Back to Article

  • 36772 - Saturday, May 30, 2009 - link

    Hi, I own a pair of Athlon MP 1600+ processors. I got them as a gift a year ago. They're "slow" but very responsive. There's only one problem I have so far. The PCI on my MB is only 2.1 and I need 2.3 for my TV card. As one of the first consumer SMP platforms - can the MPs be benchmarked once again? Thank you :) Reply
  • Seramics - Friday, May 29, 2009 - link


    according to AMD official website here,">,,30...

    LE-1640 2.7 GHz 512KB 45W
    LE-1620 2.4 GHz 1MB 45W

    so the bench info about LE-1620 having 512kb L2 cache is wrong.
  • Hrel - Friday, May 29, 2009 - link

    I'd really like to see them compared to P4 Hyperthreaded 3.0GHz and the Pentium D 930, as the Pentium D 930 was the first processor from Intel that I feel meets my now heightened performance requirements; and I won't really be interested in anything less than a 3.0GHz Hyperthreaded P4; I don't care how little power it uses. HTPC CPU's are different than mobile CPU's, and I just don't really care about power consumption when I have a wall outlet. Reply
  • DerwenArtos12 - Wednesday, May 27, 2009 - link

    I know the lower is better graphs are a pain and confusion for everyone, and I've had an idea that I hope can help. Instead of doing them by pure time increment, do them as either a percent improvement chart or as a less time chart, essentially, set the graphs so that the largest number is zero, every other processor will then be scored on a negative scale, invert the scale and show seconds of improvement over the baseline.

    For example, the Windows Media Encoder 9 x64 test, the core i7 965 had a low score of 24 seconds, The Atom 230 had a high score of 271 seconds, under the scale that I'm suggesting the Atom 230 would have a score of 0 and the Core i7 965 would have a score of 247, and the chart would be labeled as "seconds of improvement over a baseline, higher is better." The concept would take some explaining in articles at first, but it would give people a defined scoring metric that would remain linear(simply oposed) to the lower is better score.
  • croc - Tuesday, May 26, 2009 - link

    While I am sure that you will get many requests for re-running older CPU's, I am sure that you won't get very many volunteers to help in this endeavour. I doubt if one in one hundred readers understand the time required in setting up a system then running the suite of tests against that CPU, and recording the results. What has been done so far must represent 100's of man-hours to have accomplished, and I for one applaud that effort.

    I am looking forward to the HD benches, as their benchmark suite shouldn't change near as much over time. I hope you include some SCSi / SAS drives for completeness.

    GPU tests? Seems as difficult of a task as trying to measure the climatic changes of one butterfly flapping in the Amazon... But I commend you for at least thinking of making the attempt.
  • NICOXIS - Tuesday, May 26, 2009 - link

    I'd like to see, single core:

    1.6 P4-M
    1.6 Athlon XP-M
    1.6 Atom
    1.6 Turion 64
    1.6 Nano
    1.6 Pentium M
    1.6 Athlon Neo
    1.6 C7
    1.6 Core Solo
    1.6 Phenom (one core enabled)

    I'd love to see performance/watt figures on that.
    Keep up the good work,
  • Johnmcl7 - Tuesday, May 26, 2009 - link

    Once choice I've seen debated is a new Atom notebook against an older refurbished ULV system, it would be interesting to see how the likes of a Pentium-m 1.1Ghz or Core Solo 1.33Ghz processor compared to the Atoms. Reply
  • 9nails - Tuesday, May 26, 2009 - link

    How close are we to being able to predict CPU and GPU scores combined? At some point all of this data is going to establish a trend. With some math that trend could show predictions for hardware combination's which are synthesized from these results. For those of us who are system builders, it would be nice to be able to pick our CPU's and add in some expected GPU and see where the graphs lie. Reply
  • MrPoletski - Tuesday, May 26, 2009 - link

    how about including some of the other processors, such as arms cortex, vs atom in cross platform benches. Cortex is a competitor of atom after all. Reply
  • jjuhl - Monday, May 25, 2009 - link

    > Any suggestions from the crowd?

    Yeah, I've got a suggestion.
    I own a 'fit-PC slim' ("> with a 500MHz AMD Geode CPU. I'd love to see that included in the benchmarks to see how it stacks up to the competition.
  • mino - Monday, May 25, 2009 - link

    You may still print the results but the score shall be allways normalized for easy visual comparizon.

