Comments Locked

68 Comments

Back to Article

  • mbf - Monday, July 23, 2007 - link

    I accidentally posted this in the comment section for the earlier article, but it seems to fit better here. Sadly I cannot delete the other comment...

    I've been wondering how older motherboards will work with the new FSB1333 processors. Specifically I'm interested how an ASUS P5W DH Deluxe without the latest BIOS would react to having e.g. an E6750 dropped in. ASUS claims support for FSB1333 processors for the P5W DH Deluxe as of 2205 beta.

    Would the system boot and run using a pre-2205 BIOS (although not at peak performance), so a BIOS upgrade can be performed? Or would the system fail to boot at all, like when the first Core 2 Duo processors surfaced and needed a BIOS upgrade to run at all on certain boards.

    The reason I ask this is that I've my eyes set specifically on that board (I have several reasons, ECC memory support being one of them). I had originally planned on getting an E6600 after the July 22 price cuts, but right now there's nearly no FSB1066 processor to be had locally. Also, I'd of course love to have a access to the latest processors in any case.
  • number - Wednesday, July 18, 2007 - link

    Marvelous article. However, one benchmark is missing. Quad core processor may be used in the following way: two cores are working on a job that utilizes them to the max, while remaining two run a game. How well the processors fare under this scenario?
  • strikeback03 - Wednesday, July 18, 2007 - link

    Was the Q6600 compared to the E6850 at stock speed, and not with the boost from setting 1333 FSB? Is it really 10% faster in CS3 than the 3GHz E6850 (and therefore a lot quicker than my E6600)? And would similar improvements carry over to CS2, or did they improve multicore support in the transition from CS2 to CS3?
  • IntelUser2000 - Tuesday, July 17, 2007 - link

    quote:

    Despite theoretical showings on paper, the 1333MHz FSB appears to do very little for performance even when feeding four of Intel's fastest cores.


    Maybe you should also add scores of 1066MHz FSB on the P965 rather than showing 1066FSB on P35 only. Your earlier tests with P35 have shown that there is performance improvement just by moving from P965 to the P35 chipset. Moving to P35 showed greater improvements than changing P35's supporting CPU from 1066FSB to 1333FSB.
  • Bozo Galora - Monday, July 16, 2007 - link

    http://www.tgdaily.com/content/view/31822/135/">http://www.tgdaily.com/content/view/31822/135/
  • Bozo Galora - Monday, July 16, 2007 - link

    (no edit button?)
    http://www.thetechrepository.com/showthread.php?p=...">http://www.thetechrepository.com/showthread.php?p=...
  • scott967 - Monday, July 16, 2007 - link

    Seen various comments to "wait for G0" stepping coming out now or very soon. What stepping was used on this comparo and any comments on this stepping issue?

    scott s.
    .
  • clairvoyant129 - Monday, July 16, 2007 - link

    I love how all these AMD proponents claim that Intel motherboards are too expensive. Then they claim the difference can buy them a much better video card.

    There are tons of cheap Intel motherboards that are just as good as comparable AMD motherboards.

    These fanboys are pathetic. Whats the next excuse?
  • mamisano - Monday, July 16, 2007 - link

    Curious, anyone find these new Intel CPUs in stock, and if so are the prices in line with what has been listed?
  • Bozo Galora - Monday, July 16, 2007 - link

    I have to agree this was a superb article. Well thought out and logical in all repects. Answers all buying questions from various angles.
    Of course, IBM's 300GHz CPU perhaps may make all this irrelevant - heh
    http://www.newtechspy.com/articles06/crystalcomput...">http://www.newtechspy.com/articles06/crystalcomput...

    While I am at it, the article on 32 bit addressing was also very clear and informative explaining the 4GB "wall", and the 2GB/2GB split. I am sure many on the web will make reference this article in the future.

    And the first power supply review was the third in this triad of superior investigation by the AT crew. Those graphs showing the PSU voltage vs wattage load are simply the best insight I've ever seen in a PSU review anywhere. Also interesting was the fact the Silverstone was advertised as a single 12V rail, but was in fact 4 separate.

