Comments Locked

48 Comments

Back to Article

  • Future1investor - Tuesday, August 11, 2009 - link

    I bought this unit in Oct 2006 except instead of the ATI video it has the GF7950GX2-1GB NVIDIA card. I also have a Depot or Onsite 3yr Warranty, which is the issue here.

    After the liquid cooling pump went out and I started experiencing various problems likely to be related, ABS would not give me onsite service. Instead they wanted me to remove the Liquid Cooling unit, and ship me an air-cooled replacement. After much pain in communications, they then expected me to send it back and then ship me another liquid cooler. But I said with the other problems the whole system needs to be diagnosed properly on the bench.

    I shipped the whole computer back at my expense of $87 through UPS. When they received it they again balked saying that they "don't do that" which I interpreted as bench diagnostics. This was confusing because under warranty, the risk of the average user damaging something can be rather high if they are asked to do their own diagnostics and replacements (even given phone assistance). But what of the 3Yr Onsite warranty?

    Anyway, long story short; they agreed to bench test and fix what was needed. They told me after a week that they had replaced the liquid cooling unit and the graphic card. I also had a problem with the OS recognizing an additional 4GB Ram same timing, same brand. They said that they'd test for that too but I havn't heard the results. They have closed the RMA and I assume the computer is on route back across the country to me.

    Communication has been poor. I don't know if my additional 4GB Ram will be utilized. They changed the Onsite warranty to Depot only. I had to pay for packaging and shipping back to their dock at $87. A few months ago, a phone call to support, to talk about the Ram was met with odd laughter. The person on the phone said it was to be expected for buying that operating system (Windows XP Professional) the only thing available at time of purchase.

    This was my second computer purchase from ABS. The first with a 3Yr Onsite Warranty. The tech support is sketchy. They will only call you back from an unidentifiable phone number. They want you to pay for shipping along with any replacement parts, saying that they'll reimburse you once they receive the defective pieces. They want you to do your own repairs. And only if you push the issue to the wall will they attempt normal customer satisfaction (though I paid for it with additional costs to me).

    Perhaps ABS' business is not as good as it used to be, and perhaps they are cutting corners and cutting up warranties at the expense of future business.

    Considering ABS for a future purchase looks doubtful. Like customer service from Dell. It is a customer nightmare scenario. Though Dell is much much worse! I will be focusing more attention on the warranty service given though Digital Storm. They have one of the absolute highest customer service ratings of any boutique computer retailer. Though I will still monitor the rating in this area for ABS, as perhaps things will improve greatly?
  • appu - Saturday, August 26, 2006 - link

    Just came across this yesterday - http://www.velocitymicro.com/">http://www.velocitymicro.com/

    How about test-driving one of their systems (although I think we can hazard guesses at the results for the most part)?

    These guys also seem like a really nice alternative to building your own box if you don't like generic OEM stuff and don't have the time to put one together yourself. I was definitely impressed with the 15-20 mins I spent on the site looking at various options.
  • Justin Case - Sunday, August 20, 2006 - link

    One month ago all articles around here were saying what a brilliant combination Core + Crossfire was. Now AMD buys ATI, and all of a sudden the conclusion changes to "Crossfire is immature". Interesting...
  • yyrkoon - Sunday, August 20, 2006 - link

    "The second item worth nothing is that the NVIDIA SLI solution outperforms all of the X1900 CrossFire configurations, including the 20% overclocked ABS, even when using a slower CPU. Other games might perform better on CrossFire, but there certainly appears to be plenty of room for improvement in the ATI drivers, particularly when it comes to Core 2 optimizations."

