 0

    I for one welcome our new "Bang for the Buck" overlords.

  6. No Brainer by XPeter · · Score: 1, Insightful

    It's been like this for the past few years. If your on a budget, go with AMD. If you have a little more dough to spend, go with Intel.

    --
    "The difference between genius and stupidity is that genius has it's limits" - Albert Einstein
  7. This is a TERRIBLE comparison by junglebeast · · Score: 5, Interesting

    It is wrong to compare performance/price because this assumes price scales linearly with performance, which is clearly false. Nobody expects to get 50% more performance when they pay 50% more. But if there is a $100 process having a performance of 1000, then we would normally consider it an excellent deal if we could pay $150 for a performance of 1300. The value for your money therefore scales in a non-linear way, and it's better to just have everyone look at the scatter plot and choose their own price point based on their personal internal scaling function. The core i7 has the greatest discontinuity in jumping ahead of the rest of the crowd in this regard.

    1. Re:This is a TERRIBLE comparison by slashkitty · · Score: 4, Interesting
      Actually, this is a very important comparison. If you're considering any sort of clustering, like what I do, you absolutely need it. Do you need 100 opterons, or 4 core i7's?

      In the cluster I run, I've been quickly swapping out old xeons for new Core i7's. With just the 4 920's I have running, I've been able to remove 20 old xeons, all while improving the overall performance of the cluster.

      Price / Performance also helps you judge how fast the computer will be antiquated. If you now need only 500 gflops, and this computers offers 1000, you know that it should sustain you into the future.

      --
      -- these are only opinions and they might not be mine.
    2. Re:This is a TERRIBLE comparison by Anonymous Coward · · Score: 0

      price/performance IS useful if you are planning to spend at that price. most people have budgets and many will have a limit to the amount they can spend. Also, the scale is definitely non-linear but it's also not one directional. Newer procs may have much better ROI due to better technology advances but it could very well be worst due to having to pay for being an early adopter. Performance/Price pretty gray with random deals (especially online) too but it's still useful as a reference. It is still better to do your own comparison before purchase though especially for those on tight budgets.

      There is also the issue of depreciation where newer products tend to drop is price early on in it's life. Spending 500 bucks every two years is quite a bit different then spending 1000 every 4 years.

    3. Re:This is a TERRIBLE comparison by Anonymous Coward · · Score: 0

      Nobody expects to get 50% more performance when they pay 50% more.

      Why not? There are quite a few pairs of processors on the scatter plots that satisfy this. Some even give you more performance if you pay less.

    4. Re:This is a TERRIBLE comparison by Kjella · · Score: 1

      I think his general point is that most of us are only able to use one primary computer at a time, you can have supporting ones like file server etc. but we humans don't scale well. So if I got 1000$ to spend and the choice is between Intel at 1000$ and AMD at 500$ then my only real options are to buy the Intel or buy AMD and pocket the 500$. Buying two AMD machiens might net me more gigaflops than the Intel but I don't have four arms and 30fps on each screen is not the same as 60fps. I know that I'm quite a bit past the sweetspot of price/performance and more on my price/utility sweetspot where I can't justify spending more on a computer, but it's all along the one computer axis.

      --
      Live today, because you never know what tomorrow brings
    5. Re:This is a TERRIBLE comparison by lopgok · · Score: 1

      I am not sure what you are doing with your cluster, but I am sure your cluster is now unreliable. It is most likely your old slow xeons had ECC memory. They might even have had chipkill like all the opterons and phenom II's have. But without ECC, I suspect you have a very big, somewhat hot computer that has a bunch of random errors.

      Making a bunch of assumptions, 20 xeons, 4gb of ram each, means 80gb of memory. Assuming you still are using 80gb of memory, and with IBM's estimate for memory errors (1gb = 1 error per week), you will see have about 80/7, or 11 errors per day. That is almost an error every 2 hours.

      There was once a 'cluster' made of 1u macs. Apple ended up replacing every one with a newer model with ecc. It is true that there are fault tolerant algorithms that can deal with low error rates, but if aren't coding with them, I would have little confidence in the correctness of your output.

      Perhaps the i7 is cool for a gamer, but when I paid my money, I bought a phenomII 940, because it has chipkill ecc and background memory scrubbing (it scrubs all memory every 8 hours). This is pretty much mandatory for reliable computing, let alone a cluster.

    6. Re:This is a TERRIBLE comparison by INowRegretThesePosts · · Score: 1

      Read the article. They don't merely divide performance per cost and proclaim the processor with the largest ratio as the best buy. They provide all the data, advise that the price of other computer parts should be accounted for, advise about the power consumption, and then pick *a set of* processors that seem particularly attractive.

      I agree that buying based on "bang per buck" is irrational, and specially irrational when "bang" is a single metric that does not come close to measuring the real value. For example, a 400GB HD may have 5% better capacity/dollar than a 200GB HD, but since 200GB is enough for me, buying the 400GB would pretty much be overpaying by 90%. Not to mention that speed and reliability also count.

      But the article does not advocate buying based on "bang per buck". Read it, or at least the first and final pages.

  8. Price is all-important by BadAnalogyGuy · · Score: 1, Interesting

    None of that is important. Any modern x86 CPU is going to have enough performance for anything you want to do. You'd get more benefit shaving a baby's ass than squeezing the cost/performance ratio on these chips these days. Better to throw more money at a separate server if you really need more power than trying to boost the speed of any single computer.

    Really, the thing that will make the biggest difference is the OS, but if you're running any modern OS you're already wasting most of those CPU cycles on platform overhead.

    1. Re:Price is all-important by 644bd346996 · · Score: 5, Insightful

      Buying a faster chip is a lot cheaper and faster than rewriting something to be multithreaded.

    2. Re:Price is all-important by bami · · Score: 2, Insightful

      But buying a faster multi-core (as in, 3 or more cores) chip isn't going to do you any good if your application only runs on one or two threads.

    3. Re:Price is all-important by ArcherB · · Score: 1

      Buying a faster chip is a lot cheaper and faster than rewriting something to be multithreaded.

      Mostly true, but if the OS is multithreaded, and anything beyond WindowsME is, you can run multiple separate non-multithreaded apps on a single machine and get better performance from each of them.

      I say "MOSTLY" true because there are situations where it is better to go ahead and rewrite the application, especially if you are running several thousand copies of it on multiple machines. When it comes time to upgrade your systems, it will be cheaper to upgrade to 500 multicore processors than 1000-2000 single cores.

      --
      There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
    4. Re:Price is all-important by ArcherB · · Score: 4, Insightful

      But buying a faster multi-core (as in, 3 or more cores) chip isn't going to do you any good if your application only runs on one or two threads.

      Very true if your system only runs that single application. However, everyone I know runs multiple applications just by booting their OS.

      --
      There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
    5. Re:Price is all-important by SuperAndy · · Score: 2, Insightful

      Which is even more the case if you virtualise an operating system.

    6. Re:Price is all-important by Anonymous Coward · · Score: 0

      But the actual choke in an application still usually occurs when a single app (explorer typically) goes balistic on something and obsesses on it at 100% of a thread. The initial advantage of multi core was to allow (other) windows to not freeze in this situation, so you could still save your work before rebooting. However, these days most single apps still choke on 100% of a single thread.

      To revist the GP:

      Buying a faster chip is a lot cheaper and faster than rewriting something to be multithreaded.

      That's no longer really an option. The faster chips are pretty much hitting a speed ceiling (at least with silicon). In order to utilize the new chips applications have to be multi-threaded. My only hope (and this should be borne out by cost-benefit analysis) is that multi-threading only happens where it's absolutely needed. Specifically I'm looking at you explorer. Every ass that event attempts to multi-thread the explorer.exe should going to get hunted down and gutted to make an example.

      multi-CPU's/Cores/Threads are wonderful things that protect us from poorly written software (ie: windows).