    I.e. 0% to 100% where the faster CPU gets 100%

    This will remove a lot of initial confusion as with real numbers as bar sizes every second score has a different meaning (lower/higher is better).

    Another, imediate, option is to group the lower-is-better and higher-is-better.
    That way the visual info would be much more usable without going to the small print...
  • Skiprudder - Monday, May 25, 2009 - link

    I know World of Warcraft isn't a particularly demanding game, but it is the most popular game in the world currently. It has also had some significant updates with two expansions and several major patches. This has included implementation of multi-core cpu support and significant graphical enhancements.

    It's hard to find benchmarks that reflect either these changes in performance demands or benchmarks that rate the game on recent hardware. Due to its MMO nature WoW also places different demands on CPU and memory usage when compared to many other games. It would be nice to have a reference where one could see exactly what one could get say upgrading from an older CPU to a Phenom II, for example.
  • 91TTZ - Monday, May 25, 2009 - link

    I've always had to manually look at the comparisons from years past to figure out how my new component would compare to my old ones. If you wanted to compare your old CPU "A" to the new one of "C", you couldn't find it all in one review since reviews usually don't include products from older generations. You'd have to read through the review archives to see how A compares to B, then how B compares to C. Finally you could figure out how A compared to C.

    This idea is great.
  • cfaalm - Monday, May 25, 2009 - link

    This is just what the doctor ordered. Excellent work. It will be so much easier to determine the overall performance of a cpu. The additions would be some older stuff indeed: Athlon XPs and P4 (nonHT/HT, Northwood/Prescott) and the first dual cores of AMD (X2) and Intel (Pentium D). Any other x86 chip like VIAs Nano would be nice too. Reply
  • Crassus - Friday, May 22, 2009 - link

    I think it would be good to get some mainstream CPUs from the past days in there - Athlon 64 3000+ and 4200+ come to mind, some Northwoods, certainly the 1GHz Tualatin and maybe the 600MHz Katmai :) Reply
  • abnderby - Friday, May 22, 2009 - link


    Hey quick question, Why is it that CPU comparisons never have xeons or opterons? It is nice that you have all of these desktop and laptop processors, but all of the sites seem to forget that many of us out here have bought and use many dual cpu systems. I use 2 dual xeons for desktops and many of my friends and co-workers do also. Just look at ebay dual xeons sell like hot cakes. It would be nice to know how our systems stack up to the latest and greatest. I have seen several of the motherboards used in your drive tests are ones that I have used and or are using.

    Like now I run flawlessly 64-Bit Vista/Windows 7 on the following:

    Intel SE7525RP2
    Dual Xeon 3.00 GHz 64-Bit 800 FSB 2MB L2
    8GB DDR2 ECC
    AMD HD 3870 512MB PCI-E
    2 Dell FP1800 18" flat panels 1280x1024
    X-FI Sound Blaster
    Adaptec SA-2610 RAID 5 with 6 320GB WD

    I play Crisis at max settings without stutter. But if I had to go by what is listed it would seem like i run a junk system. It is old but it runs great. So without really being able to see how I stack against newer systems I find it hard to justify buying into a whole new rig. I have nothing to compare to. As this is the problem for many of us out here.

    Just like adding the Intel V8 would be nice etc...

    Does anyone else agree?

  • Lazlo Panaflex - Friday, May 22, 2009 - link

    If possible, could you guys please add a few of the older 1st gen Quads (Extreme/non extreme) to the mix?

    Was also curious where the QX6700 would end up (~Q9400?)?

    Thanks much!
  • aj654987 - Friday, May 22, 2009 - link

    I would love to see some tests done on a few p4's or especially athlon xp era systems.

    I am wanting to replace my current athlon xp laptop with one of the dell 12" atom netbooks because funds are low and I cant get anything else near the price that will run 5+ hours on battery life.

    I really want to see the performance difference. Even the sempron is killing the atom in all the tests so it would be nice to see exactly what an atom is comparable to in performance.
  • BushLin - Friday, May 22, 2009 - link

    I've only just upgraded from an Athlon XP-M @ 2.7Ghz and apart from a couple of applications I've been underwhelmed by Core2 @ 4Ghz, would be interesting to see what the benchmarks make of the XP at stock.

    BTW, if you do this, please run the benches on an Nforce2, I can't remember a better chipset for the Athlon.
  • FlameDeer - Friday, May 22, 2009 - link

    Thanks for providing this useful & convenient interactive CPUs Bench with regular update.