    Saying that ATI needs to iron out thier drivers for core 2 duo is hardly concluding thats 'Crossfire is immature'. Of course, I can pretty much twist something someone says in any dirreciton I want to also. However, I think taken in with the rest of the context, they are trying to say something along the lines that performance wise right_now, Crossfire boards seem to underperform the nVidia varients, but this doesnt mean its going to hold true in the future.
  • JarredWalton - Sunday, August 20, 2006 - link

    On AM2/939, CrossFire often beats out 7900GTX SLI. On Core 2 *right now* it doesn't, at least in the tests I looked at. (I don't have a full SLI 775 Core 2 setup right now, unfortunately.) I've *always* felt CF left a bit to be desired, and I'd say the same thing about .NET and CCC. For whatever reason, some of the issues with CF were magnified on this ABS system. Seriously, THIRTY SECONDS after XP shows the desktop before CCC is finished loading. That's terrible. It's also about 3X as long as a 939 ASUS A8R32-MVP for the same thing, which is why I conclude that CF on C2E has some issues. Of course, the 6.8 drivers may have just addressed a bunch of my concerns - I have to retest that before sending the system back.
  • yyrkoon - Monday, August 21, 2006 - link

    I will agree with you that 1 minute OS load times is terrible, hell even my budget Asrock AM2NF4G-SATA2 / 3800+ system boots XP in 18 seconds, and thats from IDE. Not that its cutting edge technology either . . .
  • JarredWalton - Monday, August 21, 2006 - link

    It's about 18 seconds to load the OS, but it's 33 seconds of POST codes and 35 seconds of ATI/.NET after XP is basically loaded. With RAID 0 Raptors no less. And people wonder why we at AnandTech often say NVIDIA seems to have more robust solutions....
  • yyrkoon - Monday, August 21, 2006 - link

    Well, when I said 18 seconds, its 18 seconds to boot to desktop, and about 1-2 seconds more and its ready to use, however, I just installed the 2.0 framework for microsoft web dev express, and havent rebooted yet. so perhaps it'll take a little longer now. I still dont have 1.1 installed, but on this machine (my gaming rig) I dont plan on installing it, thats what my dev box is for :)
  • JarredWalton - Monday, August 21, 2006 - link

    .NET should only affect load times when it is used - i.e. by the ATI CCC drivers. I don't think just installing it really impacts performance. I have it installed on most of my PCs - both 1.1 and 2.0 - and have never noticed any delays on booting, except when using an ATI card.
  • KrazyDawg - Monday, September 4, 2006 - link

    I never really noticed that the ATI drivers utilizing .NET was the cause of a longer boot time. In the past, I used omega drivers for my Radeon 9800 Pro but stopped using them when the latest release failed to install properly. I haven't tried other drivers since but a driver review would be interesting and beneficial to many.
  • yyrkoon - Tuesday, August 22, 2006 - link

    Ah, I was wondering why you were mentioning .NET in technologies you were unhappy with. .NET is really a boon for programmers, and even more so for hobbyist programmers such as myself (it makes things much easier, and faster to code very usefull applications ).
  • JarredWalton - Monday, August 21, 2006 - link

    Of course, after looking at some of the results on the AM2 RD580, maybe SLI *is* faster in most situations. We need to test additional games to say for sure, which I think Derek will be doing on his next GPU article.
  • yyrkoon - Sunday, August 20, 2006 - link

    I think they were talking about the intel crossfire implementation is immature. Irregaurdless, comparred to nVidia technology, Crossfire technology IS Immature.
  • giantpandaman2 - Sunday, August 20, 2006 - link

    Here's my suggestion to have better system reviews:

    1) A ratings system where you break down things like value, game performance, work performance, ease of use/documentation, customer service and the like. Similar to CNET and HardOCP.

    2) A greater focus on the complete end user experience. How easy was it to order the system? How long did it take to get it after ordering. Was the packaging good? How was customer service when you had a part fail (made up or real)?

    3) A handy list of comparable systems from other vendors w/reviews of those systems if you have them. Or if there are other respectable sites that have reviews of them. (Now this is a pie in the sky type of request, but it'd be an extremely nice little function.)