    7. Re:Price is all-important by geekoid · · Score: 2, Insightful

      multithreading is not that difficult, and there are ways to multithread an app without rewriting. Granted it's hackish, but then so is almost all of software.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    8. Re:Price is all-important by Waffle+Iron · · Score: 4, Informative

      However, everyone I know runs multiple applications just by booting their OS.

      But not everybody knows what their system is doing. I do: I always keep a CPU load chart in my KDE taskbar, and for interactive usage I probably use one core less than 3% of the time, and both cores about 0.1% of the time.

      If I am transcoding a video, then one core gets pegged. However, I would never notice load on the processor on even with a single-core CPU if I just ran the transcode under "nice". It might take 3% longer to run because it waits for the interactive stuff, but that's insignificant.

      But I don't run transcoding on my workstation anyway. Why? Because all the I/O continuosly flushes out my disk buffers for other processes. That makes my interactive apps seem slower than crap anyway because they have to hit the disk every time some of their data or program image gets flushed out. My dual cores do nothing to address that issue. I run transcode jobs on a server box where they won't bug anybody.

      The only place dual core would really help most people with typical single-threaded apps is if they run at least two copies of programs heavy on number crunching but light on I/O and memory bandwidth, like $FavoriteCause@Home. Other than that, people will have to wait for multithreaded user apps to get much real-world benefit out of multicore CPUs.

    9. Re:Price is all-important by robthebloke · · Score: 1

      The only place dual core would really help most people with typical single-threaded apps is if they run at least two copies of programs heavy on number crunching but light on I/O and memory bandwidth

      Not true, one core runs at 100% for the app you're trying to run, the other runs at about 90% with crapware, virus scanners, limewire or whatever crap the kiddies are installing on their PC's these days. As far as your average user goes, they probably will see a doubling of performance...

    10. Re:Price is all-important by Tycho · · Score: 2, Interesting

      There is another barrier that we will eventually hit. The current process schedulers in use on modern operating systems have a problem. Attempting to use more than roughly 38 logical processors will result in the additional processors either waiting to run the process scheduler, waiting for a memory access, or waiting for I/O. Currently this is sidestepped on mainframes using virtualization and low latency I/O. I have a hard time seeing how virtualization or lower latency I/O could be adapted for use in desktop computers. Running Windows Aleph-Null, MacOS X 12.8, and Linux 2.8.1853 would probably not have much appeal to average users that would still need the power a desktop computer offered at that point. Intel, the main driving force of the PCISIG keeps pushing newer versions of PCI-Express that do not address the interconnect's inability to play nice with multiple masters, providing any type of packet routing, or deal in any way with its absurdly high latency, which is over 100ns even with PCIe 1.1, and gets worse with each newer version. Some sort of low latency sideband channel would work, but figuring how to maintain backward compatibility with current PCIe cards and motherboards is not easy. Instead Intel has added only DRM features, but no actual security for the computer's user on the bus itself (think Firewire and writing to whatever memory locations you want).

      Bad form ahead:

      If embryonic stem cell research does not make you uncomfortable, you have not thought about it enough. --James Thompson

      Better:
      If embryonic stem cell research makes you uncomfortable, you have thought about it far too much. Try researching the actual potential feasibility of the scare stories, and consider that adult stem cells have never been made totipotent, only pluripotent.

      Yeah sorry about that.

      --
      Impersonating Tycho from Penny Arcade since before there was a PA.
    11. Re:Price is all-important by AcidPenguin9873 · · Score: 1

      There's obviously a trade-off. For some problems, buying a faster chip is MUCH lower-return than rewriting something to be multithreaded, and you must weight that with the cost of rewriting.

      Buying a faster chip gets you on average about 10% speedup. But if your application gets, say, a 200-300% speedup on a quad-core processor (from being multithreaded and having nearly-linear scaling behavior as you add cores), it might be worth the investment. Depends what you're doing - how much parallelism is inherent in the problem, how much it costs for a someone to parallelize it, and how much you value having the results sooner.

    12. Re:Price is all-important by Anonymous Coward · · Score: 0

      Sounds like you need to get a better application for transcoding your videos. Say one that is actually multithreaded.

    13. Re:Price is all-important by foobsr · · Score: 1

      Any modern x86 CPU is going to have enough performance for anything you want to do.

      Rendering ???

      CC.

      --
      TaijiQuan (Huang, 5 loosenings)
    14. Re:Price is all-important by Attila+Dimedici · · Score: 0, Offtopic

      In response to your "bad form--better" If embryonic stem cells are so much better why are there not yet any treatments that use them and yet there are multiple treatments using adult stem cells?

      --
      The truth is that all men having power ought to be mistrusted. James Madison
    15. Re:Price is all-important by Ant+P. · · Score: 2, Interesting

      But I don't run transcoding on my workstation anyway. Why? Because all the I/O continuosly flushes out my disk buffers for other processes.

      Which is exactly the reason why posix_fadvise(2) exists.

    16. Re:Price is all-important by serviscope_minor · · Score: 2, Informative

      Attempting to use more than roughly 38 logical processors will result in the additional processors either waiting to run the process scheduler, waiting for a memory access, or waiting for I/O.

      [Citation needed]

      SGI sold systems with 128, 256, 512 and even 1024 processors running a single Linux image.

      --
      SJW n. One who posts facts.
    17. Re:Price is all-important by Tubal-Cain · · Score: 1

      Pfft. Even though I don't like embryonic stem cell research, I don't like your analogy. Here's a counterexample:

      If adult stem cells are so much better than organ transplants, why aren't we using adult stem cells?

      Mostly, the answer is time and money. Transplants have gotten more research $$$ and a lot more time to work out the details of practical application than stem cells have. Similarly, adult stem cells promise much of the benefits and carries few of the stigmas attached to embryonic, so it gets all the money and effort. And, as a result, breakthroughs.

    18. Re:Price is all-important by dbIII · · Score: 1

      This is not really aimed at the above poster but at the same attitude expressed by others.
      There's a python developer in my company that turned off the second core in his laptop's CPU so that his code would run reliably. It's not the fault of python or the CPU, it's just somebody that needs to be dragged screaming into the 1990s or possibly even the 1960s when people wrote a lot about race conditions. Multiple CPUs are in my nintendo DS and are probably in a few mobile phones by now, and soon they will be in most desktop PCs - it's time to learn how to write code that can do something with it.
      Even if your code can't do anything with the extra CPUs it allows other CPU bound stuff to run at the same time. There are also a lot of tasks that are trivial to run in parallel - just about anything involving images or video or 3D representations of things.

    19. Re:Price is all-important by dbIII · · Score: 1

      None of that is important. Any modern x86 CPU is going to have enough performance for anything you want to do.

      No. When you get a faster CPU you just find more things you want to do with it.

    20. Re:Price is all-important by lvv · · Score: 1

      It is not just flushing the buffers but also swaping out now-inactive (interactive) processes. When I run some streaming app, I just disable swap. Also ionice and "sysctl -w vm.swappiness=1" might help too.

    21. Re:Price is all-important by 644bd346996 · · Score: 1

      Good point. There's no excuse for application code breaking in a multiprocessor environment.

      By the way, the piece of code I had in mind with my above comment is a simulation that's been developed, optimized, and debugged as a single threaded app over the past few years, with several million dollars poured into the project. The code may seem inefficient by being single threaded, but in this application, getting the right answer extremely important. With the economy the way it is, we can't afford to do the debugging that would be necessary to validate a multithreaded version. (That is, unless we get some of the stimulus money, which is still up in the air.) We're far from the only company in a situation like this.