    AT seems to be reinvigorated for some reason - kudu's to you, top notch work.
  • DolphinAMD - Monday, July 16, 2007 - link

    nooo! I just bought an Core 2 Duo E6420 for $186
    The replacement seems like the E6750 at $183

  • Sunrise089 - Monday, July 16, 2007 - link

    sorry buddy, but these prices weren't exactly a secret.
  • lennylim - Monday, July 16, 2007 - link

    If you go according to max. multiplier, an important number for overclocking, the pricing makes some kind of sense.

    Looking at dual cores only, here's the pricing by multiplier.

    7x multiplier : $163
    8x multiplier : $183
    9x multiplier : $224 (E6600) / $266 (E6850)
    10x multiplier : $316

    Still a damn good price for C2D. And the E6600 is actually a good deal for overclockers.
  • hubajube - Monday, July 16, 2007 - link

    The price of the X2 6000 is $169.99 on Newegg not $180. Even though the new stuff is quicker than AMD's best it's only a little quicker and the $10 difference between them plus the cheaper motherboards for AMD will still pretty much seal the deal for my next upgrade. If the quad cores were $180 then I would be willing to stretch for the extra cost of the Intel motherboards (Intel or Nvidia chipset).
  • MrKaz - Monday, July 16, 2007 - link

    Finally someone with brains.
    I already have said the same with different words.
    Did you also notice they only mention the extra price cost on the AMD motherboard because of Quad FX, they don’t mention Intel extra price premium on the motherboard.

    The Intel motherboards are very expensive, the ASUS P5N-E SLI for example in my country costs 120€, the AMD version the M2N-E SLI 80€. Its 40€ difference.
    Also the only interesting Core 2 Duo is the E6550 which costs $163. Lower than this you get one castrated CPU from Intel.
    AMD X2 3600 costs 60€ in my country so its 60+80=140€
    Intel E6550 168€ so its 120€+168€=288€
    Its “just” 148€ difference or one X1950PRO if you prefer.
  • Accord99 - Wednesday, July 18, 2007 - link

    quote:

    The Intel motherboards are very expensive, the ASUS P5N-E SLI for example in my country costs 120€, the AMD version the M2N-E SLI 80€. Its 40€ difference.

    Why compare the E6550, which is faster than all but the AMD 6000+ with a 3600+ that is slower than Intel's "castrated" Pentium E2160?
  • hubajube - Monday, July 16, 2007 - link

    I was looking at the ASUS M2N32-SLI Deluxe which is $170 compared to the ASUS P5N32-SLI Premium for $210 here. A $40 difference that I can use towards my 8800GTS and then add the $10 from the CPU difference and I STILL get a computer that plays BF2 or whatever at WELL over 100fps.
  • Pirks - Monday, July 16, 2007 - link

    exactly, this is why I'm going to use AMD for my gaming rig for the foreseeable future. cheap AMD cpu plus expensive nvidia 3D video equals best gaming experience, definitely better than that of intel FOR THE SAME PRICE. AMD is the best friend for the gamer, while Intel is the best friend for media encoder/3D renderer kind of guy, or anyone who loads all their four cores at 100%
  • iceburger - Monday, July 16, 2007 - link

    I've had the same reasoning for the last 10 years- AMD platform always cost less- sometimes the difference will cover nice video card. IMHO this Intel price cut may backfire for them- the reason you spend money on marketing is to be able to get higher profit margins. Instead of dropping the price, the reasonable decision would have been to "milk" the market- maintain the current separation: high- and mid-end for Intel, low-end for AMD. This price cut is strictly a hostile move aimed to bury AMD's entire line and force them to lose even more with consecutive AMD price cut. However Intel has higher fixed development cost and even if they sell more CPUs, their profit will be lower- it's a lose-lose situation. Reminds me of GWB's tax cuts.
  • Sunrise089 - Monday, July 16, 2007 - link

    Only problem with all of this otherwise sound reasoning is that the same folks who are enough of an enthusiast to know that the AMD MB's can save them a bit, and then apply that savings towards either the GPU or grabbing a higher-end AMD processor are very likely to overclock, and that sort of blows the AMD side of the equation out of the water.

    I'm and AMD fan myself, but you have to admit, anyone who even considers overclocking their CPU has no business picking the AMD side unless they just want to help the underdog.

