Quantcast

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

I need to upgrade! Advised is greatly appricated!

2»

Comments

  • swing848swing848 Member UncommonPosts: 292

    If you do not have a high resolution monitor you do not need 1GB of video RAM.

    Intel Core i7 7700K, MB is Gigabyte Z270X-UD5
    SSD x2, 4TB WD Black HHD, 32GB RAM, MSI GTX 980 Ti Lightning LE video card

  • LadyRenoLadyReno Member Posts: 218

    LOL I have a 20 inch old dell monitor  from when they first came out with the flat screens Model # Dell E2047WFP  I seriously doubt it is a high resolution monitor.

     

    Ladyreno

  • QuizzicalQuizzical Member LegendaryPosts: 22,129

    Originally posted by rarraf

    But at the end of the day Nvidia drivers are 10x better than ATI and if I had not been drawn in by all the junk that people get much better performance with ATI.  I would not have bought this asus 5870 which I am not at all happy with my GTX260 run better in some games.  Times like thease people tend to go with the cheap option but I can tell you I wish I put up a bit more cash and went for GTX460

    Here comes the FUD, I see.  I think you're confused.

    A Radeon HD 5870 is much, much faster than a GeForce GTX 260, and on average, gives a little less than double the video performance.  A Radeon HD 5870 is also a lot faster than a GeForce GTX 460 at nearly everything.  It's more expensive than a GTX 460, as well, but the performance is in line with the price tag.  But it really doesn't make sense to say you wish you had spent more for a GTX 460 if you instead bought a 5870.

    "I'll wait until Christmas for purchase and start putting those penny's away now."

    Why Christmas?  If you're going to wait until then, you might as well wait for mid-January for the launch of Sandy Bridge.

    And if you're going to wait until Christmas to buy a mostly new system, why can't you wait a couple more weeks for the launch of Northern Islands?  There's a decent chance that Nvidia will be forced to slash prices on their existing cards to compete.  I guess there's also a chance that they'll simply figure that they can't compete on price without losing money on every card they sell, so if the only people who will buy their cards are the fanboys, they might as well raise prices and at least make decent money off of the fanboys.  It really depends on how far they can cut prices without taking a loss on each card even ignoring the fixed costs.

    "is there a really big difference between the 768 mb and the 1 gig? Or is just a slight difference. "

    They're two different cards with the same model number, though they are based on the same GPU.  It's kind of like how the GTX 470 and 480 are two different cards based on the same GPU.  A more honest company would have given them different names.

    You know how the amount of memory should always, always, always be a power of two times the number of memory channels?  The ratio of 1 GB to 0.75 GB is not a power of two.  The 1 GB version has 4 memory channels, while the 768 MB version has 3.  This means 1/3 more memory bandwidth, in addition to the extra memory capacity.  It also means 1/3 more ROPs.  Performance varies by game, but if the extra memory capacity doesn't make a bit of difference, then on average, the 1 GB version performs about 10% better.  And if the extra memory capacity does matter, it can get ugly for the 768 MB version.

    Also, why Sparkle?  For the 1 GB version, it might well be the cheapest, but for the 768 MB version, this is cheaper:

    http://www.newegg.com/Product/Product.aspx?Item=N82E16814500173

  • LadyRenoLadyReno Member Posts: 218

    Ok thanks for the clerification.  Just getting an idea now of what is out there, like I said I haven't checked into any upgrades for a while.  As for sparkle, life time warrenty main reason.  But other than that I'm flexable with the the brand. 

     

    Ladyreno

  • QuizzicalQuizzical Member LegendaryPosts: 22,129

    Originally posted by LadyReno

    As for sparkle, life time warrenty main reason.

    You'd better read the fine print on Sparkle's "lifetime" warranty.