    22. Re:Price is all-important by ps2os2 · · Score: 1

      Believe or not there are studies that show (not in a VM environment) that the practical #'s of CP's limit is around 15.
      The curve just about goes flat at that time and stay that way no matter how many cp's you add (assuming non VM). The speed is there but sooner or later you have to use another CP for *SOMETHING* and there goes thru put. I am not familiar with how INTEL has its instruction buffers and the logic they employ to decode the instructions and address's but *IF* they do not do a good job in second guessing the next instruction that is coming up aqain the speed in the world will not help you if you have to flush the instruction buffer

    23. Re:Price is all-important by ps2os2 · · Score: 1

      That is part of the issue with INTEL is that their chips are not really set to run at 100 percent . Some manufactures chips and OS are designed for thruput and to run at 100 percent (all the time) so they are efficient. They also use computer resources wisely and give a lot more bang to the buck than INTEL.

    24. Re:Price is all-important by Tycho · · Score: 1

      Okay, here you go:

      http://en.wikipedia.org/wiki/Software_lockout

      I should have looked that up instead of depending on a vague recollection of the article, but this is still /. in any case. The machines SGI sold that you describe almost certainly did not resemble the average design of a current x86 PC in any way, regardless of the CPU architecture SGI used. That was no standard Linux Kernel either. Worse yet a system like that would be nearly useless for normal PC tasks like surfing the web, running office applications, or playing demanding 3D games. The current record holding general purpose supercomputer, an IBM Blue Gene system, uses Linux as well. IBM's Roadrunner is great, if the Cell processors do the work you need them to do, which is not the case most of the time. At any rate, on Blue Gene systems a single copy of Linux is run per dual core processor and standard DDR2 RAM is used as well. No attempt is made to micromanage each Linux image.

      --
      Impersonating Tycho from Penny Arcade since before there was a PA.
  9. Up front costs are a fraction of total costs by Anonymous Coward · · Score: 5, Informative

    Until electricity is free, comparing CPUs based on up-front cost of the CPUs alone ignores a major part of the expense of owning and operating computers, particularly if you're running servers.

    But that's okay, Slashdot. I understand that you live in your parents' basement and you don't pay for electricity anyway.

    1. Re:Up front costs are a fraction of total costs by Anonymous Coward · · Score: 0

      I live in my parent's attic you insensitive clod!

    2. Re:Up front costs are a fraction of total costs by Anonymous Coward · · Score: 0

      Does your other parent have a basement you could use?

    3. Re:Up front costs are a fraction of total costs by fnj · · Score: 1

      Yep. A piece of crap Intel or AMD system running at 200 watts 24x7 uses 1750kwh per year, which at 20 cents per kwh comes to $350 per year. If it lasts 5 years, you're looking at a total power bill of $1750.

      We need new blood. This garbage is pathetic. It's producing 99.9% heat and 0.01% computing.

    4. Re:Up front costs are a fraction of total costs by PitaBred · · Score: 1

      The thermal envelopes in most current processors aren't that different until you start looking at the extreme top end, and if you're looking in that price range, you don't really care about the cost anyway. Besides, a couple bucks a month isn't that much of a deal for most people, the initial capital outlay and overall performance are MUCH more important... it really only becomes a serious concern when you're running datacenter numbers of machines.

    5. Re:Up front costs are a fraction of total costs by Ant+P. · · Score: 1

      These CPUs aren't aimed at people who have a bunch of beige boxes running as servers. They're better for gamers and virtualisation.

      If you want a low-power standalone server, buy VIA. They've been doing low power chips with hardware crypto for years.

    6. Re:Up front costs are a fraction of total costs by yoshi_mon · · Score: 2, Insightful

      ...running at 200 watts 24x7...

      Unless your running SETI or some other setup then it's doubtful that your CPU is pegged at 200W 24x7. But lets for the sake of argument say that for whatever reason that some setup has it's CPU pegged at max/near-max all the time. What then is the value of the productivity?

      Now that's a huge question and something I'm not even going to being to answer but just wanted to point out that putting out a lot of big numbers without context is pretty silly.

      --

      Really, I know what I'm doing...Ohhhh, look at the shiny buttons!
    7. Re:Up front costs are a fraction of total costs by Anonymous Coward · · Score: 0

      Phenom II has lower power consumption too.

  10. AMD price : performance linear by jbeaupre · · Score: 4, Insightful

    What stood out to me is that AMD seems to have a fairly consistent price:performance ratio. Is this policy?

    Most of their offerings fall pretty close to a line (not quite a zero crossing, but close). If this holds true for all their current and future offerings, you don't have to have test metrics for every processor. You can use price as a reasonable estimate of performance. i.e. Double the price gets you twice the performance.

    Intel on the other hand, you can't trust price to indicate performance. A lot more research is involved. OR else you have to assume there's a high likelihood that the AMD offering for the same price will be better.

    --
    The world is made by those who show up for the job.
    1. Re:AMD price : performance linear by giverson · · Score: 5, Insightful

      It's because Intel dominates the high end. AMD can't sell a processor with a premium pricetag because its performance would compete with Intel's midrange which is priced pretty reasonably.

      AMD is the loveable underdog, but don't forget how expensive their X2s were when they were dominant. AMD isn't cheap because they're doing us a favor, they're cheap because they have to be.

      --

      Capitalism does not lead to corruption, lack of character does.
    2. Re:AMD price : performance linear by Bill+Dog · · Score: 5, Insightful

      Indeed. Something that stood out to me (from that scatter graph) is:
      * If you don't want to spend more than $100 on a CPU, AMD wins.
      * If you don't want to spend more than $150 on a CPU, AMD wins.
      * If you don't want to spend more than $200 on a CPU, AMD wins.
      * If you don't want to spend more than $250 on a CPU, AMD wins.
      * If you don't want to spend more than $300 on a CPU, Intel's (cheapest) i7 wins by far.

      --
      Attention zealots and haters: 00100 00100
    3. Re:AMD price : performance linear by Kjella · · Score: 3, Interesting

      AMD is the loveable underdog, but don't forget how expensive their X2s were when they were dominant. AMD isn't cheap because they're doing us a favor, they're cheap because they have to be.

      That's a thing that people don't seem to get - prices are what they must be in the market. The question is, can you skim off enough to keep designing new chips and developing your foundries? Already they've failed at the latter and is trying a huge bet trying to make a foundry company spin-off. No matter how badly they're really doing, in the "here and now" they'll be competitive right up until they file for chapter 11.

      --
      Live today, because you never know what tomorrow brings
    4. Re:AMD price : performance linear by paziek · · Score: 0

      Yeah, as long as you already have mobo + RAM for i7, but then - why would you buy yet another i7?

    5. Re:AMD price : performance linear by citizenr · · Score: 1

      the most important is missing. All this info assumes people dont overclock. Intel wins at every price range when you overclock. Currently all Ibtel chips can run at 4GHz with little to no effort.

      --
      Who logs in to gdm? Not I, said the duck.
    6. Re:AMD price : performance linear by fnj · · Score: 2, Insightful

      Now run your comparisons based on 64 bit code instead of 32 bit. Intel shits the bed running in 64 bit mode.

    7. Re:AMD price : performance linear by funkatron · · Score: 3, Funny

      Intel shits the bed running in 64 bit mode.

      Is that a good thing?

      --
      "Welcome to our world. We are the wasted youth. And we are the future too." Yes, I know these are stupid lyrics.
    8. Re:AMD price : performance linear by Anonymous Coward · · Score: 0

      I was wondering the same :D

    9. Re:AMD price : performance linear by TwistedSymmetry · · Score: 1

      Intel shits the bed running in 64 bit mode.

      Is that a good thing?

      Funkatron is obviously not running in common sense mode.

    10. Re:AMD price : performance linear by Anonymous Coward · · Score: 0

      Intel shits the bed running in 64 bit mode.

      So AMD ftw? Or do we hate beds, and Intel ftw?