    Here are few suggestions:
    1. Increase View Comparison function to compare up to 5 CPUs each time. Comparison for 2 to 5 CPUs, let users choose from drop down list the number of how many CPUs want to compare at a time.

    2. Adding bench data of Intel Pentium Dual Core E5300 from the review here.">Zotac Ion mini-ITX Ion Board - Page 7">Athlon X2 7850 vs Pentium E5300 - Page 3

    3. Adding bench data of Intel Atom 330 (Zotac Ion) from the review here.">Zotac Ion mini-ITX Ion Board - Page 7

    4. Adding non standard reference platform bench data section, using grey chart bar. Suitable for old CPUs tested in different platforms & drivers, also for partially tested CPUs with only few benchmark results. Just like the Intel Pentium 4 2.66 GHz data use here and other old CPUs that plan to test later.">Zotac Ion mini-ITX Ion Board - Page 5

    5. Let users choose with a drop down list of bench data sections like below.
    - "Standard Reference Platform" (data using now, show with blue chart bar)
    - "Non Standard Platform" (for old CPUs or partially tested CPUs, show with grey chart bar)
    - "Show All Results" (display all CPUs results regardless platforms or sections)

    6. Concerntrate on desktop CPUs for this interactive CPUs Bench. Only try to create mobile CPUs or server CPUs bench data as different tools in future, because with different functionality & platforms involved, needed different sets of test software.

    7. Highly anticipated the storage & GPUs bench data interactive tools. Here also can use the approach mention at point 5 above, create different sections - standard, non standard & show all. This can help to increase the number of items covered while easier to update & maintain.

    8. Changing Intel Core 2 Duo E7200 frequency to 2.53 GHz for review chart drawing bench data in future use.">Athlon X2 7850 vs Pentium E5300 - Page 3">Phenom II X4 955 Black Edition - Page 8

    These are the few suggestions that I can think of & find out. Thanks Anand, take care.
  • Zak - Friday, May 22, 2009 - link

    Very useful, thanks!

  • lemonadesoda - Friday, May 22, 2009 - link

    I dont think it is necessary to have so many CPUs of the same architecture in the database, but just with different Ghz speeds. IMO, you only need 2 CPUs of the same architecture; a low speed one and a high speed one. That information is sufficient for us to see if the CPU scales well with speed.

    However, at the same time, I think there should be more architectures in the database; e.g. Tualatin, Northwood, Gallatin, and DP Xeons, like Sossaman, Woodcrest, Clovertown, Harpertown, Gainstown.

    The table would then be very thorough/comprehensive.
  • Denithor - Friday, May 22, 2009 - link

    This is exactly the direction my thoughts were going as well.

    Instead of cluttering things up with 50 different CPUs just pick a standard speed and compare in a variety of single and multi-threaded apps.

    Chips with HT should be tested with & without to show the impact in both single & multi-threaded apps.
  • LtPage1 - Friday, May 22, 2009 - link

    I'd be curious about an Athlon XP (specifically, the 2400+), because that's the chip I got into enthusiast computing with. Reply
  • JonnyDough - Thursday, May 21, 2009 - link

    I would really like to see one of each of the following listed: High clock and low clocked single and dual-core athlons (that's 4 Athlons total - use DDR@400mhz of course), and a Pentium 3 @ 1Ghz for good reference if you can. A couple Pentium 4's are fine as well, although we already know they place a bit behind the Athlons. If these chips are too old or too insignificant, I might remind you that dual and single core Athlons are still quite capable of playing a lot of modern day games, and many of us are still on them.

  • Randomblame - Thursday, May 21, 2009 - link

    athlon xp 2500 plz Reply
  • lemonadesoda - Thursday, May 21, 2009 - link

    ... but please also add P4EE Gallatin (if you can) to the one or two P4s. Please also add a P3-Tualatin-high end... since these things still sit around in many servers. The performance and power compared to the Atom may determine many upgrade decisions.

    And for enthusiasts, please add dual Xeons and Nehalem-EPs. I chose the route of dual Xeon 5420's at stock and silent rather than the cheaper but noiser route of Core Quads on major overclocks.

    Comparing dual Xeons to i7 to N-EPs would be very helpful indeed!
  • Etern205 - Thursday, May 21, 2009 - link

    Just when to Intel's site and found out the Atom N330 is a dual core with HT.