    4) A small separate box for the "reviewers tilt." In other words, whether the reviewer would buy the system or get something else. For example, for the added cost would you buy a mac? Would you prefer a different type of memory? What concerns do you have with the system? IE-Case cooling would be poor if you added a second video card. Etc. This would be totally subjective, hence why it'd have to be offset from the normal review.

    5) A chance for the vendor to respond to any questions or criticisms you had with the system and/or customer experience.

    Now, honestly, given your focus on less consumer oriented computer stuff I don't know how far you'd want to go with any of these. As much as I understand why people dislike HardOCP a couple of things they've put in that are nice is a focus on customer experience in system reviews and they're slight separation of consumer oriented reviews as compared to enthusiast reviews.

    Personally I'd love it if Anandtech created more focused regular content based around Business IT/Consumer/Enthusiast rather than your current back end classification of stuff. (Motherboard/CPU/Video/etc.) Why? Well, I think you'd be able to up your number of reviews. Quicker less technical reviews for a lot of the consumer stuff. IE-You could quickly do reviews of a lot of cheap digital cameras and OS iterations. And far more technical and in depth reviews of expensive cameras or OS iterations when used in business/network settings.

    Heck, you could almost do both types of reviews at once, get out the consumer one early, then the more in depth one later. That way low end consumers don't have to wade through a lot of stuff they don't understand or may not be interested in, and gurus can get the fix they need. Anyhow, just a few ideas out of many, but this post is getting too long. :)
  • yyrkoon - Sunday, August 20, 2006 - link

    I'd have to agree somewhat. Breaking down how a system performs into categories I think is a good idea. Something along the lines of how TH does it but perhaps with your own little twist. Doing two seperate reviews on a product however, I would think is a bit overboard. People like me, who are very interrested in the technical aspect of hardware, will pick and choose thier articles, which may give you less readers for those type of articles, but I think the over all experience for everyone would be fine. My suggestion for this type of thing, would be maybe to make a dumbed down review, if you think readers of said article may not be interrested in the technical aspect, and just leave it at that. *shrug*

    I still think you guys do a bang up job, and it isnt fanboyism when I repeat the words "I'll read your articles over tomshardware.com's reviews any day". It just seems to me, that you guys at AT are less worried about how the manufactuers feels about your reviews, and call it how you see it.

    Now, for a couple of gripes:

    1) PLEASE work on your forums, its ugly, not very organized, and the over all experience just doesnt "leave a good taste in my mouth".

    2) I've sent you guys an email concerning this issue, Animated ads within the text area of your articles are very annoying, and make it sometimes very hard to read / concentrate on the article on hand. Since this, I've disabled all images within my browser from your site, but what IF I want to veiw a photo relating to the article ?!

    Take these gripes for what they are worth, and not personally please, it is my hope that these gripes will help you improve your reader overall experience whilst on your site, and forums.
  • chunkychun - Sunday, August 20, 2006 - link

    I am a non techie and I am glad you review the OEM systems, high end or low end. Although I don't have the time or patience to learn how to build a computer, I enjoy reading which components are perfered by more experienced builders. Keep up the good work.
  • SunLord - Saturday, August 19, 2006 - link

    Now we jsut need Anandtech to get a ABS Ultimate M6 Sniper AM2 speced out with as much incommon as possible and see what the performance difference is on a real world systems...
  • JarredWalton - Saturday, August 19, 2006 - link

    It depends on what you're doing. If you're playing games at high resolutions, the difference in performance is negligible. If you're doing video encoding or 3D rendering or certain other tasks, Core 2 Duo is still quite a bit faster -- and even more so when overclocked to 3.52 GHz. If you want to buy a $4000 computer, I really see no reason to purchase an AMD system right now. However, I don't think most people should spend that much money on the computer system regardless of processor type. You can build a system that is almost as fast (in games) for about $2000 -- using either Intel or AMD processors.
  • pottervillian - Saturday, August 19, 2006 - link

    One Error:

    Page 2: Features and Price
    Component Price List
    "Power Suply: Enermax Liberty 620W Modular PSU 155"

    Other than that, this is a great article. Too bad I don't have a rich uncle.

