Back to Article

  • trexpesto - Tuesday, March 08, 2005 - link

    Or you could just ask on the forums like everyone does..

    #7 and others:
    it is interesting to note that if you do have a fair degree of confidence in parts of an equation, the more factors you include, the more likely the errors will cancel out.

    Unless you are consistently an optimist or pessimist!

    I believe Enrico Fermi popularized this strategy, famously estimating the yield of an atomic blast by throwing torn-up strips of paper in the air. That part sounds suspiciously theatrical.

    A classic one you hear about is a test question that asks the circumference of Earth. Well there are 3 one-hour time zones in the continental US, which is about 3000 miles across. Every thousand miles = 1 hour * 24 hours/day = 24000 miles.
    Actually it's 24,901.55 miles at the equator but that's not toooo bad.
  • Gioron - Sunday, February 06, 2005 - link

    I'm a bit late commenting, but oh well.

    Looking through this, its an interesting way to view upgrading, but I feel its a bit oversimplified and overcomplicated at the same time. Its oversimplified because you're just assuming a linear fit for most of the graphs, when things aren't really linear. Its overcomplicated because you're trying too hard to account for every single variable and making your explanations too complex for most people to really grasp. On the bright side, the basic concept you're trying to get across is something that a lot of people could really use and seem to overlook when making buying decisions.

    As an example of how you're oversimplifying things, consider your "cost to not upgrade" that you're considering a flat $0.25. In reality, the cost to not upgrade is going to increase every day you wait. Its more of an exponential increase instead of a linear line, but there are large steps in the value as you start running into more things your current computer just can't do. As an example of this, my computer 3.5 years ago was fairly high end, it could play all the games, ran the current windows version well, and there wasn't a whole lot out there that was better. Sure, I could upgrade it, but the cost to me would only be about $0.01 a day, mainly from bragging rights. A year later, it was still a good computer, but new games had come out, a new version of windows had come out, and it was being asked to do new things. The games still ran at decent frame rates, but they could be better, windows didn't spend too much swapping out its now medium amount of RAM, but it was noticable now. At that point, upgrading the computer would be worth about $0.05 a day for me. The next year, new games came out where its performance dropped, new software came out that taxed it a bit, and I would definately see an improvement if I upgraded. At that point, it would be worth $0.25 for me to upgrade. A year later, you start hitting things is just plain _cannot_ handle. Can't run the latest games, processor can't handle real time video encodes that I wanted, etc. If I were still using it, it would be worth at least $2.50 a day for me to upgrade it. This is not a linear trend, but over the short term you can fool yourself into thinking that it is one. Assuming tomorrow is like yesterday, turning down the graphics in unreal tournament is worth the same amount of money, but once HL2 comes out the price suddenly jumps, and they'll constantly be coming out with more new software that is more and more taxing on the system. The same can also be said for some of the prices, since the "next best thing" you keep telling people not to wait for tends to push prices down suddenly, but in between prices fall at a slower rate. The period of time you're looking at for video cards and processors is really farily stable, but that doesn't mean its like that all the time.

    I'll spare you the explanation for why I feel you're overcomplicating the issue, but suffice it to say that it shouldn't take that many pages and charts to explain when to buy. You might need the charts if you were trying for a definitive "buy X in Y weeks" article, but you're aiming for a general "this is what to consider when buying" article, and that can be done in a lot simpler words and with less graphs.

    So... what would I recommend instead? A more relaxed approach, but one that considers some of the same things as the article. I guess the heart of what I would say is "don't forget that having something now compared to later has value", it seems to be the one thing many people overlook. Aside form that, I can't really think of a mathematical model that would give an accurate depiction of the many variables, so I guess I'll leave it at that.
  • JarredWalton - Wednesday, February 02, 2005 - link

    For what it's worth, I don't know that a full-blown model of all potential upgrades would really be feasible or terribly useful, PrinceGaz. There are *SO* many factors to consider, and while ceratin tests will show a difference in performance, otheres might not change much at all.