    11. Re:AMD price : performance linear by INowRegretThesePosts · · Score: 1

      AMD is the loveable underdog, but don't forget how expensive their X2s were when they were dominant. AMD isn't cheap because they're doing us a favor, they're cheap because they have to be.

      The people who advocate AMD do not think that AMD would sell cheap processors even without competition. In a monopoly, they would most likely charge as much as the market could pay.
      The point is that AMD is a very positive force in the market, and I buy from them not only because their processors have better value/dollar (and ditto for the motherboards for these processors) than Intel's, but also because I and every other consumer would be in a much worse situation if not for the little competition that still exists in the x86 market*.

      A second point is that Intel has used anti-competition practices, and should be punished for that.

      * Then again, I would certainly prefer not to be obligated to buy x86 processors in the first place. I see ARM advancing, and I love it. I wonder that with ARM's advance, cross-platform will dominate, and this could allow even other architectures to compete, making me smile like a baby. Not to mention that Free Software is far better at cross-platform than proprietary software; Free Software helps underdog arches, and underdog arches help Free software, in a positive feedback loop of market competition.

  11. Best performance per dollar ... by psergiu · · Score: 1

    Best performance per dollar: the 486 i got for free. Do the math yourself :)

    --
    1% APY, No fees, Online Bank https://captl1.co/2uIErYq Don't let your $$$ sit in a no-interest acct.
    1. Re:Best performance per dollar ... by jbeaupre · · Score: 5, Funny

      0/0 gives me a headache. Have to keep carrying the zero.

      --
      The world is made by those who show up for the job.
    2. Re:Best performance per dollar ... by Archangel+Michael · · Score: 2, Informative

      But but but but ....

      Any number divided by itself = 1

      THEREFORE 0/0 = 1

      See, maths is easy!

      --
      Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    3. Re:Best performance per dollar ... by alexhs · · Score: 2

      But but but but ....

      Zero divided by any number = 0

      THEREFORE 0/0 = 0

      See, maths is contradictory!

      ( Those mods that didn't #DE or #Z already will now #MF )

      --
      I have discovered a truly marvelous proof of killer sig, which this margin is too narrow to contain.
    4. Re:Best performance per dollar ... by Archangel+Michael · · Score: 1

      Wikipedia has all the answers!!!!

      "Nullity, a pseudo-mathematical concept proposed by James Anderson, defined as the value 0/0 and represented by the character Φ"

      --
      Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    5. Re:Best performance per dollar ... by Ant+P. · · Score: 1

      Zero divided by any number = 0

      Any number divided by itself = 1

      THEREFORE 0/0 = (0 && 1) = 0

      See? Easy.

    6. Re:Best performance per dollar ... by Anonymous Coward · · Score: 0

      0/0 is, if treated as a legitimate operation, equal to every number at the same time.

      Take for example: ab=c

      If this is true, then a=c/b

      However, if we substitute in 0 for b and c...

      0a=0, which is obviously correct. But if we divide by zero here...

      a=0/0

      Except you can substitute any number you want for a and the equation will still be correct!

      Division by Zero is undefined. It's not that we don't know what it is, it just doesn't have any meaning.

      Division is a function. Just like every function, it has a domain (a set of inputs) and a range (a set of outputs). When we say division by zero is undefined, we say that, for all a, the result of D(a,0) is not within the range of D(a,b).

      Now, we can't just make up a value for a division by zero and just keep going, since division by zero being undefined is actually quite important. If you try to define it, a lot of things we're used to in mathematics suddenly stop working, like fields. Of course, you can always build a new mathematics on a new set of axioms under which a division by zero is defined, but...

  12. Let's stop making reviews for gamers by Xtravar · · Score: 4, Interesting

    I recently had to make the tough choice of a Phenom 2 vs Intel Core Quad. I went with the Intel because I somehow came to the conclusion that they run cooler.

    You see, I'm building a recording PC, so I want to have as few fans as possible. I plan on having a huge heatsink with NO fan. Most reviews, if they focus on heat, focus on the overclocking aspect.

    If wattage correlates to heat like I think it does, I may have been better off with a Phenom 2. But, then again, the wattage test was only run during one task in this review. I read another review where it was different.

    There just aren't enough review sites out there for... ahem... "grown ups". Maybe I should start one that takes a look at performance with DAWs like REAPER.

    In the end, I don't care about best performance per dollar, or wattage per dollar. I care about performance per degree of heat, because heat = noise. Performance of modern CPUs is good enough these days.

    Oh well, that's my rant of the day.

    --
    Buckle your ROFL belt, we're in for some LOLs.
    1. Re:Let's stop making reviews for gamers by diskis · · Score: 4, Informative

      You didn't google enough.
      These guys are really anal when it comes to finding quiet parts. Following their advice, I now actually have an overclocked PC, that I can't hear if it's on or off.

    2. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      Are you kidding? I visit that site every day. While they're great, they don't really have as large of a hardware selection as most review sites. Also, they're still (as you've demonstrated) somewhat geared toward the gamer/overclocker/max performance crowd.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    3. Re:Let's stop making reviews for gamers by jbeaupre · · Score: 2, Interesting

      I'm curious what you think of the Atom processors. Passive cooling seems to be a nice benefit if you can live with less performance.

      --
      The world is made by those who show up for the job.
    4. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      Didn't even know about Atom processors. I just did a quick search and it seems like they're for an entirely different market segment. Although, I'll have to keep that in mind if I ever build a HTPC.

      I was specifically looking for a quad core processor with this project, since I need to process effects on multiple audio streams at a time. The actual speed of those cores isn't too important to me as long as they're better than a Pentium 4. I also need relatively quick memory and the bus bandwidth/speed to input multiple audio streams.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    5. Re:Let's stop making reviews for gamers by powerlord · · Score: 1

      What do you mean by "Recording PC"?

      I'm not sure what you mean by quiet but if you need the least noise possible, look at the Atom architecture from Intel.

      If its Beefy enough then you could probably build a PC without a FAN (they use Atoms in most of the popular Netbooks which don't have fans, just some thermal material to passively cool the chips).

      Combine that with an SSD if you can afford one (again, depends on what you mean by "Recording PC"), OR some of the Western Digital "Green" drives (their 5400 RPM and run easy on power so their spin low and are relatively quiet), OR if you ABSOLUTELY need quiet in a specific area, try combining approaches. Use a smallish SSD to boot the local system, and feed the data to a big honking RAID in another room (possibly soundproofed if need be).

      Again, depending on how "beefy" a PC you need (and other requirements), some netbooks are relatively beefy and are incredibly quiet. Either think about using them, or learn from what they do. :)

      --
      This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
    6. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      While a real recording studio may not care so much about the noise (since they would just move the PC to another room), hobbyists have things a little difference.

      Anyway, my NewEgg shipment is already being delivered as we speak, so most of this discussion is moot. You can see more about the actual specifications I was looking for here: http://slashdot.org/comments.pl?sid=1168821&cid=27269123

      The point of my main post is... review sites are always geared toward getting the most performance regardless of how loud the computer has to be. In addition, most benchmarks focus on things that are not that important to real people.

      Another hiccup in my quest was finding a low-end video card. I wanted a separate video card because recording software/effects/drivers mostly only work in Windows XP 32-bit. Which means I am limited to 4GB ram. Which means I don't want to share memory with the video card.

      Well, of course, if you don't want on-board video then you MUST be a gamer seeking maximum performance!! Which means even the lowest low-end cards without fans are going to probably run very hot, raising the temperature of the entire case.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    7. Re:Let's stop making reviews for gamers by Zashi · · Score: 1
      Have you considered getting one or two PS3s? I'm not sure what audio processing software your running, but if you just need more cores you might be better off with a solution like linux clustering.

      You could also look into CUDA/OpenCL using one or more graphics cards.