  • eastvillager - Saturday, August 19, 2006 - link

    If you had a rich uncle, you'd be better off getting cash and building it yourself, or paying a friend to do it. These prices are crazy.
  • JarredWalton - Saturday, August 19, 2006 - link

    As I pointed out on page 2, the price charged by ABS really isn't bad. I mean, $300 extra to build a super high-end setup, relative to what you'd pay at Newegg? That's not to say I think buying ultra-high-end isn't a bit crazy anyway, but getting all the watercooling and such installed isn't a 30 minute task. Anyway, you can always build a better PC for less money on your own, but you also then are fully responsible for supporting that PC if anything goes wrong. I'm fine with that, most of the people posting are fine with that, but there are a lot of people out there that would rather pay someone else to do the work. I've charged people $200 or so in the past to build a moderate system, and that's probably less than I would charge now.
  • plewis00 - Friday, August 18, 2006 - link

    Does anyone else think it's absolutely criminal to spend that amount of money on a system which would then turn out not to be 100% stable? In fact, any system which isn't 100% stable out of the box (unless by some freak one-off accident) I would consider a failure, like this one.
  • DigitalFreak - Friday, August 18, 2006 - link

    If you're going to spend that much on a system, you should be dealing with Alienware, VoodooPC, etc. This ABS system appears to be junk.

    BTW, you guys should take a page from Hardocp on how you conduct your system reviews.
  • rjm55 - Saturday, August 19, 2006 - link

    -H- has turned into a joke in the mind of any real enthusiast. Stupid reviews like the Core 2 Duo launch "proving" AMD outperforms Conroe are posted, and then any one who questions his majesty is banned from making comments. You end up with King Kyle and his adoring "yes massah!" loyalists around his feet. Every REAL enthusiast I know has been banned from -H- at least once for asking questions - it's almost become the red badge of courage to be banned form -H-. Compare that to AT where hard questions and comments always stay up, are encouraged, and actually debated by Editors.

    Every one knows -H- only rewards Editors Choices to whoever bought his last truck. His recent Editors Choice to the ECS KA3 MVP has everyone I know rolling on the floor laughing. You start with the idea you need to give an Editors Choice to ECS. You run benchmarks that are meaningless and very few, that you know will "prove" the ECS was a great board. You ignore the fact this is the AM2 board that is the WORST overclocker anyone has ever tested - and mention quickly it doesn't even have the means to adjust CAS latency and runs Corsair 1066 at 800 at CAS 5 timings AT BEST (every other AM2 board can run it at 3-3-3). You can bet the award was payment for the new truck ECS bought this year. Based on the fact -H- has recently moved from giving anything Abit an Editors Choice to now trashing Abit, so it's a pretty safe bet Abit isn't buying the trucks any more.

    You don't have to believe what everyone who tells you about -H-. Just go to -H- and ask a hard question about test methods or how a conclusion was reached. You'll be banned so fast it makes your head swim. Hitler also surrounded himself with yes men, as apparently do many other Texans who avoid reality.
  • JarredWalton - Saturday, August 19, 2006 - link

    Careful! We might ban you from our forums for turning on white text! ;)
  • rjm55 - Saturday, August 19, 2006 - link

    has turned into a joke in the mind of any real enthusiast. Stupid reviews like the Core 2 Duo launch "proving" AMD outperforms Conroe are posted, and then any one who questions his majesty is banned from making comments. You end up with King Kyle and his adoring "yes massah!" loyalists around his feet. Every REAL enthusiast I know has been banned from {H] at least once for asking questions - it's almost become the red badge of courage to be banned form . Compare that to AT where hard questions and comments always stay up, are encouraged, and actually debated by Editors.