    Now does that mean if HT is enabled then users will have 4 logical threads?
  • Etern205 - Thursday, May 21, 2009 - link

    Add some Celerons to the mix. Reply
  • tynopik - Thursday, May 21, 2009 - link

    you will quickly limit yourself and make it hard to see differences at the top

    instead of minutes to complete task convert it to frames/sec or MB/sec or something so longer is better
  • tynopik - Thursday, May 21, 2009 - link

    in case i'm not making myself clear, your time based charts are NON-LINEAR

    at the high end, going from 10 seconds to 9 (10% improvement) is far more impressive that going from 200 seconds to 190 (5% improvement), yet the bar representing that 5% improvement is 10 times longer than the bar representing the 10% improvement
  • tynopik - Thursday, May 21, 2009 - link

    where are the T3400's and T6400's and the P8400's and Celeron 585's and Athlon X2 QL-62's and Turion X2 RM-70's and their ilk?

    the atom is mainly a MOBILE processor

    so where are the other MOBILE processors?
  • casteve - Thursday, May 21, 2009 - link

    Another +1 for adding an old Athlon or two into the mix. Looking forward to the GPU ratings as well. It would be nice to include a WoW benchmark with your midrange gaming cards - it may not be a graphics hog, but it's the most popular multiplayer game out there... Reply
  • pattycake0147 - Thursday, May 21, 2009 - link

    I'd really appreciate an Athlon XP or two in the mix the 3000+ would be a decent place to start. I'd also like to see a couple non-HT P4's maybe 3.2 and 2.4 GHz.

    On the side there is currently an error in Bench.
    The notes for the Athlon LE-1620 are just a long list of procs instead of the info on the system.
  • Eri Hyva - Thursday, May 21, 2009 - link

    a Pentium3 @ 1Ghz and a single-core A64 3000+ would be nice.

    They run WinXP very well.
  • solarisking - Thursday, May 21, 2009 - link

    Hell yes, include the old P4s.

    I bought the P4 2.8C, the first mainstream cpu with hyperthreading. It was the last one that held a performance advantage over AMD until the Core processors came along. So yeah, include them. In fact, one like mine with HT would be particularly interesting to see simply because of the HT.
  • Lord 666 - Thursday, May 21, 2009 - link

    Actually the 2.4C was the first maintream Intel HT. Since the 2.4, 2.8, and 3.0 are the same CPUs, but just binned, any of them would be valid.

    I agree with your position as previously noted due to HT's revelance with Atom and i7. Not to mention corporate America still uses them.
  • DLimmer - Thursday, May 21, 2009 - link

    I'd like to see a minimal dual core like the D820 (2.8) Reply
  • faxon - Thursday, May 21, 2009 - link

    I have an Athlon 64 S939 4000+ San Diego which cost like $300 back in 2006 (pre Core 2 /cry). It is currently doing loaner duty with 2GB ram and an HD4670 on an evga NF47 Ultra board. could you guys possibly bench some of these old CPUs? there are probably still a lot of people out there who got these as gaming rigs back when dual core cpus all still cost like $6-900 and didnt give a major boost in gaming performance yet (pre core 2), and im sure there are still a couple of people who are using these rigs even today. it would be nice to compare the atom vs something like this since i would consider buying an atom computer for the same purpose (loaner/backup) as i would consider using my 4000+ rig for. Also, i still have a 2600+ T-bred rev-b AthlonXP with 1GB of ram and a 9800PRO 128mb. if you guys have something like this lying around still, that would be a sick thing to see benched, since im looking to donate this box to a friend soon if i cant find a better use for it (its an emachine with a new gpu ram opticals and PSU). If an atom machine would be a better performer, i might be able to convince him to shell out for something a bit better and get a new system instead of dealing with this old beater lol Reply
  • TA152H - Thursday, May 21, 2009 - link

    I'd like to see a couple of Tualatins thrown in there, for comparison sake.

    I currently run a Celeron 1.4 GHz for one of my machines, and it's actually pretty good running Windows 2000. I don't know how it would run the bloated, but now somehow classic Windows XP, and I'm sure Vista would be horrible on it, but it's fine for the stuff I do on it.

    A 1.4 GHz Pentium III-S would be nice to add to. My feeling is, this will beat the single core Atom, but I am curious about it.