      --
      Skiffy is Spiffy, but Ort is tort.
    8. Re:Let's stop making reviews for gamers by Fweeky · · Score: 1

      The last AMD system I built is quieter and uses about an order of magnitude less power than my PS3 (~30W idle; about the same as a friend's Atom board manages).

    9. Re:Let's stop making reviews for gamers by Alarindris · · Score: 1

      If you're doing serious recording, your computer should be isolated from whatever you are recording anyway.

      Noise shouldn't be an issue.

    10. Re:Let's stop making reviews for gamers by orzetto · · Score: 1

      If wattage correlates to heat like I think it does, [...]

      Er... in case you had doubts, wattage correlates with heat 100%. When energy enters a computer case, it can only leave as heat, except possibly for the small amount of airborne kinetic energy impelled away by the fans, which is anyway very small.

      If you are talking about temperature, you obviously have to consider both heat input (=wattage) and dissipation capacity. In your particular case, in which you want to minimise fan usage, low wattage is very advantageous. You should consider how CPU-intensive your typical recording application will be, and find the CPU that delivers that performance with the minimum wattage.

      If you really want to go for something radically silent, try heat pipes or liquid cooling, either convection-driven or with an acoustically insulated pump.

      --
      Victims of 9/11: <3000. Traffic in the US: >30,000/y
    11. Re:Let's stop making reviews for gamers by jargon82 · · Score: 1

      If you are going to use windows 32bit, and you are going to have 4GB ram, you are also going to give up that ram one way or the other. 32 bit consumer windows doesn't support PAE and as such you have 32bits (4GB) of addressing space. If you throw a 512MB graphics card in, it uses (at least) 512MB of that addressing space, even if it's not using the RAM. The most you can reliably count on using in 32 bit windows is a tad above 3GB.

    12. Re:Let's stop making reviews for gamers by fyleow · · Score: 1

      Silent PC Review catering to the gamer/overclocker/max performance crowd?? Are we visiting the same website?? There's an entire forum on that site dedicated to green and eco friendly computing. There are also recommendations to use slower but quieter 5400 RPM WD Green drives and notebook drives on desktops. That's hardly max performance. There are also plenty of threads about underclocking and undervolting processors to save power and heat.

    13. Re:Let's stop making reviews for gamers by fnj · · Score: 1

      I'm with you on power drain. Intel has continually shot themselves in the balls since the early Conroe era. The last good thing they did was adopt the superb development work that came from their Israel team. The T2500, T7200, and E6850 rocked; it's all been downhill since then. We're repeating the Pentium 4 gag-fest debacle, churning out space heaters that do a little computing on the side. They laid a giant shit egg with their 45nm process.

    14. Re:Let's stop making reviews for gamers by geekoid · · Score: 1

      A solution is to put a fan in a different room that circulates cool air into a 'closet' you put your computer in so is is silent during recording.
      Or put your computer in a different room, and then just remote into it from a less powerful computer in the studio.

      Or lease a computer housed in a different building and just remote into there.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    15. Re:Let's stop making reviews for gamers by Anonymous Coward · · Score: 0

      Well, of course, if you don't want on-board video then you MUST be a gamer seeking maximum performance!! Which means even the lowest low-end cards without fans are going to probably run very hot, raising the temperature of the entire case.

      Not necessarily. You said you weren't concerned with xHz/$. It may be worth your while to check them out. None of the cards on the linked page have fans. I've used a few different matrox cards in the past and have been very satisfied. They make reliable hardware, with good solid drivers.

      Of late I've been eyeing their DualHead2Go. Yes most laptops (these days) will let you work across both screens, but rarely can you get a desktop screen to:
      1. Sit evenly with the laptop LCD
      2. Be the same resolution at the same pixel density.

      I don't work for Matrox, I'm a public servant. I just really like Matrox's products.

    16. Re:Let's stop making reviews for gamers by fnj · · Score: 2, Insightful

      I'm curious what you think of the Atom processors

      The Atom is an efficiency crap fest. It's garbage. I'd rather have an Arm chip any day. Let's get away from this absurd undead i386 architecture garbage. Linux runs on any architecture.

    17. Re:Let's stop making reviews for gamers by geekoid · · Score: 2, Insightful

      What real difference do you get from over clocking.
      Don't get me wrong, I don't have anything against overclocking, and did it back when perfomance gains was going from Doom as a slide show to a smooth running game.
      I also understand overclocking for the sake of overclocking. But is getting 15% increase in MHz really noticeable without testing?

      And does it still ahve the same direct relationships now that a lot of work is being removed from the CPU, and the CPU's are tasking across several cores?

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    18. Re:Let's stop making reviews for gamers by fnj · · Score: 1, Troll

      The Atom is GARBAGE. Use an Arm chip.

    19. Re:Let's stop making reviews for gamers by A+Friendly+Troll · · Score: 1

      Are you kidding? I visit that site every day. While they're great, they don't really have as large of a hardware selection as most review sites. Also, they're still (as you've demonstrated) somewhat geared toward the gamer/overclocker/max performance crowd.

      I don't think you have ever visited that site.

      It's geared so much towards silent/quiet computing that they have their own anechoic chamber and a special system for testing fan airflow. They even test hard drives for vibration (subjectively, but still).

      "Most review sites" give absolutely no indication of hardware noise, and if they do, it's almost entirely worthless.

    20. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      Take a look at their latest CPU review article. It only compares the top tier processors. Their "CPUs and Motherboards" page is all... motherboards.

      I'm afraid that they just don't receive enough funds/free stuff from manufacturers to do a thorough job.

      Everyone on the forums is making a HTPC or a gaming rig. I don't have the patience or time to grep through all of that and abstract the necessary data.

      Yeah, I'm being picky. So? Like I said, this was my rant for the day!

      --
      Buckle your ROFL belt, we're in for some LOLs.
    21. Re:Let's stop making reviews for gamers by Jah-Wren+Ryel · · Score: 1

      I also understand overclocking for the sake of overclocking. But is getting 15% increase in MHz really noticeable without testing?

      Try closer to 50%. You have to pick the right model of cpu to get that kind of increase, but it is always the case that the best overclocking chips are in the cheap section. So it isn't like you have to pay for the privilege if you do it right.

      And yes it can be very noticeable, depending on what you are doing. For example, it can mean the difference between smooth video playback of HD video and a herky jerky mess of dropouts.

      --
      When information is power, privacy is freedom.
    22. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      Hi, friendly troll!

      Yes, all of those metrics are great. But where are their reviews/comparisons of the current line up of CPUs? You know... relating to how this whole topic started.

      There's lots of nice information on PSUs, fans, cases. That stuff's infinitely helpful, and for that I am indebted to silentpcreview.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    23. Re:Let's stop making reviews for gamers by aaronl · · Score: 1

      I took my i7 920 from 2.66GHz to 3.32GHz by upping my base clock from 133MHz to 166Mhz. This changes the QPI bus accordingly, making system transactions much faster as well. It cost me nothing over the original system cost to do this, and I didn't have to change voltages to make it work.

      This makes all single core bound tasks (which are frequent) substantially faster, and I spend far less money to do it.

    24. Re:Let's stop making reviews for gamers by A+Friendly+Troll · · Score: 1

      They don't measure performance, only power consumption and noise. Naturally, a CPU won't produce noise, but a fan will :)

    25. Re:Let's stop making reviews for gamers by regular_gonzalez · · Score: 1

      On CPU-bound tasks like transcoding, it scales 1:1. I have a Q6600 overclocked from 2.4 ghz to 3.0 ghz - a 25% bump which is all-but-guaranteed with this chip (even higher overclocks are easily attainable, the Core line overclocks ridiculously well). This results in a 25% reduction in transcode times. Whether this is worth it to you, I suspect, is dependent on how many cpu-limited tasks you do.