    Pointing out Anandtech's failure to mention the cheaper AMD platform is fair enough, but it was AT LEAST equally damaging to the Intel side to not show benchmarks from an overclocked 6000+ vs a 6850. Does anyone think it would be close?
  • MrKaz - Tuesday, July 17, 2007 - link

    I don’t know but I have very doubts that someone who buys some premium and highest end CPUs will even bother in OC.
    If I was going to buy one AMD CPU for OC I would choose one of the single core or one dual up to the X4400+, higher than that I was shooting myself in the foot.
    With Intel I would go for one of the lowest FSB versions (800/1066) or the slowest of the 1333Mhz (but I doubt I would go for one of this). Going for the 3.0Ghz Intel versions I was again shooting myself in the foot. Why OC something that is already so fast and already in its limits.
  • Pirks - Monday, July 16, 2007 - link

    quote:

    the same folks who are enough of an enthusiast to know that the AMD MB's can save them a bit, and then apply that savings towards either the GPU or grabbing a higher-end AMD processor are very likely to overclock
    VERY far from truth - I'm kind of enthusiast myself, but I specialize in silent & inexpensive gaming computers, and AMD gear is VERY solid choice here, I pick old single core AMDs for nothing on ebay, like $45 for a fast gaming San Diego 4000+, pick older 7900GTX cards on ebay as well for cheap (only the ones I know are silent 'cause reviewers say so) and then I stuff it all in Antec P182, do some other voodoo with Cooler Master or ThermalRight gear... voila, a SILENT gaming rig, chews through S.T.A.L.K.E.R. just like that! and cheap, compared to some intel rigs from overclocking Intel freaks - it is DIRT cheap, cause there are no water, no overheating from fashionable overclocked quad-core intel shit, nothing like that.

    so you are TOTALLY wrong when you classify all enthusiasts as overclocking intel freaks - there are a lot of us who specialize in silent & inexpensive PCs, including gaming ones. it's easy to pay intel $$$ $$$ $$$ and get watercooled quad extreme blah blah blah, or get a cheap and noisy overclocked aircooled rig. but to get 1) gaming 2) silent 3) cheap PC - this is REAL ART, btw it's not covered at anandtech at all
  • relic2279 - Monday, December 17, 2007 - link

    A User said:
    "the same folks who are enough of an enthusiast to know that the AMD MB's can save them a bit, and then apply that savings towards either the GPU or grabbing a higher-end AMD processor are very likely to overclock."

    Pirks replied to this:

    "VERY far from truth - I'm kind of enthusiast myself, but I specialize in silent & inexpensive gaming computers."
    -----------------------------------------------------------


    Very far from the truth? Possibly for you specifically but he was generalizing and I believe he is correct. People who do care enough and are being specific, building their own PC's tend to be the same people who tinker, and OC their computers. People intrigued enough to read this whole article and pay attention to the benchmarks are more likely to overclock then not.
    So to say that it's "VERY far from the truth" is not only incorrect, but ignorant. It's just more fanboys spouting propaganda for their favorite company.

    I don't have a preference personally. I buy what the best is for my money, at the time. If it's AMD, then I buy them, if it's intel, then them. I suggest everyone do the same. I've purchased 6 intel chips and 7 amd chips in my life. Most of my intel chips were 286's 386's or pentium 1-2's. Lately I was buying AMD cause they were the better buy, but not now. For the money, I get alot more with intel. I have noticed an increase in reliability as well, after switching to intel.
    I may have to take that into consideration on my next chip purchase which (if the wife allows me) will hopefully be soon. :)

    Oh and I noticed that some people mentioned that the price cuts would be bad for intel as far as profits go etc... The price cuts benifit us, and thats what matters. I don't care if amd/intel's revenue is down this year by 100% and neither should you. What matters is that we get a good price and a good cpu. If someone brings that up, it just further proves they are fanboys and care more for that particular company then they probably should. Again, getting a decent price is what should matter, not profit margins of a huge company.
  • Sunrise089 - Monday, July 16, 2007 - link

    please learn to tell the difference between "most likely" and "all"
  • Pirks - Tuesday, July 17, 2007 - link

    quote:

    please learn to tell the difference between "most likely" and "all"
    doesn't matter if you classify all enthusiasts as OC guys, or just "most of them" as OC guys - this is still your subjective opinion, you have no facts to prove it.