    Most importantly, I'd add the Centaur chips in as well, since they are competitors. Although, really, it seems the Nano (stupid name, is it not?) is somewhere between the Atom and the Core. The previous generation was very simple though, and I foolishly bought one at 800 MHz, figuring it wouldn't be THAT bad. A 400 MHz K6-III+ easily is faster :-P . I would be curious how they ran at high clock speeds though. So, I think one Nano and one C7 would be interesting.
  • swaaye - Thursday, May 21, 2009 - link

    Would definitely like to see Pentium M in there. Athlon XP 3200+ perhaps too. Reply
  • fenderkb76 - Wednesday, May 20, 2009 - link

    I would love to see an Athlon 64 3200+ as this is the first Athlon 64 released in mass (other than the FX that was really a server part). Test it on a K8T800 chipset. By the way, be sure to make it the sledgehammer core 2.0 GHz with 1 MB L2 cache. I built mine in December 2003 with a MSI K8T Neo FIS2R and it's still running strong. I would like to see the difference quantified from something I can buy today. Reply
  • DeadpunkDave - Wednesday, May 20, 2009 - link

    Great feature. I was wondering, not if you could retest everything for folding performance but rather if you knew which of the tests that are used is the most useful for judging folding performance (in Standford's folding@home clients)? Reply
  • escott888 - Wednesday, May 20, 2009 - link

    This is a great quick scan tool. I know that I tend to favor older hardware since it still works for many applications. How about adding few older items just for a relative comparison? i.e. a P3 Tualatin (1.13GHz) or an Athlon MP 1800+ or maybe in a dual or quad arrangement. Maybe it would help people let go of the older tech or give them a reassurance that maybe some of it is still good enough (especially when it really cheap or free) Reply
  • flipmode - Wednesday, May 20, 2009 - link

    Seems it might be helpful to have the option to see the results as percentages, if that is an easy thing to do.

    A Pentium D 3.0 might be nice to have in there since CPUs seem to have more or less topped off at that speed.

    Is there a Via Nano in there? I did not notice. It may not be worth it, dunno.
  • marc1000 - Wednesday, May 20, 2009 - link

    i'm just looking at it and it feels nice. the horizontal height of the bars seems to be a little too big tough.

    about the CPUs, i would sugest something like the 2.4ghz or 3.0ghz Northwood cores P4 (no HT) and some Athlon 64 3200+ single cores. IF (and this is a big if) possible, add a high-end dual-core P4 (like the Pentium-D 950 or 960) and some of the very first Banias CPUs (Pentium M variants).

    that would be nice, because the older CPU's are very common in countries other than the USA. I for example, have 2 computers: a C2D e7200 for me and a Pentium-D 945 for my girlfriend. I just bought my C2D and gave her my old Pentium-D to replace an even older Athlon64 3200+ (that was the replacement for an even OLDER Duron 2400+, who came after a OLDER-THAN-OLDER Pentium3 800mhz...) Well, you got the point.. hehe
  • Nat495 - Wednesday, May 20, 2009 - link

    Perhaps something really old like a 486 or Pentium 1 just to see how all the new stuff stacks up in comparison. Be kind of interesting, I think. Reply
  • judasmachine - Wednesday, May 20, 2009 - link

    I'd like to see maybe a couple of AthlonXPs in the mix too. Thought totally understand that this may be too much work, as you guys are probably busy enough. Reply
  • SpeedEng66 - Wednesday, May 20, 2009 - link

    could you throw in a pentium-m? banias and dothan Reply
  • deputc26 - Wednesday, May 20, 2009 - link

    I'd like to see the top and bottom Proc from each previous major architecture. That way we could get a good feel for historical performance data. Reply
  • lopri - Wednesday, May 20, 2009 - link

    Can we have:

    - Possibly a couple more benchmarks that measure purely single-threaded performance? Other than Sysmark (which tangles single-thread and multi-thread) and games (which are heavily affected by GPUs)

    - Some sort of notes on the platform a CPU was tested on? Granted the difference may be small between 975X and X48, but we're looking at very granular list of CPUs. I think it'd also be relevant, especially if you add power consumption figures. It's not likely anyone will mate E2140 with an X48 board.