      --
      Due to circumstances beyond my control, I am master of my fate and captain of my soul.
    26. Re:Let's stop making reviews for gamers by Anonymous Coward · · Score: 0

      I have a Q6600 which runs at 2.4 Ghz stock. I overclock it to 3.6 Ghz.

      That's a 50% increase in clock speed (not your marginal 15% increase)

    27. Re:Let's stop making reviews for gamers by Xtravar · · Score: 1

      A CPU will produce the heat that will produce the noise! You need to know the size of the dog shit before you know what kind of pooper scooper to get.

      I think it would be rather advantageous to know which processor gets the most performance per unit of heat, don't you? Good! Then we're in agreement. In fantasy land, that data would also be relevant to real world usage scenarios instead of being synthetic benchmarks and games.

      --
      Buckle your ROFL belt, we're in for some LOLs.
    28. Re:Let's stop making reviews for gamers by diskis · · Score: 1

      Oh, there is a difference indeed. I have a Xeon 3210, stock 2.13GHz, clocked to 3.0 or so. I'm using Lightroom a lot, and difference between normal and overclocked is annoying lag vs. almost instantaneous response. Lightroom uses all 4 cores evenly, all of which gain 0.8GHz of processing power. That's 3.2 GHz free. Then again using Photoshop, which is singlethreaded, I won't notice the difference without a stopwatch.

      But, it doesn't matter. Overclocking is free anyways.

    29. Re:Let's stop making reviews for gamers by torkus · · Score: 1

      The 80's called, they want their recording studio back.

      Besides, not everyone *has* two separate rooms for PC equipment and a studio but modern PC hardware will still give them very, very nearly studio quality sound for an incredible price.

      --
      You can get rich if you own a politician, but you have to be rich to buy one in the first place.
    30. Re:Let's stop making reviews for gamers by hawk · · Score: 1

      *sniff*

      Obviously, *your* computer doesn't have bay-mounted gauss guns . . . or even a serious laser defense system!

      amateurs . . .

      hawk

    31. Re:Let's stop making reviews for gamers by MrKaos · · Score: 1
      I've sooo been where you are now. I had enough when the rendered music I produced and had frequency range where I eq'ed it louder because it happened to be the same range as the fan noise.

      blech

      Do yourself a favor and get yourself a TNN 500AF. Since I bought one of these a few years ago I have never looked back and my ears have been served well, even with 6 drives on board.

      Overclocking? No problem, and it's still quiet. Actually when I upgraded the motherboard (in 2006) in the machine I found that I had pushed it so hard the motherboard came out warped around the cpu area. Both have been intel chips wound up. Funny thing is I was looking at the i7 920 as my next upgrade with a gigabyte GA-EX58-EXTREME.

      The only issues I have with the case are 1. the power supply can sometimes have a mind of it's own but I hope Zalman have rectified the issues in the meantime (I intend to have a spare power supply handy for when the original unit dies). 2. when upgrading you need to order the motherboard rear mount thermal blocks well in advance of your upgrade. 3. It fuckin hurts if you stub your toe on it. 4. sometimes you can go to bed and forget that the machine is on, it really is that quiet - I love it for music production.

      But you are right though, music production is more demanding than game play, my music prod box handles games easily. In some respects I think the kernel tuning, for music production, makes the games run better than they do under windows - but I've never done anything but a cursory comparison. Many of my game player friends look at my music prod box with envy, and whilst I let them use it to play the odd game - I never let them put that shit through my carefully room eq'ed production monitors - this machine is for real work.

      Plus the case makes a great space heater during the winter ;-)

      --
      My ism, it's full of beliefs.
    32. Re:Let's stop making reviews for gamers by INowRegretThesePosts · · Score: 1

      But, it doesn't matter. Overclocking is free anyways.

      It costs a lot of time during many "does this combination of voltage and frequency work?" tests, and then a lot of torture tests to "validate" the end result. Also, even after you "validate" the end result, your computer might still crash during a peculiar compilation, transcoding, number-crunching, gaming or whatever, sending time and data to /dev/null. There are *a lot* of ways a CPU can be stressed, and a simple torture test will never catch it. Remember that a computer is a general-purpose computation machine *designed* to operate on a nearly infinite input space. Properly testing a processor requires an engineering team (and as an Electronic Engineer I can tell you there is specialized theory on this), proper equipment, and a lot of time.
      Not to mention that torture tests often merely stress the CPU with a tight loop and see if it crashes or not - they don't even check that the small loop body is correctly computed, not to mention a more complex calculation. A computer that silently miscomputes data is far worse than a computer that loudly crashes.

      The time you lose due to this probably negates the time you gained with your overclocking (remembering that even if you overclock your CPU by 30%, your computer doesn't become 30% faster - the HD, network and video card are still at the same speed, and often the memory is not overclocked as much as the CPU).

      Also, increasing the frequency will probably increase power consumption at least linearly, and increasing voltage will probably increase power consumption quadratically (that is, 15% greater voltage means 32.25% more electrical power) - although I don't know how this works with today processors where frequency and sometimes even voltage changes with demand.
      And in any case the torture tests will be particularly expensive.

      Also: the only CPU I have ever seen to be damaged was one I had overclocked (when I was a teen). Note, I didn't even increase voltage, only frequency, and the computer was stable. Even in this conditions the CPU was lost after some months (although it is not certain it was due to overclocking).

      Also: if you don't plan for overclocking, you hardly get a good benefit like a 40% clock increase. But if you plan for it, you run the risk of paying a premium for an overclock-friendly processor, motherboard, heatsink/fan, a powerful PSU and then... it doesn't overclock nicely. You don't know how much it will overclock until you try it, but at this point, you already spent your money.

      In summary: overclocking costs time, electricity dollars, stability, the risk of hardware damage, and probably money on overclock-friendly parts, without even knowing in advance if you will succeed.

      The last point: I haven't heard of a server admin, number-crunching cientist, office computers admin or any other professional overclocking a computer. But I do see teenagers doing it, then bragging about.

    33. Re:Let's stop making reviews for gamers by olman · · Score: 1

      I also understand overclocking for the sake of overclocking. But is getting 15% increase in MHz really noticeable without testing?

      You forget whole geek bragging rights-aspect.

      Another point worth keeping in mind is that the last 15% actually would cost you a whole lot of money. If you buy a CPU/GPU that is stock clocked 15% faster, you can easily end up paying 100% more.

      Which is of course why CPU manufacturers have gone to considerable trouble to stop you from overclocking. Anyone remember overclocking Athlons with a pencil? Or the silly superglue-method tom's hardware guide came out with to overclock the CPU that made pencilling impossible?

      Same with GPUs of course. And as everyone already pointed out, you can often get quite a bit more than "15%" more. My 3GHz Dual core is running at 4GHz, for example, just because I can. Low end parts tend to be more overclockable than high end.

    34. Re:Let's stop making reviews for gamers by fnj · · Score: 1

      To the drooling imbecile who thought that was a troll: read up, idiot. Anyone with a clue knows an Arm gets all the typical lightweight use done for around ONE TENTH the power consumption of the Atom, which is ... wait for it ... aimed at LIGHTWEIGHT USE.

  13. Only one benchmark? by kill-1 · · Score: 0

    Did anyone notice that the only benchmark they base their chart on is Cinebench? Coincidentally, this is a benchmark that makes the Phenoms look especially good. There is also no commentary whatsoever why they chose that benchmark. So IMO the article is completely useless unless you only use Cinema 4D on your machine.

    1. Re:Only one benchmark? by Anonymous Coward · · Score: 2, Informative

      They're just using that chart as an example to illustrate their ranking system. You see the little next page button at the bottom of the page?

    2. Re:Only one benchmark? by Anonymous Coward · · Score: 0

      You might want to read past page one.

    3. Re:Only one benchmark? by Anonymous Coward · · Score: 1, Insightful

      Try reading the whole article!
      Page 2: http://techreport.com/articles.x/16570/2

      They are Multimedia Centric but not just 1 benchmark...
      "We used the following versions of our test applications:

              * WorldBench 6 beta 2
              * Half-Life 2: Episode Two
              * Crysis Warhead
              * Far Cry 2
              * Unreal Tournament 3 1.3
              * Valve VRAD map build benchmark
              * Valve Source Engine particle simulation benchmark
              * Cinebench R10 64-bit Edition
              * POV-Ray for Windows 3.7 beta 29 64-bit
              * notfred's Folding benchmark CD 9/28/08 revision
              * The Panorama Factory 5.2 x64 Edition
              * Windows Media Encoder 9 x64 Edition
              * x264 HD benchmark 2.0 with x264 version 0.59.819
              * LAME MT 3.97a 64-bit
      "

    4. Re:Only one benchmark? by poetmatt · · Score: 1

      Really? I noticed they benchmark about 25 different programs or so, if you had read past the first page.

      Sheesh, why don't people ever just skip all the shit and go to the conclusion page?

    5. Re:Only one benchmark? by Jeng · · Score: 1

      I want to spend mod points on you, but there is no category for WRONG.

      --
      Don't know something? Look it up. Still don't know? Then ask.
  14. Missing the best value for the buck, AMD Kuma 7750 by sricetx · · Score: 2, Interesting

    The article is missing the best CPU value for the money, in my opinion. The AMD Kuma 7750 AM2+ processor. It's dual core, but at around $60 shipped (Newegg) nothing else touches it from a performance to dollar perspective. They should have included the 7750 in the comparison rather than the Athlon X2 6400+ (the 7750 is K10 architecture vs. K8 for the 6400+, has 2MB level 3 cache, is not discontinued, etc.)

  15. Just about threading by astra05 · · Score: 5, Interesting

    I just redid my system line by putting an e8400 in my desktop, where I mostly game. I switched out a Phenom 9600 (cause the tlb erratum for vista x64) to my file server/media center (which runs Ubuntu). and really as this graph will tell you: fast dual-cores are going to blow away slow quads in gaming because most games are not programmed for multiple threads and take advantage of a higher clock core. However, for most other tasks I do, like compiling the Linux Kernel (I run gentoo side by side with vista), the Quad Core Phenom 9600 seems to be much faster. Plus, I had a hard time overclocking the 9600 to anything past 2.6 ghz whereas the 3.0 ghz stock e8400 easily clocks up to ~4.0ghz on air. I should also note that I picked the e8400 over the q8200 because of the virtualization tech as I do alot of virtual systems for testing.

    --
    Live Free
    1. Re:Just about threading by Ant+P. · · Score: 1

      I thought this was common sense.

      You wouldn't use a sports car to tow an 18-wheeler trailer, nor would you go racing in a truck. Most of the time, anyway.

    2. Re:Just about threading by LackThereof · · Score: 1

      fast dual-cores are going to blow away slow quads in gaming because most games are not programmed for multiple threads and take advantage of a higher clock core

      This is becoming less true, things are beginning to change now. New PC games are coming out which are heavily multi-threaded.

      Left 4 Dead, for example, performs better on a slow quad-core than on a fast dual-core. To be fair, this is an extremely recent game, but we can only expect this trend to continue.

      --
      Legalize recreational marijuana. Seriously.
    3. Re:Just about threading by TheThiefMaster · · Score: 1

      A lot of games written for the xbox 360 will run best with three or more cores when ported to pc, purely because the 360 has three cores. Assuming said game makes use of all three 360 cores.

  16. Single core performance? by Anonymous Coward · · Score: 0

    I want to know what's the best processor to buy if one's going to use mainly only one core?

    1. Re:Single core performance? by etymxris · · Score: 1

      Core 2 Duo 3.33Ghz. Supposedly these overclock better than most processors so you can get the single core performance even higher.

  17. Now something about that linked site... by Anonymous Coward · · Score: 0

    What is it with sites like that and insisting I click "next page" 11 times? No, that drop-down isn't enough. I want everything on one page, TYVM. Usually "print this" is a good get-out-of-dodge for this sort of thing, but not here. It gets me four pages, two without content, and together they contain only the first "page" and a printed version of the drop-down bar to find the next "page". I'm not your clicky-slave, techreport. I'll go read somewhere else.

    1. Re:Now something about that linked site... by Anonymous Coward · · Score: 1, Interesting

      I would suggest the repagination extension for Firefox. With it, you just right click any "next" and it automatically loads every page in the series of whatever you are looking at onto the current page. So, you can just keep going down the page to see every next page. Works great for sights like failblog.org, etc.

  18. AMD vs Intel in "floating point" operations by pxlmusic · · Score: 1

    ..like audio software and soft synths.

    Ableton Live will support multiple chips/cores, so I want something that will really make short work of those VSTs that are currently beating my CPU like a redheaded stepchild.

    Any ideas?

    --
    "If for any reason you're not satisfied with our service, I hate you."
    1. Re:AMD vs Intel in "floating point" operations by astra05 · · Score: 1

      Get a core i7. But if you really need a lot of smp'ing, you should check out sun's niagra t1 processor: http://www.sun.com/processors/niagara/ or the Niagra 2.

      --
      Live Free
    2. Re:AMD vs Intel in "floating point" operations by geekoid · · Score: 1

      i7

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    3. Re:AMD vs Intel in "floating point" operations by fnj · · Score: 1

      AMD mops the floor with Intel for floating point.

    4. Re:AMD vs Intel in "floating point" operations by Anonymous Coward · · Score: 0

      IBM's Power6 destroys everything out there for FP (and has for several years now) but you need like 5 grand+ for an entry level system. Here are the specfp2006 rate results:

      http://www.onscale.de/specbrowser/2006-fr-004.html

      Your best bet is probably an i7 though.

    5. Re:AMD vs Intel in "floating point" operations by TrancePhreak · · Score: 1

      [citation needed]

      --

      -]Phreak Out[-
  19. Unlcocked i7 is the pwnge by Anonymous Coward · · Score: 0

    The fact I can now unlock my memory and QPI now completely reevaluates the worth of the 920. The performance increase over my old quad is scary. I was gonna upgrade my 9800 round xmas and see absolutely no need to. I run everything cranked on a 28 inch 1920 x 1200 monitor and it never even hiccups. I think factoring in the fact that I no longer need a GPU upgrade makes the 920 more appealing as I just saved 5 bills (CAN) from having to get a 285. Plus the option on current mobos to go Crossfire or SLI and my P6T SAS compatibility make the platform fairly immune to obsolescence for a while.

  20. AMD didn't drop their prices for a solid 18 months by Anonymous Coward · · Score: 0

    When they ruled the roost late in the Netburst era.

    I want them around, and like them to some degree, but they rested on their laurels to a big big degree.

    Granted, this was after Intel rested on their laurels and gave us the engineered-by-marketing NetBurst.

  21. Re:Missing the best value for the buck, AMD Kuma 7 by mdm-adph · · Score: 1

    It's probably because at stock clocks (what they have to run for this test), the 7750 runs a lot slower than the 6400+, while probably using the same amount of power.

    Other than that it's a darn good chip.

    --
    It is by my will alone my thoughts acquire motion; it is by the juice of the coffee bean that the thoughts acquire speed
  22. not too bad by mu51c10rd · · Score: 1

    I saw an i7 motherboard/cpu combo for $534. The prices does seem to be coming down.

    1. Re:not too bad by i.of.the.storm · · Score: 1

      Sure, but I could build an entire PC for that much, and it would be a pretty good PC too. The prices on the whole i7 ecosystem are kind of ridiculous. My theory is that Intel invented triple channel RAM as a bailout to the DRAM companies.