    I classify "most of enthusiasts" as silent PC guys, not OC guys, so what? here ya go, my subjective opinion versus yours. enjoy your meal :P

    yeah, and when you get some solid arguments besides your opinion - don't forget to post them here, I'm interested! maybe I'm wrong about most of us being silent PC people, who knows ;)
  • Accord99 - Wednesday, July 18, 2007 - link

    quote:

    I classify "most of enthusiasts" as silent PC guys, not OC guys, so what? here ya go, my subjective opinion versus yours. enjoy your meal :P

    Well, silence enthusiasts would be better off with Intel seeing as how most of their Core based dual-core lineup now uses less power under load than the 4000+ San Diego, and with the G0 stepping Intel has only increased its performance/watt.
  • Pirks - Wednesday, July 18, 2007 - link

    quote:

    silence enthusiasts would be better off with Intel seeing as how most of their Core based dual-core lineup now uses less power under load than the 4000+ San Diego
    nice shot, but, alas, a miss - you have no idea how much AMD 65nm and 35watt dualcore chips consume under load. get back to school, read you hardware docs, come back - we'll talk again. good luck ;)
  • Accord99 - Wednesday, July 18, 2007 - link

    http://www.matbe.com/images/biblio/art_core-se-dec...">http://www.matbe.com/images/biblio/art_...e-en-pen...

    Complete domination of Intel Core processors in full load power consumption.
  • Pirks - Wednesday, July 18, 2007 - link

    quote:

    Complete domination of Intel
    you're right about consumption, intel is slightly ahead, but if we take prices and upgrade scenarios (like copious amount of old DDR RAM in the system) into account, the picture is not so rosy for Intel
  • Accord99 - Wednesday, July 18, 2007 - link

    quote:

    you're right about consumption, intel is slightly ahead

    In the same way that AMD's K8 was slightly ahead of Prescott.

    quote:

    but if we take prices and upgrade scenarios (like copious amount of old DDR RAM in the system) into account, the picture is not so rosy for Intel

    How so? There aren't any of the low-voltage or 65nm X2s on Socket 939, they're all high-power 90nm models where even the lowly 3800+ uses as much power under load as the fastest dual-core C2D. Meanwhile there are a few DDR1 MB that support C2D if you want, and with the excellent power usage of the C2D, passive cooling is a piece of cake with a half-decent tower heatsink.
  • Pirks - Wednesday, July 18, 2007 - link

    again, you're probably right if the silent PC is build with dual core CPU. in my case I've got single core CPU since I don't need any dual core functionality (I mostly play games like Doom 3 and S.T.A.L.K.E.R.), and while dual cores from intel are very solid choice for silent PC - single cores from Intel pretty much suck. hence my choice of AMD San Diego single core chip - that chip turned out to be the best "price/performance/watt for gaming" ratio I could find on ebay :) again, this is all about single core CPUs. I have no idea how the picture looks for dualcores, probably Intel got stronger offer here - by the time I'm about to upgrade to dual core Phenom would be around and we'll see again who wins - AMD very often wins by better price, even when their CPUs are slightly inferior to Intel ones
  • utube545 - Thursday, July 19, 2007 - link

    Oh, STFU already, you dumb fanboy
  • Pirks - Thursday, July 19, 2007 - link

    what, you forgot to put some lube on your blue intel dildo again?
  • Zak - Monday, July 16, 2007 - link

    Looks like the Core 2 Duo E6850 (3.00GHz) is a decent pick for gaming machine until games take full advantage of quad-core CPU.

    Z.
  • jay401 - Monday, July 16, 2007 - link

    Or if you're a budget-conscious gamer, pick up an E4400 for less than half the price and run it at 3.0GHz/1333fsb (drop the multiplier to 9x) which seems to be a pretty common and easy oc with any motherboard capable of 1333fsb.
    Should show little or no performance difference considering the only hardware difference is it has half the cache which doesn't seem to impact games much if at all.
  • sprockkets - Monday, July 16, 2007 - link

    Yep, you want a nice fast platform, get Intel. But I've tested power usage of my Abit nview + 3800+ dual core 65nm processor from AMD, and it takes around 115 watts of energy at full load.

    I think nowadays either you get an ATX gaming system or now try to build the smallest and quietest and coolest mini itx system since they are powerful enough now for most.
  • bobbyto34 - Monday, July 16, 2007 - link

    Yeah for normal home usage, it is interessant to find the best power consumption/ price / performance Ratio.
    AMD X23800+ @ 65€ was an excellent bargain (good performances ingame, though not as good as C2D).

    In some cases, you want only raw performance: usually for work, less time spent waiting, gives you more productivity. At work we have to treat 1gigabyte text files, so the E6700 rocks vs other stations we have (A64 3000+ or P4 3.2Ghz) !
  • kataras - Monday, July 16, 2007 - link

    Hello, just wanted to know is this the Core 2 Duo CPU which has no IHS (internal heatsink) on. Was it hard to remove it, do you need special tools? Did it decrease the temp significantly? I am asking this because i am thinking of either removing IHS or lapping my E6320 as it runs really hot indeed. i would be very pleased if you could answer my questions regarding IHS.
    Thanks
    Ron
  • AMDfreak - Monday, July 16, 2007 - link

    You can find lots of info about removing the IHS over a the xtremesystems.org forums.
  • microAmp - Monday, July 16, 2007 - link

    "... whole 7MHz faster than its predecessor may..." on page 2. Should be 70Mhz faster.
  • gigahertz20 - Monday, July 16, 2007 - link

    Ahhh man I loved the article but I was hoping for some overclocking benchmarks, very disappointing. I wanted to see an overclocked Q6600 vs. overclocked e6850, there were reports the E6850 can OC up to around 4GHz. I was hoping this article would show us some OC results. Is this planned for later or something?
  • cpeter38 - Monday, July 16, 2007 - link

    Another Ditto!!!!
  • Sunrise089 - Monday, July 16, 2007 - link

    and another!

    All the early Conroe reviews had OC numbers. And yes, by now we realize that as far as dual-cores go, any of the new parts should overclock about the same. It's really important to a lot of users however to know how the overclocking of these new parts matches up to the quad-core part.
  • Frumious1 - Monday, July 16, 2007 - link

    You've got numbers http://www.anandtech.com/mb/showdoc.aspx?i=3020&am...">like this that have quad core hitting 3.60 GHz with appropriate cooling. Q6600 should easily hit 3.0-3.3 GHz with a reasonable HSF, and if you want something high-end like the Ultra-120 Extreme, maybe even 3.60 GHz (9x400). One more reason to go quad - just mind the energy bills!
  • Sunrise089 - Monday, July 16, 2007 - link

    yes, but what about if the dual-core part is overclocked to 4.0ghz?
  • Frumious1 - Monday, July 16, 2007 - link

    Given that overall performance goes to 2.4 GHz quad core over 3.0 GHz dual core (a 600 MHz difference), I can guarantee that a 3.6 GHz quad core would be better than 4.0 GHz dual core (a 400 MHz difference). That said, for gaming it really wouldn't matter much right now - no games even try to utilize more than two cores that I'm aware of. And don't even think about a 3.6 GHz overclocked quad core chip unless you have a beefy PSU!
  • BikeDude - Tuesday, July 17, 2007 - link

    Flight Simulator X SP1 adds multicore support and should benefit from additional (beyond dual) cores.

    "As I stated previously, our multi-core support will take advantage of both 2 and 4 cores today, and more cores in the future when they become available."
    (http://blogs.msdn.com/ptaylor/archive/2007/05/14/f...">http://blogs.msdn.com/ptaylor/archive/2007/05/14/f...

    I am a bit disappointed that Anandtech doesn't bench FSX. I also miss the compiler benchmarks they used to do. AMD used to do quite well in those... (I say this as someone who uses compilers, not as a former AMD fanboy which I probably am)
  • RamarC - Monday, July 16, 2007 - link

    anyone know what oc results i can expect with a p35-based mobo and a q6600? is 3.0ghz as easy to reach as it is with an e6600?
  • ZDNetReader1 - Monday, July 16, 2007 - link

    Ditto!! What he said!!
  • MrKaz - Monday, July 16, 2007 - link

    Price-wise, the only AMD/Intel competition we have here is between the FX-74 and the Q6700. Do keep in mind that as the FX-74 is a dual-socket configuration, the motherboard is a bit more expensive than what you can use with any of the single-socket quad-core Intel solutions.