    We really only looked at two of the major components in a computer, as they are often the bottleneck. RAM capacity is really the only other major factor. If we were to try to add in HDD, motherboards, PSUs, etc. then the model quickly becomes something that not even a mathlete would properly understand without staring for a while.

    As for the Quality to Price topic, for upgrading it becomes very difficult to model properly while including your present Quality to Price. If you assume your current system as 100% performance with 0 cost, you get a divide by zero error. In fact, any price for your current hardware other than its original price is going to skew a graph heavily in favor of not upgrading. Which leads to my take on the situation.

    The impetus for an upgrade has to be that you're unsatisfied with the current level of performance. If you're more or less happy, don't bother upgrading! Once you decide to upgrade, however, Forget about selling old hardware, forget about all the other stuff, and just pretend you're going to ditch what you have and buy something new. If you try to take all of the other variables into account, you again end up with a confusing model.

    If you want to be "fair" in the model, you can always take the price for all hardware and add the MSRP for your current hardware to it. So if you have a $100 9600 Pro, rather than saying it's "free" (and getting divide by zero), say it costs $100 but the 9800 Pro costs $300. Three times the price for maybe double the performance. The 25 cents per day CNU then changes as well, I think. If you want to play a game like Doom 3 and it runs poorly on your 9600 Pro, CNU is going to be more than 25 cents each day.

    My final comment (for now) is that the hypothetical system we were going to upgrade was chosen for a reason: we could forget motherboard, RAM, and many other components for an upgrade. A more realistic upgrade would have an older mobo, PSU, RAM, etc. and would need more than $200 to get a lot better performance. If you actually have an A64 2800+ with a 9600 Pro, you're probably going to be quite happy with it. :)
  • MadAd - Wednesday, February 02, 2005 - link

    #11 Hahahaha, right on - statistics sux huh :-) Reply
  • LordConrad - Wednesday, February 02, 2005 - link

    I think this article should have been called "The Mathletes Guide to Upgrading". As this article proves, it is certainly possible to overcomplicate things. For those of you (like me) who hate calculating and charting stuff, check out posts 31-32 for a much simpler way to accomplish the same tasks. Reply
  • PrinceGaz - Tuesday, February 01, 2005 - link

    Thanks for the explanation. Perhaps the article should be retitled "The Economic Guide to Building a New System", instead of the misleading "Economic Guide to Upgrading" as it most certainly isn't about upgrading.

    The only part of the article which even considers what you might upgrade from is where you set the quality of that at 100% and every option is relative to that. The graphs would be identical (except for the scale) if you just suggested people put the raw framerate from benchmarks in, as setting an arbitrary level to 100% (what you are upgrading from) doesn't affect the results at all.

    I was under the impression this was a guide to economic upgrading, and it could so easily have been if you'd deducted the quality and (optionally) the second-hand value of whatever you upgrade from. Upgrading is about replacing something old with something better, and the quality and price of an upgrade is therefore the quality of what you buy minus what you have, and the price is the cost of the upgrade minus what you can sell what you have for.

    "I actually modeled the "Quality - 100%" approach, and while the Quality to Price graphs changed in terms of numbers, the overall slopes were about the same.". Did you try that with the example I gave, or anything approaching a wide-range of upgrade options? The slopes might be similar but where they are on the graph are totally different. Your current sheet could very easily recommend someone to upgrade to something slower, and this is an article about upgrading! In fact I'm sure it would recommend a low-end Sempron as the best upgrade choice for someone with a fast Barton or Athlon 64. You really need to make it clear that the quality and value of what you have must be deducted from any upgrade option.

    I'm sorry to go on about this, but whilst it was a very interesting (if heavy going) article, it was so flawed from being "The AnandTech Guide to Economic Upgrading" that it really needs correcting. Either change what the article addresses (new system builds instead of upgrades), or correct it to reflect upgrading.
  • JarredWalton - Tuesday, February 01, 2005 - link

    38 - PrinceGaz, the current model does actually reflect more of a "what part should I purchase" mentality as opposed to an upgrade. So if you consider that the model is based off of that, the charts are still valid. The simple approach would be to take the spreadsheet and plug in your own quality, price, and depreciation values along with your own CNU to see how things look.