    Every one knows only rewards Editors Choices to whoever bought his last truck. His recent Editors Choice to the ECS KA3 MVP has everyone I know rolling on the floor laughing. You start with the idea you need to give an Editors Choice to ECS. You run benchmarks that are meaningless and very few, that you know will "prove" the ECS was a great board. You ignore the fact this is the AM2 board that is the WORST overclocker anyone has ever tested - and mention quickly it doesn't even have the means to adjust CAS latency and runs Corsair 1066 at 800 at CAS 5 timings AT BEST (every other AM2 board can run it at 3-3-3). You can bet the award was payment for the new truck ECS bought this year. Based on the fact has recently moved from giving anything Abit an Editors Choice to now trashing Abit, it's a pretty safe bet Abit isn't buying the trucks any more.

    You don't have to believe what everyone who tells you about . Just go to and ask a hard question about test methods or how a conclusion was reached. You'll be banned so fast it makes your head swim. Hitler also surrounded himself with yes men, as apparently do many other Texans who avoid reality.
  • Kalessian - Saturday, August 19, 2006 - link

    whoa, how did you manage to break the comments system?
  • JarredWalton - Saturday, August 19, 2006 - link



    Urg... just to recap, don't post the {H} abbreviation with brackets instead of braces. You know how {B} and {I} do bold and italic? {H} does highlighting, which in this case means white text. We should probably disable that feature, though it does make things somewhat entertaining. ;)
  • rjm55 - Saturday, August 19, 2006 - link

    What a f$#&ing joke!! has turned into a joke in the mind of any real enthusiast. Stupid reviews like the Core 2 Duo launch "proving" AMD outperforms Conroe are posted, and then any one who questions his majesty is banned from making comments. You end up with King Kyle and his adoring "yes massah!" loyalists around his feet. Every REAL enthusiast I know has been banned from {H] at least once for asking questions - it's almost become the red badge of courage to be banned form . Compare that to AT where hard questions and comments always stay up, are encouraged, and actually debated by Editors.

    Every one knows only rewards Editors Choices to whoever bought his last truck. His recent Editors Choice to the ECS KA3 MVP has everyone I know rolling on the floor laughing. You start with the idea you need to give an Editors Choice to ECS. You run benchmarks that are meaningless and very few, that you know will "prove" the ECS was a great board. You ignore the fact this is the AM2 board that is the WORST overclocker anyone has ever tested - and mention quickly it doesn't even have the means to adjust CAS latency and runs Corsair 1066 at 800 at CAS 5 timings AT BEST (every other AM2 board can run it at 3-3-3). You can bet the award was payment for the new truck ECS bought this year. Based on the fact has recently moved from giving anything Abit an Editors Choice to now trashing Abit, it's a pretty safe bet Abit isn't buying the trucks any more.

    You don't have to believe what everyone who tells you about . Just go to and ask a hard question about test methods or how a conclusion was reached. You'll be banned so fast it makes your head swim. Hitler also surrounded himself with yes men, as apparently do many other Texans who avoid reality.
  • rjm55 - Saturday, August 19, 2006 - link

    has turned into a joke in the mind of any real enthusiast. Stupid reviews like the Core 2 Duo launch "proving" AMD outperforms Conroe are posted, and then any one who questions his majesty is banned from making comments. You end up with King Kyle and his adoring "yes massah!" loyalists around his feet. Every REAL enthusiast I know has been banned from {H] at least once for asking questions - it's almost become the red badge of courage to be banned form . Compare that to AT where hard questions and comments always stay up, are encouraged, and actually debated by Editors.