    This is obviously true, but it’s also true that Intel motherboards for single socket VS AMD motherboards for single socket are also more expensive.
    In my country equivalent ASUS AMD motherboard VS ASUS INTEL motherboard is around 30% to 50% cheaper.
    Do keep that in mind too.
  • Darkmatterx76 - Monday, July 16, 2007 - link

    Nice article. I would like to point out 1 small inconsistancy. On page 12, 4th graph down you have the order for that particular "Lower is better" reversed compared to the others in the article.

    Also, I do have 1 question. Any idea when Intel will offer non-extreme quad cores at 1333 FSB?
  • zsdersw - Monday, July 16, 2007 - link

    I don't get it. Both are listed as 2.33GHz with 1333FSB and both with 4MB. What's the use of having two models?
  • zsdersw - Monday, July 16, 2007 - link

    Nevermind. I found the answer. The 6540 doesn't have Intel Trust Execution technology.. or so I read elsewhere.
  • jay401 - Monday, July 16, 2007 - link

    So how does the E6850 ($266 3.0GHz 1333fsb) compare to my existing E4400 ($133 running 1333MHz fsb with a 9x multiplier = 3.0GHz)?

    That's the test I'd like to see. Half the price but half the cache: Which is better.
  • bobbyto34 - Monday, July 16, 2007 - link

    Your o/c CPU might just be a little hotter :)
    Otherwise, it should have the same performance approximatively (less cache in E4xxx). But other tests showed that the E4300@3Ghz and could approach the performance of the X6800 !
  • lplatypus - Monday, July 16, 2007 - link

    Here's a little error I spotted on page 2, in case you want to fix it: the QX6850 is not 7MHz faster than the QX6800; it is 70Mhz faster.
  • Gary Key - Monday, July 16, 2007 - link

    Fixed.
  • 96redformula - Monday, July 16, 2007 - link

    I also think the scale would be better from -100 to 100. It makes it easier to distinguish and more visually pleasing.
  • ManuelX - Monday, July 16, 2007 - link

    I don't post here much but I had to this time. I simply loved the article. The logic behind the comparison was explained nicely, and the comparisons themselves were super easy to grasp. Good stuff.
  • just4U - Monday, July 16, 2007 - link

    I am going to have to agree here. Nicely laid out article with easy comprehensive graph comparison(s). Well done Guys!
  • xsilver - Wednesday, July 18, 2007 - link

    One question has still yet to be answered:
    how far does the e6850 overclock vs how far the q6600 overclocks

    from previous articles the q6600 doesnt reach much beyond 3ghz unless you have supercooling?
    but the e6850?
  • Slash3 - Monday, July 16, 2007 - link

    I know this is a bit after the fact, but would it be possible on the "vs" charts, to plot the negative performance improvements (read: performance loss) in a left-of-center fashion, instead of having both extending to the right of zero, with a negative sign tacked on? It makes it pretty difficult to scan visually. Go from -100 to 0 to +100 in the same X axis, and just increase the granularity a bit to fit things on, in cases where there are significant negative values. The E6850 vs Q6600 is a good example. Negative and positive, all over the place. Just friendly commentary. Excellent writeup, otherwise. :)
  • DerekWilson - Monday, July 16, 2007 - link

    as was explored in a previous video artilce, we could simply add 100 to each of these and compare the bars with 100 percent meaning eqivalent performance. negatives would be less than 100 while positives would be greater than 100 ...

    personally, i don't mind the negaive numbers in a different color paradigm. if the readers would prefer the "centered at 100%" style, we will certainly adapt.

    i don't know how the other editors here feel, but marketing guys like to show us graphs around 100% performance of something ... because of that, it just ends up feeling wrong to me. :-)
  • dev0lution - Monday, July 16, 2007 - link

    No red lines? That's a pretty impressive lineup for the prices Intel has. Looks like there might be Q6600 in my future very soon :)
  • tuteja1986 - Monday, July 16, 2007 - link

    where is this price cut.. i don't seem em in newegg.
  • webdawg77 - Monday, July 16, 2007 - link

    July 22nd
  • DerekWilson - Monday, July 16, 2007 - link

    fixed the red lines issue
  • Thatguy97 - Thursday, June 18, 2015 - link

    back then i stuck to dual core with my e6600 going all the way up to 4ghz ish speeds

Log in

Don't have an account? Sign up now