    "Lifetime replacement warranty does not cover items out of production if the company no longer stocks them. (Lifetime is defined as the lifetime of the product on the market. Outdated technology is not covered by lifetime warranty if the item is no longer available on the common market as a new product.)"

    http://www.sparkle-usa.com/warranty.asp?seq=3

    Sometimes Sparkle's "lifetime" warranty doesn't even last a year.  If you're wondering what a real "lifetime" warranty means on a discontinued card, basically they give you a newer card with roughly equivalent performance.  For example, someone with a Radeon HD 2900 XT that dies might get a Radeon HD 5670 as a replacement.

    If you want a real lifetime warranty that means what you think it means, there are companies that offer them.  EVGA, XFX, and BFG used to be the big three board partners for Nvidia that enthusiasts who wanted a lifetime warranty went for.  Then XFX started producing AMD cards, BFG went bankrupt because Nvidia didn't have much to offer and AMD wouldn't let them produce AMD cards, and now XFX is stopping on producing Nvidia cards (and filling warranties for dead Nvidia cards with new AMD cards).  That just leaves EVGA.

    Though I don't think a lifetime warranty is really all that important.  If a card dies three years from now, do you really want to replace it with something of equivalent performance, or with something better?

  • swing848swing848 Member UncommonPosts: 292

    Quizzical,

    She needs to understand that screen resolution makes a difference.  There are far fewer pixels to push around on a low resolution screen as there are on a high resolution monitor.  Your formula for how much local RAM is needed on a video card does not include this.

    I selected the fastest of each class that Tom reviewed.  The overclocked HD 5870 with 1GB memory beat the HD 5870 with 2GB of memory.  The fastest HD 5850 and GTX 460 are also on the list.  The OP probably does not know that all numbers are frames per second except for 3DMark, heat, and noise levels.

    Benchmark Scores:

    http://www.tomshardware.com/charts/2010-gaming-graphics-charts-high-quality/compare,2487.html?prod%5B4551%5D=on&prod%5B4549%5D=on&prod%5B4560%5D=on&prod%5B4511%5D=on

    If LadyReno

    As are you, I am not a fan of Sparkle.  And, I agree with you that she should wait just a little longer to see what the HD 6000 series brings to the table.

    If someone is an nVidea fan I do not try to diswade them from making a purchase, however, in cases such as LadyReno I believe at least a little information/education is important.  I suggest AMD video cards over nVidea most of the time.  Although the GTX 460 can be cooled properly some of the cooling solutions on cards at NewEgg.com are questionable unless the computer case is well cooled.  An example is the video card you suggested because it has a video out in the way of the exhaust port.  However, between the price range of $170 to $190 the GTX 460 is not a bad card at medium screen resolution, although it is a little weak with BF2.

    I had a couple of other points, but, I have to leave again.  If I have made some typos please excuse me as I have to leave now.

    Intel Core i7 7700K, MB is Gigabyte Z270X-UD5
    SSD x2, 4TB WD Black HHD, 32GB RAM, MSI GTX 980 Ti Lightning LE video card

  • LadyRenoLadyReno Member Posts: 218

    Ok thanks guys for all the advise, I will check into the amd as well.  I'll start another thread when I am ready for purchase.

    I just checked your chart and boy the 460 was at the bottom of everything, so guess I'll check out the amd as well. 

    Thanks a bunch

    Ladyreno

  • NudlesNudles Member UncommonPosts: 54

    Just some last words with advises before purchasing new pc components (and this apply to many other things outside the computer world).



    Define your budget and use of the PC.

    For a wiser purchase, research existing technology offers (ex:AMD/ATi/Nvidia/Intel) using different sources of information (ex:website reviewers, forums..) and crossmatch info gathered to draw your own conclusions. Computer performance values reviewed are not definitive since many things can affect these like different hardware setups, drivers, revisions, OS, games, etc... and the end user may get different results.



    It's also a good idea to research technology roadmaps to see what's coming next and when.

    Sometimes, waiting for a new chipset (like motherboard/cpu/gpu) is worthwhile since it gives you the chance to upgrade into fresh technologies which can be supported for a longer period of time or buy cheaper previous tech. It all depends on your budget, usage and urgency of use.