    - At least one or two Netburst CPUs in the list? There are many folks with P4's and P-D's, and it'll help them to know what to expect. (Believe it or not, people still buy white-boxes with P4's and P-D's)

    Regardless, it's an excellent tool overall and I applaud you guys for a great job.
  • LostPassword - Wednesday, May 20, 2009 - link

    bench the nano Reply
  • Azsen - Wednesday, May 20, 2009 - link

    Nice work! Reply
  • GeorgeH - Wednesday, May 20, 2009 - link

    I'd actually prefer to see a ~3GHz Prescott instead of a Northwood. From where I'm sitting, the most interesting comparison is between Intel's last real attempt at MHz-centric long pipe spaceheating and its current adventures with super high efficiency. Reply
  • ClownPuncher - Wednesday, May 20, 2009 - link

    If not, that was a pretty popular chip for a long time. Reply
  • casey0102 - Wednesday, May 20, 2009 - link

    I would like to see some laptop benchmarks added, considering most people are buying laptops nowadays to replace their old desktops. This would give people an idea of what kind of performance to expect out of a laptop CPU versus a desktop CPU that they may be familiar with. Reply
  • cparka23 - Wednesday, May 20, 2009 - link

    I like the idea and the exhaustive list of benchmarks, but it'd be nice to be able to whittle down the number of processors to view. For example, an option to show only Core2 Duo processors would be extremely useful if I already have a motherboard to use. Sorting by manufacturer would also be welcome for less specific situations. I realize that there is already a button to hide individual processors, but that's a little impractical with the size of this list, which will only continue to grow.

    Yes, the results will be more telling if you're deciding between two different microarchitectures rather than different clock speeds. I also know that others will know what's what by simply looking at the product number (E#### vs. Q####). For those of us who don't follow that closely but still have a general idea of the improvements between platforms, it's about real-world practicality. I may not need to see the results of Ion vs. Atom mixed in with P4 vs. Athlon. Having a way to sift through the growing amount of data quickly will make this a better tool for the casual-interest crowd.

    Similar to an article that details the performance gains of i7 over its predecessors, I'd like to be able to have the option to see just that data presented in Bench rather than the raw numbers for everything ever tested. If I want to see those results a couple years from now, I want to look it up without having to search through all the old articles in full detail. I might just not need the full article after a quick glance at Bench.
  • ltcommanderdata - Wednesday, May 20, 2009 - link">

    So I'm guessing this will be similar to the interactive charts available at Tom's Hardware?

    Any chance you could through in the best of Netburst architecture as a comparison point. ie. the 3.73GHz Pentium Extreme Edition 965 and the 3.73GHz Pentium 4 Extreme Edition. It'd be good to bring out the old top-of-the-line 2.8GHz Athlon 64 FX-57 as well.
  • mforce - Wednesday, May 20, 2009 - link

    I'd like to see how the Pentium III @ 1 GHz does because I have a strange feeling that even that might beat the Atom. I think there were also some Celerons based on the P III at over 1 GHz.
    If you could compare it to the 1 GHz CPUs of the old , the P III and Athlon that would be very cool.
  • Carnildo - Saturday, May 23, 2009 - link

    The Atom 230 is said to be comparable to a Celeron 900. I'd like to see that tested, as well as a selection of other old CPUs for comparison.

    (Specifically, I've got a Pentium MMX 233, a Celeron 450, and a 500MHz UltraSparc IIe that I'd like to see it compared to.)
  • MrBlonde - Wednesday, May 20, 2009 - link

    For some people seeing data for P4 era chips is still relevant. My boss still runs a P4 2.8Ghz Northwood (B I think) non Hyperthreading model. Of course if you do a Pentium I'm sure somebody will want to see data for an Athlon XP. Reply
  • mino - Monday, May 25, 2009 - link


    Actuall I feel it much more relevant that havin 5 different Sempron-LE from the ssame generation ...

    I would go for at least one result from the:
    Coppermine (i.e. 1Ghz/100)
    Thunderbird (i.e. 1GHz/266)
    Throughbred (i.e. 2GHz/266)
    NetBurst-ed Celeron 3Ghz/400/128k -> this score is VERY important for many BFU's ...
    Northwood C (i.e. 3GHz/800)
    Pentium D (i.e. 840/940 - 3GHz/800)
    A64 754 (i.e. 2GHz/512k)
    A64 939 (i.e. 2GHz/1M)
    A64 X2 939 (i.e. 3800+)
    Conroe 2GHz/2M
    Banias 1.5GHz/400
    Dothan 2Ghz/533

    Simply said, every major CPU line since PIII/Athlon days shall have ist numbers.

  • whydoyoucare - Wednesday, May 20, 2009 - link

    The Atom 230 comes out on top against the Atom 330 in the majority of the benchmarks when I compare them.

