System Builder Marathon, June 2010: $2,000 Performance PC

[Crashman]

[citation][nom]p4l1ndr0m3[/nom]Civility stopped when Crashman was replying to a4mula on the first page.[/citation]There you go, blaming someone else for your own bad behavior without even pointing to a specific example. In other words, making stuff up, or prevarication. I think someone should ban you simply to remove all the dishonest statements and unsubstantiated accusations you've been making.

I just went back to the first page and saw the post you appear to be referring to with some kind of reference to unsubstantiated claims, I guess you didn't actually look at the LINKS in that post. So I decided to help you out: Look for the "PROOF FOR p4l1ndr0m3" tags

 

[Userremoved]

[citation][nom]Crashman[/nom]There you go, blaming someone else for your own bad behavior without even using an example. In other words, making stuff up, or prevarication. I think someone should ban you simply to remove all the dishonest statements and unsubstantiated accusations you've been making.[/citation]
Next time just disable the comments and they wont whine.

 

[notty22]

In other words, making stuff up, or prevarication. I think someone should ban you simply to remove all the dishonest statements and unsubstantiated accusations you've been making.

Took the words right out of my mouth, just the trolling on this page alone, don't bother acknowledging him.

It's the internet
, grow a pair.

When was the last time a PSU review article was published on Tom's? Yeah, that's what I thought...

The case doesn't even have E-SATA... ???????

I'd hate to see someone with this machine run F@H on their GPU and CPU with any regularity.

 

[notty22]

Really Crashman thinks that JohnyGURU and Hardware Secrets are bad reviewers? Then who those Crahman think are good reviewers? Just wondering not trying to start a fight.

Not trying to start a argument, then why are you feeding a troll ?
What the hell do those web sites have to do , with this build ?

 

[Crashman]

[citation][nom]p4l1ndr0m3[/nom]You keep quoting that same phrase yet you seem to be in short supply of facts yourself. Are you suggesting that the "guru" website I mentioned does not present results based upon fact? Here I thought it was a consensus that those guys had the best load testing equipment and methodology of any PSU reviewers around. When was the last time a PSU review article was published on Tom's? Yeah, that's what I thought...[/citation]No, I'm merely suggesting that I have a bunch of equipment and a meter. When you try to argue against the meter, you get that kind of response from me. Easy no? Well, perhaps not so easy for someone looking for a fight.[citation][nom]p4l1ndr0m3[/nom]Banning people who disagree with the author in the comments section is a good way to attract new readers. You aren't obligated you reply, though it is more fun when you do. I have to agree with Userremoved, disable comments if you can't handle it. It's the internet, grow a pair.[/citation]You saw what I said, so quit lying. I didn't call you out for disagreeing, I called you out for making things up about what I said. I think that the old fashioned "bearing false witness against your neighbor" offense should be automatic qualification for banning.[citation][nom]notty22[/nom]Took the words right out of my mouth, just the trolling on this page alone, don't bother acknowledging him.[/citation]Sorry notty22, I already popped the cork on this one. Some people around here seem to know that misrepresenting what someone said ranks up there with giving backwards advice to confuse a newb on my list of hot buttons. And some of those people love pushing buttons.[citation][nom]notty22[/nom]Not trying to start a argument, then why are you feeding a troll ? What the hell do those web sites have to do , with this build ?[/citation]Now you get to see how "the sausage is made". The article took power readings directly from the system at peak load. p4l1ndr0m3's argument "against the power chart" was that some other sites had shown these graphics cards could exceed their stated TDP. That is to say, his actual argument would have made sense if the power number had been calculated using the stated TDP for all the parts. The reason his argument fell apart was that the power was actually measured, rather than calculated.

So, his retaliatory statement is that I'm arguing against a site that specializes in some of those parts. Userremoved appears not to be following that entire conversation (who could?) and p4l1ndr0m3's statement "pricks his ears".