    For a budget conscious buyer to get more accurate results is to use similar technologies at around same price bracket using a similar setup formula. This, disregarding any unjustified biased preferences (read: ATi/Nvidia/Intel/AMD hype flamewars).



    To illustrate what i mean, take this chart as an example: http://www.tweaktown.com/articles/3460/gainward_geforce_gtx_460_1gb_goes_like_hell_glh_edition_in_sli_overclocked/index19.html

    When i built my system in August i followed this kind of mindset according to my needs and budget. There are many good websites which review hardware decently like http://www.tweaktown.com, http://www.anandtech.com, http://www.guru3d.com and so on. I particularly like http://www.bit-tech.net because of their testing methods and tech analysis but i also use other sites. Google is your friend.



    I previously recommended the Nvidia GTX460 1GB and the Radeon HD5850 1GB but using the above formula, i should have suggested the Radeon HD5830 instead.

    According to overall prices practised in my country (it depends on offerings price in your country):



    HD5830 slightly cheaper than GTX460

    HD5850 slightly cheaper than GTX470



    You can also take into account SLI/XFire setups to match or exceed more expensive single card setups while spending the same or less but there are other things to take into account like motherboard support, power supply, cooling, etc...



    Regardless of your brand decisions in the end, i personally think both AMD/Ati and Nvidia offer good gaming products for different budgets.

    Just set DX11 support and 1GB GDDR5 (more memory only really noticeable with multiple monitors, higher resolutions than 1920x+ and some professional programs atm) as base requirements and work out the rest.

     

  • QuizzicalQuizzical Member LegendaryPosts: 22,129

    There are some huge problems with that chart.  One is that it's mixing overclocked cards with cards running at stock speeds.  Overclocks on review sites are not representative of what you can expect to get yourself.  Video card companies that know that a site will overclock their card can bin chips to pick out one that will overclock further than most, and send the site that particular card so that they'll be impressed with the huge overclock.

    Second, prices change.  The chart says that a Radeon HD 5870 for $500 isn't a good value.  But what about a Radeon HD 5870 for $340?  The chart doesn't say, because it doesn't get updated.

    Third, mixing SLI/CrossFire results with single GPU results is a big no-no.  Average frame rates in big name games with heavily tuned drivers aren't representative of real-world usage.  Sometimes SLI and CrossFire simply don't scale at all, or at least not nearly as well as they ought to.  But AMD and Nvidia know which games tend to get used by review sites, and can work very hard to make SLI and CrossFire scale well in those particular games.  There are also issues with the frame rate being more erratic (e.g., microstutter), which overestimates how good frame rates are if you only go by the averages.

    Fourth, sites have a lot of ways to skew ratings if so inclined.  A common one is using different image quality settings on different cards.  A card running at higher image quality settings tends to give lower frame rates, but that doesn't mean it's a worse card.  Some sites have tested games with GPU PhysX and then averaged that in with actual video card results.

    Another common one is exploiting the arithmetic mean-harmonic mean inequality to skew things whichever way you want.  This one is subtle enough that it needs an explanation.  Suppose that your review suite consists of two games, and you're comparing two cards.  The first card is twice as fast as the second card in game A, and the second card is twice as fast as the first card in game B.  Which card is faster overall?  The obvious answer is that they're tied.

    Suppose that a site decides to scale such that the first card is 100%, and the second card is whatever percentage, and then average the percentages.  The second card gives 200% of the performance in the first game, and 50% of the performance in the second game.  Average this to get that the second card gives 125% of the performance of the first card on average.  If you use the second card as the baseline, you can likewise get that the first card gives 125% of the performance of the second card on average.  Which card do you set as the baseline?  The one you don't like.

    If you're going to advise particular review sites, you might want to at least mention Hard OCP, which seems to be the only review site that tries actually playing games on the video cards as part of the review process, rather than only running synthetic benchmarks.