    Every one knows only rewards Editors Choices to whoever bought his last truck. His recent Editors Choice to the ECS KA3 MVP has everyone I know rolling on the floor laughing. You start with the idea you need to give an Editors Choice to ECS. You run benchmarks that are meaningless and very few, that you know will "prove" the ECS was a great board. You ignore the fact this is the AM2 board that is the WORST overclocker anyone has ever tested - and mention quickly it doesn't even have the means to adjust CAS latency and runs Corsair 1066 at 800 at CAS 5 timings AT BEST (every other AM2 board can run it at 3-3-3). You can bet the award was payment for the new truck ECS bought this year. Based on the fact has recently moved from giving anything Abit an Editors Choice to now trashing Abit, it's a pretty safe bet Abit isn't buying the trucks any more.

    You don't have to believe what everyone who tells you about . Just go to and ask a hard question about test methods or how a conclusion was reached. You'll be banned so fast it makes your head swim. Hitler also surrounded himself with yes men, as apparently do many other Texans who avoid reality.
  • JarredWalton - Saturday, August 19, 2006 - link



    Don't use the (bracket) H (closebracket) abbreviation in our comments - that's the code for highlight enable, which makes your text white. LOL
  • JarredWalton - Saturday, August 19, 2006 - link



    Multiple uses cause even more troble. :p
  • JarredWalton - Saturday, August 19, 2006 - link





    Have I shut off all text highlights yet? Enquiring minds want to know!
  • JarredWalton - Friday, August 18, 2006 - link

    Take a look at similarly configured Alienware, Falcon NW, VoodooPC, etc. systems and see what they cost. $4500 for the ABS is an absolute bargain by comparison. And don't think that all of those systems are going to be more stable than this ABS. I tried a Quad SLI config for a while that was much, MUCH less stable than this system. The review would have been very unfavorable, but as the company basically pulled the configuration I had there was no reason to "review" a phantom product.

    Alienware is about $4300 without overclocking or watercooling.
    FNW Talon with E6700 and 7950GX2: $6000
    FNW Mach V with X6800 watercooling and CrossFire: $8230
    VoodooPC Omen CrossFire: $7600

    None of those come factory overclocked as far as I can tell, but I'm sure they could be. Still, there's a big difference between $4500 and $6000+. Are they more stable? I don't know - I haven't tested any of the above configurations.

    As for ways to improve system reviews, I'm open for suggestions. Simply stating "be like HardOCP" doesn't really help much. What would you like to see added? Compare what this review told you about gaming to what others show, and let me know what's really different about the conclusions. Where do we fall short? I have plenty more system reviews, so if there are specific improvements to be made I'd love to try and make them.

    Thanks,
    Jarred Walton
    Editor
    AnandTech.com
  • yyrkoon - Friday, August 18, 2006 - link

    Personally, I find your articles fine the way they are, whihc is why I'll read YOUR articles over Toms any day of the week (because you guys seem to tell it how it is, and how you'd think the company you're reviewing would want to hear it.
  • yyrkoon - Friday, August 18, 2006 - link

    err and NOT how you'd think the company you're reviewing would want to hear it (left out 'not'), sorry.
  • MarkHark - Friday, August 18, 2006 - link

    Thumbs up for you, Jarred, twice in fact!

    First for a well-thought, well-written review, second and most important for your last comment.
  • samuraiBX - Friday, August 18, 2006 - link

    now if one of you will please run over me so I can get the money from my insurance... :P
  • JarredWalton - Friday, August 18, 2006 - link

    We've got a $1400 system review coming soon, for "the rest" of our readers. :)
  • yyrkoon - Friday, August 18, 2006 - link

    Sub 1,000 system review would be much bettter ;) Anyhow, nice clean looking wiring (until you look into the drive bays), additional fan on a water cooled CPU ?! 3000-4000 Price ?!