 

[Userremoved]

I have to agree with Userremoved, disable comments if you can't handle it. It's the internet, grow a pair.

Well I disagree with you I said that because people like you are flaming Crashman.

 

[Userremoved]

Not trying to start a argument, then why are you feeding a troll ?
What the hell do those web sites have to do , with this build ?

good point I forgot not to feed the trolls.

Also Crashman a few days ago you said you would ask if a pair of 5870s would give any advantages over two GTX470s to one of the GPU gurus at Tom's Hardware US did he reply? If yes is it worth it?

PS: I'm guessing all the trolls above are only one person with multiple accounts.

 

[Userremoved]

So, his retaliatory statement is that I'm arguing against a site that specializes in some of those parts. Userremoved appears not to be following that entire conversation (who could?) and p4l1ndr0m3's statement "pricks his ears".

Oh my I'm so sorry it gets hard to follow every single discussion after a few pages.

 

[Crashman]

[citation][nom]Userremoved[/nom]Oh my I'm so sorry it gets hard to follow every single discussion after a few pages.[/citation]Hey, with all these different directions a person almost needs a flow chart to follow it!

BTW on the power supply, I think I said somewhere that it's barely adequate and I probably would have picked something else if I'd known the power numbers were going to be this bad, that's true. Consuming 858W to produce around 670W (according to the manufacturer's efficiency specs) looks really bad, and an 80+ Silver rated unit would have dropped power consumption to around 790W for that same 670W output. Looking for a great price on a solid part to boost the system's value ranking, this looks like a better choice: http://www.newegg.com/Product/Product.aspx?Item=N82E16817256058 I haven't been able to find an "in the middle" 800W unit of the same quality for a lower price.

 

[Relayer]

[citation][nom]Crashman[/nom]At super-peak 100% CPU + GPU load testing the power supply drew 858W from the wall, which according to the power supply manufacturer equals an output of roughly 670W. So what's the question?Are you trying to imply that the graphics cards are consuming more power than the power supply is outputting?[/citation]

Unlike nVidia claims, the 470 can draw substantially more than 225W. It can draw ~300W. My question is, was the use of a 750W PSU in a system w/ SLI 470s influenced by nVidia? I don't know why you were surprised by the amt. of power that it did draw. It's well established the power demands of the 470. After seeing the peak draw you should have gone with a bigger PSU. As has been pointed out, over time, if someone OC'd and did F@H with this system that PSU would be inadequate.

 

[Crashman]

[citation][nom]Relayer[/nom]My question is, was the use of a 750W PSU in a system w/ SLI 470s influenced by nVidia?[/citation]No, I never base any of my decisions on Nvidia recommendations. I don't believe in it. As a matter of fact, I've been at odds with Nvidia over its recommendations and refuse to even look at them. I base my parts selection on previous test data generated by professional review sites, usually Tom's Hardware. In this case, I contacted Don Woligroski.

I don't use Folding@home so I'm not certain how many weeks any specific power supply will last at 89% of its rated "continuous" load. I do know what the word continuous means however, and I'd love to see a company brought to the spotlight for fraudulent power output claims.

Having said all that, PCs aren't even designed for 100% load 24/7, that's what severs are for. I've been running my 4.0 GHz overclocked system 24/7 for over two years, but it only sees full CPU load for a few hours a day. It never sees full CPU and full GPU load simultaneously. I think that's a more realistic idea of what to expect from a highly-overclocked machine. If you want perfect reliability at 100% load for years on end, don't even bother overclocking. But wait, wouldn't "not overclocking" solve your questions about the longevity of the power supply?

