Price-Performance-Analysis for CPUs & GPUs

[dracoonpit]

Hi!

I like to give system builders a help, when it comes to buy or upgrads the system regarding CPU or graphic card(s).

How do CPUs and GPUs rank by performance? What about price-performance-ratio? Which GPU needs the most juice, which one is best while writing an email? This and much more is shown in several charts you will find by clicking one of the links below:

GPUs (post #1 German, post #3 newegg/English)
CPUs (chart #1: sort by performance, chart #2: sort by price-performance-ratio, not yet newegged)
DDR2-RAM | DDR3-RAM (chart #1: Performance default clocks, chart #2: Performance OC, chart #3: highest clocks, chart #4: price-performance-ratio, not yet newegged)

If you have questions concerning the charts or how the data is managed and how we come to those conclusions, I am glad to answer and to translate the relevant information for the non-German people here

Best regards,

dracoonpit

[dracoonpit]

Absolutely no comments on this?

[dracoonpit]

Another quick update: GTX 260 (SLI) and HD 4870 (CF) now included. Nvidia remains performance king, but comparing prices ..

For example, the 4870 is as fast as the GTX 260, but costs about 50 euros less than the GTX 260 (in Germany) which makes it 25% cheaper than a GTX 260. Another example: An 4870 CF setup matches a single GTX 260 comparing price performance ratio.

[Paulieg]

You may not be getting much of a response since it isn't in english.

[Kursah]

It's definately interesting, I personally like W1Z's take on the same idea for the GPU end of the deal, it's quicker to read and understand, and gives a great point also. But you have a pretty good graph setup also, I just think it needs a baseline of what would be considered you get 100% of what you pay for, then have above and below that, not necessarily in-line with model numbers and such. Keep upthe great work man.

Also welcome to TPU! I look forward to seeing your project grow, I just hope you're willing to dedicate enough time to keep it updated for a while...projects like that need constant attention and updating, if you're going international markets you may have to do different graphs for different regions and such to keep accuracy and flaming down a tad, I don't know what your full project plans are or how many are involved, but it sure looks promising.

[dracoonpit]

Thanks for feedback and the friendly words. At least all of you can compare the performances of the cards with the help of the charts - this information is region-free

The threads are older than a year and are updated every month. At the beginning, we started with a baseline (took the lowest performance card as 100%) as you suggested. The problem: You can not compare cards (or CPUs also) among each other, only to the baseline you choose, in our case, the card with the lowest performance.

Thus we are now using an absolute number: The average FPS of a card gathered from 11 different Game benchmarks.

Quick example:
Average FPS HD3870: 37
Average FPS HD4850: 57
Average FPS HD4870: 70

Using your calcualtor you can do exact statements regarding differences in performance at a given resolution and quality-setting of course. The baseline is your "point of view".

Base HD3870 (100%): HD4850 54% faster - HD4870 89% faster.
Base HD4850 (100%): HD3870 35% slower - HD4870 23% faster.
Base HD4870 (100%): HD3870 47% slower - HD4850 19% slower.

Depending on which card you own or plan to upgrade, you could calculate the performance gains (or losses) which of course vary depending on the cards compared. With this information, you can pick you local prices and calculate, if performance gain meets money to spend. The price performance analysis, like we do it (best bang for the bucks) is another story and of course, prices vary from country to country.

I may do those charts for other regions too - but I do not know good foreign online price comparison sites. It would be very kind, if could show me a few.

[dracoonpit]

Originally Posted by dracoonpit

I may do those charts for other regions too - but I do not know good foreign online price comparison sites. It would be very kind, if could show me a few.

Is pricegrabber.com the deal?

[Silverel]

use newegg.com for price comparisons. Most people in the US use them.

[dracoonpit]

Hi,

Thanks. I will do an update next time also in English for the States with prices grabbed from newegg.

Other suggestions for pricegrabbers around the world?

[dracoonpit]

Hi again.

During the last update I wanted to do charts for you people here with prices grabbed from newegg.com. But: What is this "mail in rebate" about? I don't know this from Germany, it does not exist here. So everybody needs to send in such a form and then gets this low price? So what is the price I should do my calculations on?

Thx in advance for reply.

[sneekypeet]

personally I would go with the non mail in rebate pricing.

The way it works is you pay full price, fill out a form and send in the UPC from the box and HOPEFULLY you get your money back. Ususally most people dont get the MIR.

[dracoonpit]

Ok, great. That helped I will inform you as soon as I have the charts ready.

[dracoonpit]

And now they are! Please scroll down to post #3 for the English version. Please leave some feedback, comments, critics, whatever

[dracoonpit]

Hi again,

the charts for graphics cards have been updated. There are now two charts, since our data-source has now updated the testing rig and benchmarks used.

CPU-Update will follow in a few days.

[beesagtig]

For the Australian market could you please use http://www.pccasegear.com/

Thanks

Very interesting graphs

[DarkMatter]

I don't know how the average was calculated, but I have to suggest you use baselines per each game before you calculate the average. Otherwise games with very high frames are almost the only ones contributing to the average. Example:

Game 1: card 1 = 200 fps, card 2 = 160 fps, difference 20%

Game 2: card 1 = 40 fps, card 2 = 50 fps, difference -20% or 20% in card 2's favor.

Simple average:

Card 1 = 120 fps
Card 2 = 105 fps

Difference 15% in card 1's favor. Clearly they should be tied, but the end result shows a clear advantage for the card 1. Introducing more samples won't change that issue, so the average performance is not what it should be, if that is the way the average was calculated. And judging the results and what you said in post #6 I guess that's the way you did it.

For example, that's why the HD4870 X2 is so ahead of the GTX280 in performance in your charts, when in reality is not so much faster. I don't have the time to look at every card, but I'm sure the issue is on all of them.

I will gladly vote this as VERY useful, when that changes and I trully think at heart it really is useful. But as it is now I don't think so.

[dracoonpit]

Charts updated again. As you suggested, I used your feedback for the new cpu charts. newwegg/australian prices not yet updated - sorry.

[intel igent]

Originally Posted by Paulieg

You may not be getting much of a response since it isn't in english.

google translator work's pretty good most of the time