    Maybe a data entry error?
  • UNHchabo - Wednesday, May 20, 2009 - link

    I think you're seeing the 230 bar as larger than the 330 bar, but many of those tests are "Lower is better". Reply
  • Fenixgoon - Wednesday, May 20, 2009 - link

    Would it be possible to include these parameters as well? Obviously atom will be smoked by just about everything, but it's the low power consumption that makes it so attractive. Reply
  • ICBM - Wednesday, May 20, 2009 - link

    I think it would be great to throw a VIA C7 and some of the Nano models. The Nano is the real competition. Reply
  • matheusber - Wednesday, May 20, 2009 - link

    definitely Nano is a good call :)

  • Hlafordlaes - Wednesday, May 20, 2009 - link

    Me, too. Reply
  • vshah - Wednesday, May 20, 2009 - link

    i just upgraded from a p4C 3.0ghz to an i920, I would love to see my old processor included in Bench Reply
  • plonk420 - Wednesday, May 20, 2009 - link

    i love it ... except for ONE small thing:

    build 819 of x264 is pretty old... there have been quite a few peryn, phenom (i and ii), and nehalem optimizations since then. if you're looking for NO optimizations across them, it's too new.

    as of 5/5, me posting to another site, the following builds had optimizations to all of the aforementioned architectures:

    commit a5ac6a5b8688915553fe6fccee09f1272f3788ac r1019
    commit 83baa7fdd2edf3e2f9522fc8b79e0826bcf190fc r1028
    commit f9dba8bb274dffb19394db20912823464efcb8e1 r1030
    commit e1013e8152254614696bbc9d92959bc9705d98b1 r1035
    commit 77028cd3671de855affb02ffefe6bbd99ac7816e r1048
    commit 00cef64dd3fff5d4b5b9b0e63314c11bfb7d33e0 r1067
    commit 2dca5f5413051a26cbba4e20f3c77ff69b694ba3 r1122
    commit 10d6ef07409ebe38b5f1e8e4516155a2fe66d4c6 r1125

    (see changelog at"> )
  • dgtljunglist - Wednesday, May 20, 2009 - link

    I'd love to see some more Athlon 64 X2 and Core 1 Duo data, since I'm still running those after 3 or 4 years, and I think it's reasonable to expect that others are as well.

    This is a great tool though! It would be a great novelty to see every processor you could get your hands on, but that's of course not practical.
  • strikeback03 - Thursday, May 21, 2009 - link

    Core Duo was only in laptops, so would require different chipsets and motherboards and is probably not in too many desktops. A few of the older AthlonXP/A64/A64 X2 would be useful for comparison though.

    Maybe a summer intern to go through all the old tests/articles and enter numbers into the bench.
  • BernardP - Wednesday, May 20, 2009 - link

    Thanks for providing this very useful comparison tool. I hope it becomes permanent and is very regularly updated. Reply
  • MrSpadge - Wednesday, May 20, 2009 - link

    I'd like to have some veterans included, which are still used as light-duty working / surfing stations. I'd think along the lines of Athlon XP 2400+ or P3 1.0 GHz. And if you still have a P2 or P1 that would be fun ;)
    (no, I'm not just too lazy to write Phenom)

  • LuxZg - Wednesday, May 20, 2009 - link

    Well, I'd like to second an Athlon 2400+ to 2600+ for a feeling of speed. My old desktop which I used untill about a year and a half ago was around that speed, and some of the computers in my company are Athlons 2200-2500+ as well, so that would be a really nice comparision. If Atom can run as well as one of those, or even bet them - I'll be sold! Reply
  • fishbits - Wednesday, May 20, 2009 - link

    Thank goodness, at least on the GPU side. Been a rash of comparisons like this one"> lately. While I'm sure the numbers are valid, not everyone is upgrading his video card every three months and is up on every last refresh. Maybe they bought their card 2-3 years/generations ago, and want to know just how much improvement they'll get going to a newer one.

    So yeah, being able to see older scores in the same graph with newer ones is a huge benefit for some of us. Likewise, how often do the games used to benchmark GPUs really need to change? Keep a couple of old popular games for consistency, WoW, whatever, so I know the relative increase in power between the old card and newer ones. There are some sites *cough* that do have performance charts that have a meaningful history length, and it's very appreciated.