    Personally, I would never consider buying a OEM type system, but man even though it has semi clean wiring, neat gadgets like a Air cooling assisted water block, excellent non gaming performance, I would expect said system to be stable, offer the best gaming performance period, and to cook for me . . .
  • yyrkoon - Friday, August 18, 2006 - link

    Ah sorry, forgot to add, or ask rather, why do tech sites such as yours review OEMs anyhow ? Server equipment I can see, but a gaming / desktop system, I could never understand. Most people who read your site are very technical, and while I cannot speak for everyone, I think its personally a waste of time for you, and your readers(since most of us build our own systems anyhow). Also, again while I cannot speak for everyone, all my systems have_to_be stable, and when I build my own system, they usually either start off stable, or get fixed in the building the system (ironing out flaky software issues). I do not see how a OEM, who is serious about thier product, can put out something that isnt stable, I mean they have a hell of alot more resources than the average Joe such as myself, and yet, ever system I've built over the last 6-8 years (been building my own systems since 92 or earlier) has been 100% stable, usually from the get go. Then again, those of us who are tech savy, know that you never skimp on 3 parts EVER, Motherboard, memory, and PSU . . .
  • JarredWalton - Friday, August 18, 2006 - link

    I don't think ABS intentionally skimped on anything here. The problem is simply that they're working on the bugs on a design - I'm not sure if the GPU watercooling is actually coming on any systems right now, but they have said that a new GPU cooling solution should be ready in a month or so.

    As for why we review OEM systems, I tried to make that clear in the intro. 1) Not all people build their own PCs. 2) Core 2 is currently a bit difficult to find at retail, and you have a better chance of getting it in an OEM system As for the readers, trust me when I say there are a LOT of readers that never read the comments section or post here; with literally millions of readers (Anand could give actual numbers), we basically cater to every corner of the market.

    Speaking of 100% stability, I'd wager a lot of systems would be somewhat unstable under the extreme test scenario we used. Two instances of Folding@Home along with running games (real games and not just a demo loop or 3DMark) for hours on end can cause problems on a lot of the overclocked "100% stable" enthusiast systems. I've had Dell and other OEM business computers take a dive in such a scenario.
  • Xenoid - Saturday, August 19, 2006 - link

    Although not all people build their own computers, I'm sure that if you set up a poll, you'll see that a huge majority here do, so spending all the time on this article is just good comedy for me to laugh at what 13 year old's father is going to buy this for his kid so he can play Counter-Strike on while running 5 viruses simultaneously.

    As for the stability under 2 instances of F@H with a game on for hours at end... I'm sorry but who on this website is purchasing a computer here for NASA? Anyone? We're not counting the extreme nerds running F@H for a few years until they manage to lose their virginity. Running math programs for hours on end is not exactly my idea of fun or productiveness. There's a huge difference between stability and "stability in case of a nuclear meltdown while running 30 instances of F@H while your wife is screaming at you and your kids are playing with matches".

    Stability for most of us means we can run a few programs (a media player, a game, a bittorrent program, a trillian-esque program, and a few other things) while focusing on the game experience and making it top notch and not crashing.

    And I wouldn't doubt the ability of a Dell or other OEM business computer to take dives in any situation. That's why most of us don't buy OEM. We like to know how it's getting put together. If you need any more go find some pictures of those Apple laptops that used to ignite themselves and what they looked like inside.
  • JarredWalton - Saturday, August 19, 2006 - link

    Most people may not plan on running two instances of Folding@Home while playing games, but doing an SMP capable video encode (i.e. DivX) while running a game would place a similar stress on a system. Do you all want to know that a system can run stable under moderate use, or do you want to know that it is truly stable? I suppose we could just run SuperPi 32M once, and if the system completes without errors it must be fully stable, right?

    Who really knows what programs are going to be like in another year or two -- we might get games that can fully stress both processor cores on their own, along with completely stressing the GPUs. Certainly there are no current games that do that, as Quake 4 tops out at around 75% CPU usage with two cores and that's higher than Oblivion and Call of Duty 2.

    Anyway, what else are people going to use quad core or even octal core systems to run at home? At least with Folding@Home people can feel that their computer is doing something potentially useful with its spare clock cycles rather than just consuming electricity. It's also somewhat humorous that you talk about all of the people here who build their own systems and then make fun of the "extreme nerds" that run Folding@Home. (Most "normal" people in the world would consider both parties to be "extreme nerds", but I digress.)