  • NudlesNudles Member UncommonPosts: 54



    The chart was put there as an illustrative example for what i've explained previously about researching different sources of reviews and whatnot. I personally gathered the information i needed from many other sites to build my own chart but the concept was still valid : to build my own budget selection of hardware according to my needs.



    I found the chart in August which coincides with the date i was building my system, which in turn coincides with overal prices practiced then. I think it's pretty obvious to account for today's prices when building a system. And again, i revert to my explanation about using the chart as an illustrative example according to budget/performance - an actual buyer should adapt same concept to current days offerings.



    Personally, i like to have OC'ed and SLI/Xfire results included with stock/single card test results since it gives me an idea about GPU scalling which could be usefull if i'm inclined to follow that path.And yes, multiple cards can give erratic results just like Oc'ing can give erratic results if not fine tuned. In the end, it's down to the buyers discretion to decide what to do or not do but having the information available doesn't hurt in my book.



    Again, for the rest of the post, i revert to my explanation about having different sources of information to draw your own conclusions. I gave some examples of reviewing sites but also pointed out other sites reached through the "over-complicated" internet search engine...  Google - it's free.

     

    If i failed to mention one you like, then i apologise... Feel free to add Hard OCP to the list.



    About this whole post, i tried to be the clearest possible with my "advices" on the previous post and now i spent most of the time hinting about reading/understanding skills. I guess my written english is not good enough, therefore again... i apologise for it...



    Nevertheless, i hope the OP and other interested people grabbed the message i was trying to pass.

  • swing848swing848 Member UncommonPosts: 292

    The reason I have never liked TweakTown's Value charts is simple, it can give uneducated people the impression that a GTX 460 is faster or better than a GTX 480, or any other faster card, because the graph shows a higher number for the GTX 460.  The biggest problem is an HD 4650 at $34 might win on a value chart.  How important is that and what does it mean?  I know this sounds rather simplistic [a better word than uneducated, or worse, stupid] however, I have seen this asked about from people on different forums.

    I understand the chart, however, I disagree with its use and always have. 

    I am glad that you were able to research and produce the computer that you wanted.

    Most people can find reviews, look at frames per second at or near the screen resolution they will use, with the same driver numbers as compared to different reviews, with the same CPU as different reviews, with the same motherboard and RAM speed, and so on, in an effort to find an exact comparison.  Good luck with that.

    It is better to find a few sites that are well known for giving good information and go with that.  Another problem I have with many review sites is, they will get a card soon after release or even before release and use an early or beta driver, but never come back in six or seven months with a follow up review that shows how well drivers have improved.

    The task of finding what someone wants is difficult and that is often why they find forums, such as this one, for help.  While help is available here, it is limited because this is a MMO gaming site, and in that regard there are lots of opinions.

    However, when it comes to hardware, opinions should mean nothing.  Cold hard facts are what count.  I was in the computer hardware field for many years and became dedicated to the study and experimentation of hardware specifically for gaming use and overclocking results and dangers; this was for my own satisfaction.  However, my mix of computer hardware for gaming and overclocking experience is a good mix when my years of study in the computer hardware field is the base that I built upon.

    That said, no one has the answer to everything, however, building a computer is very straight forward for an expert, yet very challenging to a novice.  And, I have been known to challenge and even scold well known experts in the computer field when they did something wrong in print or video, the reason is they are experts and should know better.  Often most people found on forums are not experts, especially on forums such as this one, and I normally do not scold them for their comments, because they are not nationally known experts.  However some people that are highly skilled in the computer hardware field can sometimes be found, even on forums such as this one.

    From time to time there will be disagreements between people with varing degrees of expertise, however, with simple and accurate information people searching for answers can often find them.  Finding the truth is the problem.  I try to be careful with my comments so as provide correct advice and not to bring remarks from others that, even though someone is trying to help, actually muddy the waters for many [novice to hardware] computer owners.

    Intel Core i7 7700K, MB is Gigabyte Z270X-UD5
    SSD x2, 4TB WD Black HHD, 32GB RAM, MSI GTX 980 Ti Lightning LE video card

Sign In or Register to comment.