    There are a myriad number of ways to model the situation, and only the individual can readily determine how important an upgrade is to them. The idea behind the article is still sound, even if some of the graphs don't necessarily look right. I actually modeled the "Quality - 100%" approach, and while the Quality to Price graphs changed in terms of numbers, the overall slopes were about the same.
  • CrapONez - Monday, January 31, 2005 - link

    Good article that makes you think about cost/benefit rationalization to upgrades - for those who need it or simply are curious.

    The problem I had with this article was in determining the cost of not upgrading. So the game/encode/compile runs slowly on your computer. Where's the bottleneck? Processor? Amount of memory? Memory speed/bandwidth?

    Anandtech has plenty of reviews where various components are upgraded to determine the effect of each upgrade on total performance. We don't have that luxury. Determining which component to upgrade is often more difficult than selecting which model to upgrade to.
  • PrinceGaz - Monday, January 31, 2005 - link

    they put a space in front of cost_continuous.xls in the link on page 3
  • malikarshad - Monday, January 31, 2005 - link

    The link for excel worksheet is not working. Can somebody post a valid link Reply
  • PrinceGaz - Monday, January 31, 2005 - link

    Just to take the above example one step further, if the quality of card A in the above example was only 90% instead of 110%, so it is slower than what you already have; the graph in your sheet still shows it as the best choice.

    Spend $100 and get something slower :)
  • stephenbrooks - Monday, January 31, 2005 - link

    "What do we do to indecisive people who ask us when to upgrade?"

    "Confuse them with graphs!!"

    Sorry. It was a good idea. But as previous posters have said, what would be über-cool is to have the realtime pricing and entire benchmark database linked up to your formulae, and then let the user tweak the weighting factors on which things they find most important, and see what the site says. I guess it's a heck of a lot of work, though... maybe in 2010... :)
  • PrinceGaz - Monday, January 31, 2005 - link

    I'm afraid your calculations are fundamentally flawed from the point of view of upgrading.

    You assume that a product which is equal to what you already have (so it's not an upgrade at all) has a quality of 100%, and something twice as fast is 200%. That's fine if you are not upgrading but buying something totally new instead, but when upgrading you have to deduct the quality of what you already have from each of the potential upgrades, so subtract 100% from the quality as you already have that before upgrading. That would mean something that is the same as what you already have has a quality of 0% when considered as an upgrade, an upgrade twice as fast has a quality of 100%, three times as fast has a quality of 200%, and so on. Something half as fast would have a quality of -50% as it is not an upgrade.

    If selling your existing hardware when performing the upgrade, you should also deduct the amount you expect to sell it for from the cost of each potential upgrade option. The amount you can sell it for is likely to go down over time so that needs to be taken into account as well.

    To take a theoretical (but plausible) example and use the sheet you presented-

    Card A- $100, 110%
    Card B- $200, 170%
    Card C- $300, 240%
    Card D- $400, 260%

    the graph clearly shows card A is the best option, followed by B, then C, then D. But who in their right mind would spend $100 to upgrade to a card that is only 10% faster than what they already have?!

    Deduct 100% from the quality of each of those cards and the graph makes a lot more sense, with card C coming out on top, then D, then B, then A far behind the rest. Which is what you would expect as an upgrade to something 10% faster is a waste of money.

    Until the sheets and article are corrected, it is a very poor guide to upgrading.
  • KristopherKubicki - Monday, January 31, 2005 - link

    It should be Quality to Price - that will be fixed very soon.


  • CannonFodderjm - Monday, January 31, 2005 - link

    "Price to Quality" is best when high?!

    This confused me until the end, when I just gave up trying to understand your analyses and realized you made a "naming" mistake. It should be reversed.

    Great analysis, but this was too distracting.
    Please fix for the sake of others' sanity!
  • gimper48 - Monday, January 31, 2005 - link

    This was a great article but really leads to analysis paralysis. I am happy you guys do this for us. We really really appreciate it especially those of us who forget to carry the zero.

  • MarkM - Monday, January 31, 2005 - link

    #32, "I don't see those charts and formulas changing this all that much. You can tell which group you're in by checking your needs and your bank account," with all due respect, that was exactly the issue that the article so thouroughly addressed - for people who fit into ANY of your 3 categories, to identify the place in which to most effectively apply resources to address the perceived problem. You are the exact kind of person who could use a methodology like this, the person who's computer is slower than they want and/or need to do some specific task(s), but whose current approach to addressing a quantifiable need is nothing more rigourous than "Look at the Price Guide for the hardware you want [ed: I thought we were addressing a NEED, not a WANT?] to upgrade. Look at the components from lowest performing and go up from there. When you see the big price jump stop" Where in any of this methodology do we find ANY attempt to answer the question "will this upgrade resolve my slowdown"??? Reply
  • guest - Monday, January 31, 2005 - link

    Maybe now should be a good time to post an article about when to stop buying hardware :)
    In some cases it's better I think not to upgrade at all.
    Like when you don't buy anymore games or just do the occasional OS upgrade or just browse the internet.
  • LordConrad - Monday, January 31, 2005 - link

    People who upgrade (or buy a whole new computer) fall into one of three categories, either by choice or due to financial concerns:

    1. People who upgrade immediately when their computer starts to get a little slow.

    2. People who wait until the slowdown gets annoying.

    3. People who wait until their computer laughs at them when they try to run a program.

    I don't see those charts and formulas changing this all that much. You can tell which group you're in by checking your needs and your bank account. Why over-complicate things.
  • LordConrad - Monday, January 31, 2005 - link

    What the heck was all that crap? Upgrading is much easier than that and has two steps:

    1. Keeping your performance needs in mind, find the bottlenecks that are keeping you from reaching that performance.

    2. Replace the component(s) that are causing the biggest bottlenecks, while staying within your price range.

    Choosing Price/Performance:

    Look at the Price Guide for the hardware you want to upgrade. Look at the components from lowest performing and go up from there. When you see the big price jump stop. The item just before the big price jump is usually the best as far as Price/Performance.
  • Jedi2155 - Monday, January 31, 2005 - link

    Anand should really have a VGA and a CPU chart of some sort as well! I don't really trust tomshardware all that much.
    But then, most of us don't anyways :P.

    I like the article and its calculations tho...believe it or not....i use similar ones....
  • kmmatney - Monday, January 31, 2005 - link

    sorry about the third post - but there is one statement that I think hits the nail on the head:

    "Determining the exact performance increase from a Radeon 9600 Pro to a Radeon X850XT is not something that you'll readily see published on AnandTech or anywhere else."

    We can "almost" get this from Toms Hardware, from the VGA charts.

    They don't have half-life 2, but there are enough benchmarks to at least allow for an informed decision. In fact, I used the article to help a friend buy a Radeon 9600 Pro, which was the best bang-for-buck at the time.

    Why can't Anandtech have something similar, or the ability to create a similar graph on demand? Having the ability to chart both video card and cpu upgrades would be the ultimate tool for the upgrader.

  • kmmatney - Monday, January 31, 2005 - link

    This article reminds me of something I always wanted. The ability to pull up benchmarks dynamically. I could put in my system specs, or select a similar procesor from a list. Untested speeds can be estimated to first order. Then I could pull up benchmarks from different video cards and compare exactly what I want.

    For now, I use the Toms Hardware video charts, as Anandtech doesn't have anything so comprehensive. A "Real-Time benchmark Engine" would be awesome for this.
  • kmmatney - Monday, January 31, 2005 - link

    I'm a bit qith #19. It pains me to pay more than ~$100 for a video card. My plan was to stay with my Ti4200, but I managed to Ebay off $120 in old hardware last week, which brings my 6600GT upgrade down to an affordable $90, minus whatever nmy old card sells for. I'm definately sticking with my Athlon XP for a while, though.