    The bottom line is that you can't call a system 100% stable and less you truly stress test the system. As any experienced overclocker can tell you, there are really only two types of systems in the world: those that appear stable because they have not crashed yet, and those that you know are at least somewhat unstable because they have crashed. Under the right conditions, I think the vast majority of computers are at least partially unstable. I guess that's okay, though, and we should all return to the days of Windows 95 when daily reboots were practically required?

    Still, the system I'm writing this on is a socket 754 3200+ Newcastle overclocked to 2.42 GHz, and it routinely runs for weeks on end without rebooting. About the only time I reboot is when I have to install one of the updates from Microsoft. It also runs Folding@Home 24/7 in the background. If that eventually causes my CPU to fail, it will probably be time to truly upgrade anyway. (Odd that I haven't switched over to one of my newer systems, but this one just does everything I need it to do, other than intense gaming. It's kind of nice to have a computer where I don't need to worry about gaming performance, as probably 95% of my computer crashes that have occurred in the past five years happened when I was playing a game.)
  • yyrkoon - Saturday, August 19, 2006 - link

    Yeah games cause alot of crashes on PCs but thats not to say its the hardwares fault. If it happened all_the_time then possibly, but the random crash once in a blue moon should not be attributed to hardware issues (atleast IMO).

    Re-read the post above where I said my PCd frequently stay up for months on end . . .;)
  • yyrkoon - Saturday, August 19, 2006 - link

    PCs, sorry, you guys really need an edit function hehehe
  • yyrkoon - Friday, August 18, 2006 - link

    Well thats kind of harsh on any system, running all those apps at once, but I've always been in the school of thought that IF you're going to game, game, and dont try to run 500 application all while doing so. If you really want to crash ( or maybe just CTD a game), run Oblivion ;) Still. if you managed to crash any system running all that, I would consider that a software issue (OS, or appliction), and not hardware.

    As for ABS intentionally skimping on hardware, well, all it takes is a little research, and all those worries go away (usually). I did notice they are using a tier 1 motherboard, which is probably a good idea, but I will bet any one of us could build the same system for half the cost. Anyhow, with sites such as yours, reviewing motherboards, memory, and such, there is no excuse for anyone really, to buy 'crappy' equipment, hell, I only paid $54 usd for my current motherboard (AM2 system), but its very solid . . . care to guess who made it ? At any rate, its merely a steping stone in the dirrection of where my system will be this time next year.

    Anyhow, what I consider 100% stable, is a PC that is used normally (which for me, I use my systems all day everyday), for gaming, and work, and able to stay up for months on end without having to reboot (not including installing drivers, etc.), BSoD, or crash applications constantly. I havent seen a BSoD on any of my systems in a long time (years?), because of hardware issues. This is not to say I'm special, this is just to say that even in windows, its very possible to build a rock solid system (in-exspencively), with uptimes approaching Linux server status, you just need to pick your hardware carefully. Overclocking shouldnt even be used in the same sentance as stable in my opinion, these systems MAY run fine for a few years, but OC'n will usually eventualy kill a system, and I consider it a good thing, if systems last well past 3-5 years.

    Now, to be honest, I must say I only skimmed this article, since I'm really not interrested in OEM systems (well, I usually recommend Dell to clients for a cheap OEM systems, depends on needs), but I personally havent had any problems finding C2D CPUs (except for the 6800 extreme CPUs, which is out of my price range anyhow).

    Now, if I may offer a suggestion, I personally think it would be cool if you guys linked these comment sections to your forums.
  • samuraiBX - Friday, August 18, 2006 - link

    Hehe, thanks! I'm looking forward to the "I don't have to sell my kidney on e-bay" computers!

Log in

Don't have an account? Sign up now