 

[monteis]

[citation][nom]Crashman[/nom] Chis says SLI scales better which is why I had to broaden my estimate. Then again, his findings are based on the 480 http://www.tomshardware.com/review [...] ,2622.htmlAgain I refer to Chris, since he and Don are the graphics gurus: PROOF FOR p4l1ndr0m3-> http://www.tomshardware.com/review [...] ,2474.html[/citation]

you seem to be dodging the issue, he never said that 2 470's don't beat a 5970, what he said is that since, in your own words and in this very review you stated that 470's are about as powerful as 5850's, why not use 2 5850's since the pair is way cheaper.

i have to agree with him that 2 5850's are not the same as a single 5970, and i would think you would know this since, i got most of my information from tomshardware.com and from ya'lls reviews, so as to why you keep trying to use the 5970 as a reference for 2 5850's, i have no idea

 

[Crashman]

[citation][nom]monteis[/nom]you seem to be dodging the issue[/citation]At the outset I was trying to figure out whether to use dual 5850's or dual 5870's, so I contacted Don. His familiarity with an entire range of graphics hardware, graphics benchmarks, and test settings lead him to recommend dual 470s as the best bang-for-the-buck solution for our specific benchmark set. We all know Don's an ATI guy at heart and it probably pained him to recommend the Nvidia solution, but his integrity and professionalism always take precedence over his feelings about any particular company.

OK, so basically, Chris' test shows that the 5970 really is about as powerful as two 5870's, and that a 5850 is less powerful than a 5870. We see arguments to the contrary, but I'll take test results over opinion any day. When I don't have a complete set of test results for all the hardware I'm considering, I turn to Don.

Remember that our benchmark set is very specific, and that SLI scales better than CrossFire most of the time. I think that went into Don's recommendation.

 

[Relayer]

Just because a 470 and a 5850 are roughly the same performance doesn't mean that 470SLI and 5850Xfire will remain equivalents. ATM SLI scales better with Fermi than Xfire with Cypress.

That said, I'd still take the Xfire Cypress solution, probably dual 5870 in this situation. Then that 750W PSU would make sense, as well.

 

[hangfirew8]

[citation][nom]madass[/nom]$50 cases with side opened>hiogh end case with side closed.And no, I've never had problems with dust. And I own a dog. Go figure.[/citation]

You have no problems with dust because your airflow speed is too slow to pick up a hair and carry it all the way to your CPU/GPU coolers.

 

[hangfirew8]

The Antec 300 is a great case (I have three) and a great value for gaming and CPU and GPU overclocking... until you go SLI. The CPU will still be fine (it has a 120 and a 140mm pulling heat off of it and therefore gets most of the 2x120mm air coming in), but at that point you have only the power supply removing the heat of the bottom card and the top card cooking the bottom card. In the end you get a cooked power supply and a cooked video card or two. Putting a big fan on the side panel would help a lot... if you don't mind the noise.

In your current configuration, I hope you like the sound of the power supply and video card fans ramping up and staying fast. This kept temps stable but not low enough for stable massive overclocks.

The Megahelams is an awesome cooler. Great choice. The Hyper 212+, while a good value at its original price (good luck finding that anywhere in stock but a MicroCenter), can't hold a candle to the Megahelams. The lower mass 212+ performs well due to Direct Touch, until temps get high at which point it doesn't have the surface area to shed heat fast enough.

I understand the GTX470's but would have picked a Motherboard that left a little space in between them.

Don't care much about SSD versus disk, people will buy what they can afford or want and all will fit.

Until NVIDIA et al releases actual amps pulled instead of vague "use X watt power supplies), proper P/S sizing is still a guessing game. And lets please stop with the separate rails versus single rail flames. It is all one 12V source behind the multiple rail OCP's (so no "trapped amps"), and if your single rail exceeds 24A the wires will just melt anyway, which is why modern video cards have so many wires running into them.

I guess heat is my biggest issue with this build, a pair of GTX470's needs more space and each needs its own airflow out, and I'm not talking about a single empty PCI slot, either.

 

[Guest]

i have about the system in all the components(just using other brands in Mainboard and memory), using the corsair 750tx, and just bought 2 GTX460 cards, do u think the system will do ok with this PSU? i have four 120mm fans also