      --
      All your base are belong to Wii.
    2. Re:not too bad by torkus · · Score: 1

      Slowly perhaps but the C2D systems (and C2Q for that matter) are so powerful for day-to-day use that there's not much price pressure for the latest and greatest.

      It's kind of strange really - in my following of computers since the late 386/early 486 days I've never seen a next-gen CPU so slow to pick up interest.

      --
      You can get rich if you own a politician, but you have to be rich to buy one in the first place.
  23. '\' Re:Suprise? by Anonymous Coward · · Score: 0

    I nominate the semi-ubiquitous '\' backslash. Seen here escaped with apostrophes to provide emphasis. It could also be escaped by itself to represent itself: \\. This however leads us to my favorite regular expression in java: "\\\\\\\\" which is a pattern to match two backslashes in a String that is being searched. (the backslashes in the search String need to be escaped twice, once for being backslashes in a String and once for being backslashes in a Regular Expression.

    Can anyone familiar with groovy verify if it gets even worse at that level? What about dumping a regular expression in javascript generated by a servlet?

    1. Re:'\' Re:Suprise? by itschy · · Score: 1

      Can do:
      "\\\\\\\\\\\\\\\\"
      That would be a pattern to match four backslashes.

      But noone could ever think of a case were we need more backslashes!!!!1!eleven

  24. Re:Missing the best value for the buck, AMD Kuma 7 by citizenr · · Score: 1

    The article is missing the best CPU value for the money, in my opinion. The AMD Kuma 7750 AM2+ processor. It's dual core, but at around $60 shipped (Newegg) nothing else touches it from a performance to dollar perspective.

    E5200, not to mention you can OC almost every E5200 to 4GHz

    --
    Who logs in to gdm? Not I, said the duck.
  25. Re:Missing the best value for the buck, AMD Kuma 7 by PitaBred · · Score: 1

    I just got one of those for my media center, and it flies for 1080p decoding (mplayer-mt is awesome). The 7750 is the Black edition, so the multiplier is unlocked. It runs at 2.7GHz stock, it'll overclock to 3GHz without really breaking a sweat in my experience, not to mention having all the extra cache that sricetx noted.

  26. Re:Missing the best value for the buck, AMD Kuma 7 by robthebloke · · Score: 1

    It also has vastly increased memory bandwidth (3600 vs 2000), a much improved CPU cache, support for faster RAM, uses less power (95 vs 125W), has a 65nm die instead of a 90nm one, works in AM2+ motherboards, and contains the latest set of SIMD extensions. The 6400+ maybe slightly faster (as this shows), however it is twice the price of the 7750 for not a lot of benefit. (and the 7750 overclocks better than the 6400)

    With that in mind, I'd imagine the 7750 would have been solidly beaten the 6400+ CPU in all of the performance/$ graphs in the original article...

  27. Aha! by XanC · · Score: 2, Funny

    Slashdot's seemingly-ridiculous problems with non-ASCII characters are simply a safeguard against displaying the nullity character, which would cause the universe to implode.

  28. SCREA by pohl · · Score: 1
    --

    The "cue the foo posts in 3, 2, 1..." posts will commence with no subsequent foo posts in 3, 2, 1...

  29. ECC - i7 no, amd - yes, end of analysis by lopgok · · Score: 1

    Until the i7 has ecc, there in no winning against the phenom II which has ecc. If you are doing anything excepting gaming, you likely need ecc. Since the i7 doesn't have it (until the i7 based xeons come out), there is simply no comparison. Like bring a knife to a gunfight, you are fscked.

  30. Bad calculation by szundi · · Score: 1

    In general i like AMD better, but these comparisons are usually wrong! If you upgrade your engine in your car and get +20% horsepower, are your car really 20% better?

    Please notice that your computer is not just a processor, so when measuring performance of the processor/dollar then please use the performance/whole-machine formula instead.

    However if you use your machine for io-intensive tasks, the question is bad for you. Anyways this way of calculating is simply not correct. If you are a quantum-mech scientist maybe... even my eclipse building my java code eats the hdd not cpu.

    1. Re:Bad calculation by sznupi · · Score: 1

      If you take whole system into account, AMD is in even better position - generally cheaper motherboards, you can use cheap ddr2 memory just as well, and AMD has way better integrated GFX (might not be enough for you - but larger percentage of AMD users don't have to buy separate gfx card)

      And from savings you can buy very fast io-subsystem.

      --
      One that hath name thou can not otter
  31. Anonymous Coward by Anonymous Coward · · Score: 0

    If you look at the graphs, the 6400 the best bang for the buck...
    Look, $100 buys 100% performance. No other processor comes close to meeting that price / performance ratio. Take the top processor, it's 10x as expensive, but less than 2.5x the performance. Same goes for everything else. No other processor is twice as fast for twice the cash. Or three times as fast for three times the money... All the rest are just overpriced on a price / performance basis. Which is what the graphs should have made clear. They should have plotted price / performance ratio instead, because that is what the article was supposed to be about. Also, there are many CPU's that cost less than that 6400 and can outperform it. My Intel 2180 comes to mind ($50). And it overclocks 50%.

  32. I/O bound by tepples · · Score: 1

    Very true if your system only runs that single application. However, everyone I know runs multiple applications just by booting their OS.

    And how many of those don't become I/O bound very quickly, such as the usual gaggle of svchost processes waiting for a request to service?

    1. Re:I/O bound by ps2os2 · · Score: 1

      I/O is almost always known for its bottle necks. That is one of the things that an OS can actually help out some programs. There was a study done (eons ago) that showed 5 i/o buffers are the most efficient. Now that depends on several factors of course and AFAIK the PC world does not block their records. So I/O is a major thruput stopper. Probably trying to get the PC world to change would be like trying to sell more processors for the same or cheaper. IOW it will not happen. CPU is way to cheap and OS's are dumb to do simple things like buffering for thruput.

  33. Memory bound by tepples · · Score: 1

    Sounds like you need to get a better application for transcoding your videos. Say one that is actually multithreaded.

    In theory, transcoding video is an embarrassingly parallel problem. Split the video into n parts and transcode one part on each core. But that still doesn't help if all the threads are blocking on cache misses.

  34. I/O bound by tepples · · Score: 1

    you can run multiple separate non-multithreaded apps on a single machine and get better performance from each of them.

    If they are interactive apps, then all but one of them will be blocked on I/O the vast majority of the time, waiting for a mouse click or a keypress. So what do multiple cores buy you?

  35. Free? by tepples · · Score: 1

    Let's get away from this absurd undead i386 architecture garbage. Linux runs on any architecture.

    Non-free applications for Linux run only on the architectures that the publisher wants you to use. Free applications for Linux do not serve all niches, such as games.

  36. 64-bit? by tepples · · Score: 1

    Now run your comparisons based on 64 bit code instead of 32 bit.

    What's so special about 64-bit code? I can buy a 64-bit gaming computer for under $25 shipped. And I'd bet that any of the Intel and AMD CPUs reviewed in this article can emulate it.

  37. Re:Missing the best value for the buck, AMD Kuma 7 by owlstead · · Score: 1

    Yep, runs like a charm, really really fast processor and cost me next to nothing (~70 euro). Mine was sold as a "X2 dual core black edition" but it definitely has got an AMD 10 inside. Nice wattage (65W TDP) too, now idling along at half speed. The rest of my money is being spend on memory (8GB 1066MHz, 150 euro) and SSD (60 GB Vertex around the same). The motherboard was 90 euro, all solid caps and everything (but firewire) build in. So cheap I bought a high end power supply as well.

    Brilliant speed and silence on a budget, but look out that you don't inadvertently buy an AMD 8 instead.

  38. Which would be better for gaming? by Jourdyn · · Score: 1

    One of the x3 processors or the I7? everyone keeps telling me to get the i7.