System Builder Marathon, Sept. '09: $1,250 Enthusiast Build

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.


No problems with heat or noise demonstrated on the $1250 build. Dx11 won't be a factor for years.

More importantly though, why do you not acknowledge that this article was put into motion before the release of the Radeon 5xxx series?
 
two 5850's would make mince meat of this rig, saving money for a better cpu ( no extreme 4 x pci-e mobo needed then) and the whole would be better.
But since that is now and not a month ago this is a nice try of the beaten track to see if an alternative approach would work. It does although a bit limited.

I think a p55 based i5 build with 5850's 5870 or maybe even the 5870x2 /5850x2 would be the best bang you can get for 1300-1500 bucks.
 


No arguments here, I think a p55 with a couple Radeon 5000 series cards would be absolutely unbeatable for the price. I'll be very surprised if we don't go that way in the next SBM.
 
[citation][nom]Gulli[/nom]The difference will be small and four 4850's are actually more expensive than one 5870! Not to mention the extra costs for a bigger PSU and problems with heat, noise, games that don't do well with crossfire and lack of DX11...I could've gone with it if they chose two 4890's, but that horrible contraption they conceived here will never be recommended by me.[/citation]
people come to you for recommendations? who are these people? Barney Rubble?
the 5000 series wasnt available when they did the build so for the purposes of this particular review it's pointless to be wishing and hoping and having flashbacks. maybe you are right about the scaling but since you don't have the numbers to definitively state whether for instance 3 4890s are definitely faster than this arrangement aren't you kinda just talking out of your ass? like you could be right but it wouldnt be because of anyone definite information. it would be because there are only 2 choices, either you are wrong or you are right and you might have chosen the right side of the coin this time. show some actual numbers that backup what you say or go warm the bench.
 
[citation][nom]kaihekoa[/nom]$1265 budget and you choose FOUR Radeon 4850 1GB cards with poor scaling and 3GBs of wasted VRAM and a Phenom II X4? These choices are ridiculously irrational. This is supposed to be an enthusiast system, but performance should not be sacrificed at the cost of AMD favoritism or a "Woah, four video cards are cool!" effect. A system with an i7 and dual Radeon 4890s or GTX 275s overclocked is a much better option for this budget.[/citation]

The terms myopia and myopic (or the common terms short sightedness or short sighted) have also been used metaphorically to refer to cognitive thinking and decision making that is narrow sighted or lacking in concern for wider interests or longer-term consequences.

how many times has it been that this was an AMD build? reading isn't fundamental in your part of the world?
 
i have a 4850x2 - and i agree with an earlier poster - i think i would have gone with two 4850x2s over four individual 4850s.

i think sapphire blew the launch with bad driver support, but since then the generic ATI divers work fine - it has become one screaming card.

i'd use the savings and go for a asus crosshair iii - both slots are full x16.

good news is, any day now you should be able to et a 965BE for the same dollars as the 945 in the review.

great job don - it's gotta suck to be second-guessed by everyone like this - but i gain a lot from reading everyone's comments and putting them plus articles like this to use on my next builds.
 
[citation][nom]jellico[/nom]Well, people have been begging for an AMD build, and there it is. Soundly trounced by the Core i7 build, even with its crappy stock cooler. It is what it is and I don't think anyone here is really surprised by the outcome. Core i7 is the king of the hill... for now. I will, however, be very anxious to see a build based on the AMD 6-core processor (when they are released next year) and Radeon HD 5870 card.[/citation]

Yeah I'd love to see that trounced by the new Intel hexcore (already set the world 3DMark record) and a pair of GT300's.
 
[citation][nom]Cleeve[/nom]You don't 'have' to do anything, Kl2amer. But because you don't like it on principle, that doesn't mean the rest of us 'have' to dislike it, either.I like it on principle because it works, doesn't create too much heat, use too much power, or cost too much compared to other solutions.The number of graphics cards is irrelevant to me as long as it does the job.[/citation]

No worries Cleeve, I should have known by now that no one is allowed to have an opinion that is different than yours. Just a momentary lapse on my part. As i already stated I won't be offering my opinion again when you are involved in the article.

I did have a question though, if thats ok. What were the temps these cards were running at on load and what was the case temp? I see the CPU temps but seem to be missing the temps generated by the graphics cards. Thanks.
 


All opinions are welcome Kl2amer, just please don't suggest that your way of thinking is the only right one.



Is that facetiousness I smell? How becoming!

I believe all of the temps will be in the conclusion article.

 
Umm, no it was a legitmate question as Heat output is important to me. You referenced it twice in replies to my posts stating heat wasn't an issue and I wanted to know the temps 4 cards run at as well as the ensuing heat build up in the case you chose. Unfortunately I have re-read the conclusion and temp pages but only see the CPU temps vs. Delta. Am i missing something?
 