  • roostercrows - Monday, January 31, 2005 - link

    excellent article kristopher, as i have had all the components for my new anandtech high end computer (except the video cards) just sitting on my bench for three months now just waiting... i can't help but think about what a price gouging graph would look like for the 6800gt pci-e sli video cards...... Reply
  • Googer - Monday, January 31, 2005 - link

    #13, I agree; but for some of us doing cost benefit analysis is part of the fun too. I really enjoy doing my mathmatics when I have the opertunity to do so. For me It is all part of the research that goes in to learning about new hardware. Its a hobby for me too. Reply
  • JarredWalton - Monday, January 31, 2005 - link

    The point of this articles wasn't to give everyone a specific decision. Some people are always waiting for the ideal time to upgrade. If you're happy with your current computer, then by no means do you need to upgrade. If you're disappointed with performance, however, and you start thinking something like, "I'll wait for SLI..." or "X800XL is coming soon..." or soemthing similar, is it really worth the wait?

    That's the whole point behind the "cost to not upgrade". How much are you willing to spend per day for a better computing experience? We arbitratily chose 25 cents, but suppose you're willing to spend 50 cents or a dollar? Given the model in the spreadsheets, you can plug in various prices and estimated depreciatioin rates, and suddenly it may not look so important to wait for the X800XL or whatever.

    The article certainly wasn't meant as a joke. It was intended as some reasoning behind our price guides. We don't often suggest people wait for the Next Big Thing, and the models presented here help to show why. We didn't even touch on the topic of market segmentation, which will often keep high-end parts at a higher price despite decreased production costs (witness SCSI, for example). There were plenty of other concepts that could also be applied. Hope that helps.

    (I helped quite a bit with the article, so I feel I can safely put in my two cents.)
  • roostercrows - Monday, January 31, 2005 - link

  • Poser - Monday, January 31, 2005 - link

    Thinking about it a bit more, I guess my situation (#19) could be wedged into the model you built pretty easily -- I don't have a Cost to Not Upgrade (CNU) of $0.25 a day, instead I have a NEGATIVE CNU. The longer I wait, the more cool gotta have 'em games will be released which I can't play on my old rig. That'll push up the value of current gen hardware. Reply
  • Chuckles - Sunday, January 30, 2005 - link

    It looks like an interesting model. With some tweaking it looks like it may be valid for a variety of market segments. Cool. Reply
  • arud - Sunday, January 30, 2005 - link

  • Poser - Sunday, January 30, 2005 - link

    I've been thinking about upgrading my computer for the past 6 months or more. Gutting it, really, since once the motherboard goes, most of the other components will get upped too to prevent dumb bottlenecks -- I'm looking at around six to seven hundred dollars worth of upgrades. But, the thing that's held me back ISN'T waiting for the next big thing, or for prices to drop, it's that upgrading to Half-Life 2 or Doom 3 grade hardware is worth AT MOST $150.

    I love this site, I consider computer hardware to be a genuine hobby, but I can't justify to myself spending more than that on playing FPS video games. The price of a good PC gaming rig is so completely out of line with what it'd cost to just pick up an Xbox that I suspect I'll be sticking with strategy games for a very long time... that or buying a current or next gen console.

    Eventually, I might find some "killer app" that is actually hardware-intensive (usably good speech recognition software with excellent OS integration?), but for the moment the only thing I do that challenges even my old 1400+ Athlon XP is gaming. I just can't bring myself to think that gaming on a PC is valuable enough to justify dropping the money.

    This article was a cool read, because if nothing else, it made me think to put a number on how much I really would "value" or pay for better hardware.
  • Dragonbate - Sunday, January 30, 2005 - link

    LOL I can't help but think this article was a farce. Reply
  • cosmotic - Sunday, January 30, 2005 - link

    Next time maybe you should tell us what we should do. Like "If this is your setup, the average person would upgrade HERE" and give what you would upgrade to. This is like trying to sell something to some one but then never actually asking them if they want to buy it. You have all these details and then no real conclusion. When SHOULD I upgrade? I have no idea, and it's not worth my time to read all this stuff and then figure it all out. Again, a nice conclusion with a concrete example would be nice. And some else that would be nice would be like arrows on the graphs that say "this is when you should upgrade and for reason X, Y, and Z". The graphs mean nothing without an explination or point. Reply
  • Dranzerk - Sunday, January 30, 2005 - link

    I think the single hardest part of a PC is upgrading. If we did not have PC games how many here would be running the latest hardware? I would upgrade once every 2 years, instead of buying new hardware little at a time every month to make a new pc every 6 months. lol

  • gaidin123 - Sunday, January 30, 2005 - link

    Great article! Granted most people won't actually do the formulas but this is a great article to link to when people moan about waiting for the next big thing. ;)

    Of course if you *need* the next big thing for the purpose you will use the computer for (ie SATAII or 802.11n) you have to wait...

  • archcommus87 - Sunday, January 30, 2005 - link

    But how is this reliable? The quality percentages are for one application only and even then are very estimated. And the cost per day of one quarter of NOT upgrading can vary greatly. If I'm gaming fine just now I'm not losing out on anything by not upgrading yet. Reply
  • MarkM - Sunday, January 30, 2005 - link

    Also, I think I might add, this is a hobby for most people, not a business. The whole point is to have fun, and sometimes the excitement of researching the new hardware is the best part. A cost/benefit analysis reduces the biggest benefit for some, the fun. Reply
  • MarkM - Sunday, January 30, 2005 - link

    Uhh ... that was interesting. Man, I'm an ANALYST for my career, I write cost/benefit analyses all the time, and even I was skimming by the last few pages of that!

    The one variable you didn't figure in (I think?) was the evaluator's time. Spending hours of your time calculating whehter it's workth it to spend the extra $50 may not be cost effective, in the gneral sense of resource cost. One thing I learned very early in my career is that there is a cost/beniefit ratio even in preparing the cost/benefit. If it is a relatively minor outlay, you need to apply heuristics over full blown analysis.

    Still, I think this is perhaps a good intro to peopel not used to thinking in this way.
  • deathwalker - Sunday, January 30, 2005 - link

    Ah...the benifits of being an impulse buyer. I don't have to worry about stuff like this. If you want it...get it...trash the formulas. Reply
  • Googer - Sunday, January 30, 2005 - link

    Just kidding, a very nice article. Reply
  • Googer - Sunday, January 30, 2005 - link

    they just renamed the site Reply
  • benk - Sunday, January 30, 2005 - link

    There is a superfluous space which breaks the link for the excel sheet on page 3. Reply
  • archcommus87 - Sunday, January 30, 2005 - link

    I'm amazed that this can actually apply too much in the real world. Seems like too many numbers are fudged/made up, such as cost per day, or the quality percentages. Reply
  • faboloso112 - Sunday, January 30, 2005 - link

    it would certainly be nice to have a plug-n-chug sorta setup so we can download the app and run it and figure out what should be our next upgrade...but nonetheless...this is a very solid article...and plus im sure that making such a prog wont exactly be the easiest thing in the world. but still...if you did make such a prog you'd be helping thousands of people make a good choice when upgrading. and if not you Kristopher...maybe Anand or even an AT member could perhaps take up this little project?

    once again...great article...keep up the great work!!
  • xsilver - Sunday, January 30, 2005 - link

    right on #4 -- some noobs here may read that and go "what's a model? is it much like a supermodel?" :D Reply
  • Postoasted - Sunday, January 30, 2005 - link

    The article reminded me of an econometric class I sat in on in college. For me, a click-thru decision menu would be more practical. Generally speaking, the menu would start with one's present computing environment and uses with time/cost analysis calculated for each performance unit-increase in efficiency. Reply
  • ChrisChiasson - Sunday, January 30, 2005 - link

    Some of your graphs are labeled "Price to Quality", where the GeForce 6600 has the highest series range on the entire time domain. The graphs should be labeled "Quality to Price", if the GeForce is the best upgrade.

    1 to 2 = 0.5
    2 to 1 = 2.0

  • zetto - Sunday, January 30, 2005 - link

    all these stuff make my head hurt.... Give us a calculator please :D Reply
  • cryptonomicon - Sunday, January 30, 2005 - link

    Thanks for doing this for me in the "Buyer's Guide" so I don't have to. Reply

Log in

Don't have an account? Sign up now