    Keep at least a couple of representative cards from each generation shown, with more complete listings from the current gen. And if benchmark games aren't picked in a faddish fashion (Age of Conan? GRID? Really?) then it could be set up where exact old cards are selected from a drop-down box to minimize visual clutter in the presentation.
  • Jackattak - Wednesday, May 20, 2009 - link

    Couldn't agree more.

    I upgrade my GPU every 1.5-2 years and would love to see how it stacks up to the newer ones and how much potential performance improvement I'll be getting.

    Same goes for CPU's, although my tastes have been quenched (currently have a Core 2 Duo 6850 3GHz and was happy to see the bast improvement that a new 8600 3.33GHz would bring to the table).

    I wouldn't mind seeing an HT 3.2 GHz P4 in there just for giggles if you have one laying around, just to see how it would stack up to the latest and greatest Core 2's with their insane cache levels nowadays.

    Can't wait for the GPU's to hit!

    Thanks for setting this up, Anand. Greatly appreciated!
  • Dudler - Wednesday, May 20, 2009 - link

    All hail the new list :)
    Plz do the same with GPU's, THG's list is utter faeces. (Look it up)

    Greetings from the land og the midnight sun, Norway :-D
  • Lifted - Wednesday, May 20, 2009 - link

    I'd be more interested in seeing the highest clock single core P4 (3.8 or 4GHz?). That and a dual core P4 should be enough to get a good understanding of the performance at the time. I don't actually know anyone who bought a dual core P4, but if you have one laying around you might as well. Reply
  • Syran - Wednesday, May 20, 2009 - link

    I'd like to see an Opteron 165 or something along those lines. And maybe a low end hyperthreaded P4. Reply
  • Lord 666 - Wednesday, May 20, 2009 - link

    The Northwood 2.4C or 3.0C would be great baselines. Reply
  • chitra - Thursday, May 21, 2009 - link

    How about P4 Prescott? I still have one in my PC.. and would like to see some benchmarks with it :) Reply
  • mino - Monday, May 25, 2009 - link

    Prescott ~ Nortwood + some heat. No need to rerun the same numbers besides Pentium D 900 series.

    Also some Gallatin score would be interesting :)
  • mmp121 - Wednesday, May 20, 2009 - link

    I also vote for the P4 3.0C Northwood

    Being as I am currently still running it in my main rig, I would LOVE to see how much of a boost I'm going to get when I build me a new rig after 6+ years this xmas.
  • suppliesidejesus - Wednesday, May 20, 2009 - link

    Definitely would like to see some Opteron scores in there. Reply
  • maxxcool - Wednesday, May 20, 2009 - link

    additional thought, could there be a set of check boxes in the "comparison section" indicating the winner of the category, and if possible by what % it out performed the opposition... Reply
  • ssj4Gogeta - Wednesday, May 20, 2009 - link

    Yes that's a nice idea. The percentages could be green text when the first chip wins and red when it loses. Reply
  • erple2 - Wednesday, May 20, 2009 - link

    I really like that idea. I had a little bit of trouble looking "at a glance" which processor "won" any particular benchmark - mainly because for some benchmarks, you want lower numbers, and for others higher. Percentages are nice, too. - you can use the slower of the comparison as the "baseline" and show the percent improvement of the second one. That would give me a great idea just how much percent increase in performance moving to a "newer" processor would be.

    However, I do like this quite a bit. Someone complained about the margins - they're really not as bad as people claim.
  • maxxcool - Wednesday, May 20, 2009 - link

    IE, Desktop 35/45watt cpu's from intel and Amd in reduced speed modes (like 1ghz) to compare against dual core atoms, and or high speed 1.8 - 2.0 ghz single core atoms?

    As for old p4's? meh... I would be more interested in the single-core Core-dou variants, Core based Celerons and single core Semprons for budget builds or for cheap office pc builds running data entry/records processing/word/etc.

  • jhon1234 - Wednesday, May 27, 2009 - link">">">">
  • cosmotic - Wednesday, May 20, 2009 - link

    What's the deal with the huge margins on each bar? They are twice as high as they should be and 4x as high as they could be.

    Also is titled 'Untitled Page'.
  • evilspoons - Wednesday, May 20, 2009 - link

    I agree; there's a lot of wasted vertical space. I'd like to see it tightened up.

    Otherwise: great feature, I'm glad to see it, please keep it updated! It's the only thing you guys were missing vs. other places like THG.

Log in

Don't have an account? Sign up now