Judging from this I can only assume you're not familiar with the definition of facetiousness, or that you are unaware how parts of your post are quite appropriately described as facetious. If you're truly in the dark about this, you can PM me and I'll help you work it out.

As far as the temps, like I said - I believe all of the temps will be in the conclusion article ("will be" referring to future tense. SBM conclusion, not released yet). :)
 
[citation][nom]Cleeve[/nom]Judging from this I can only assume you're not familiar with the definition of facetiousness, or that you are unaware how parts of your post are quite appropriately described as facetious. If you're truly in the dark about this, you can PM me and I'll help you work it out.[/citation]

LOL, and your off and running yet again. I am familiar with the definition and no matter how i replied I absolutely knew this type of response was coming. You seem very defensive, I feel sorry for you since it seems you are hell bent on compensating for something. Oh, and yes I can see how you derived that from my post given your propensity for attacking everyone that dares to post something you don't like. Thanks for again showcasing your immaturity by taking comments about technology and turning it into "I googled a word you might not know, that will teach him".

Thanks for the clarification on the temps as I did miss understand what you meant by the conclusion.
 
really great article love the fact that you put four cards in there

i wonder if 3x 1gig 4870's would have done the job better

think about running folding at home on that rig
 


Heheh, I sure am. And there you are, right back again doing your part. What a sorry bunch we are. 😀




You're welcome, no problem.
 
"Umm, no it was a legitmate question as Heat output is important to me"



wouldn't the true heat output be very similar to the total wattage of the system.... oh and you forgot an "i" in "legitmate" the correct spelling is legitimate
 
Tom's should really hurry up and finish building their time machine...

That way they can buy today's stuff from the egg and use it in the article they were writing days or maybe even weeks before. Might halve the number of whiners around here.

Ofcourse, there's also the possibility they'd just whine about why Tom's didn't go further into the future and use the budget 1THz quantum chip which would pwn these ancient chips for half the price.
 
I agree in saying this is a piss poor review of amd hardware by tom's hardware. if you see the detail they put into the intel reviews....... they cut this one down.

bad choice of hardware....... i'm sure a higher phenom II x4 paired with a crossfire of 2 x 2GB 4890's would of made a better result then 4 smaller 4850's.

It's a bit of a rip on AMD really since tom's hardware compare a i7 setup and then choose mild hardware for the amd gear........ tisk tisk
 
What they did with the 4 Gigabyte 4850 cards was very ingenious in this build.It looks odd at first and then when one examines the overclocked graphics cards they chose,the power consumption,the cost and using them all in crossfire it turned out to be quite effective and efficient.
It even performed well compared to their earlier i7 build in gaming.
 
Yes. I still wouldn't do it myself, but it was well-explained and makes interesting fodder for the "index." I'd really like to see how it compares to a pair of 4890s; the further scaling tests are really needed to see where this build fits. How would three 4850s perform if the CPU were upgraded to a 955BE?
 


You're an idiot
 
I thought crossfire was not scaling well above two cards but these benches are impressive. The other interesting thing is that four 4850s can beat one 5870 (after reading your 5870 review). Finally I am not surprised that AMD beat Intel here. For gaming purposes there is not much difference between these cpus.
 
Reflecting back to some earlier articles on Tom's Hardware "Phenom II 955 Versus Core i7 920: Gaming Value Compared" here
(jtt283 might like to see this since it uses a Phenom II X4 955 BE and a pair of 4890's)

http://www.tomshardware.com/reviews/phenom-versus-i7,2360.html

and an earlier article here

http://www.tomshardware.com/reviews/cyberpower-gamer-dragon,2337.html

It's interesting how this build finally was able to get an AMD system nearly on par with an i7 build on gaming performance with the use of those 4 gigabyte 4850's.
 
[citation][nom]Cleeve[/nom]I find this surprising. A single 4870 isn't going to have any problem at all rendering 24 WoW characters. Which graphics settings, specifically? [/citation]
I bet he's using old ati drivers and has turned on all the shadows. After 3.2 patch shadow rendering on newer ati cards is broken (flickers or doesn't display at all, but eats up gpu power). In any event - wow doesn't require a huge amount og graphics power. It does need memory and decent cpu utilization though. Also the amount and quality of addons greatly affects ingame performance. I'm guessing he's just not doing stuff right! I run the game at 1680 in window mode with another monitor running facebook, winamp and all the other stuff on the side - and consequently ALWAYS hit the fps cap (60.1-60.7 per ingame measurement, ctrl+r). And taht with a puny c2d and a single 4870 512.

Anyway, as for the article - why don't you do any noise measurements? I remember when I got my 4870 I was feeling very happy about the noise level compared to a friends 4850 (some rediculously loud asus part). I'd like to know the noise difference between 4 ati cards and 2 nvidia :)
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom