System Builder Marathon, Q4 2013: A $2400 PC That Costs $2700

Page 2 - 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.
I can't fault the parts used. Nothing at all howls for replacement. My issue is with the build as a whole; way too expensive overall. It may well be a top-quality system, but it will get killed in the value comparison.
 



Asrock Xt 6 230.00
i7 4930K 569.00
Gskill F3-1866C9D Ripjaws 158.00
Gigabyte GV-N770OC-4GD x3 (SLI) 768.00
Seagate ST2000DX001 130.00
Samsung 840 EVO 256GB 170.00
Seasonic X-1050 247.00
Corsair 500R 130.00
CM EVO 212 $35
$2437

Considering the real-world cost of your build, I'd say this would smoke it. You can argue I didn't figure in case fan cost, but the case itself is available w/3 fans, nor did I calculate for a Blu-ray drive, however I don't think the BR drive will change the performance, do you? :)
 

Might want to check that. I'm sure a lot of people would love to know where to get a 4GB 770 for only $256. Those 770s are $390 on Newegg. You're definitely not packing three of them in that budget.
 
Oh, we're cost-cutting on features then? Yeh, that works but it's nothing I would have built. I like a high-end system to have all its faculties.

SLI would have worked better in two of the gaming benchmarks. I was caught off guard by the inability of F1 2012 to support these cards in CrossFire, and by the capping in FC3. But I couldn't have known that would happen when I ordered the cards.

And the Extreme6 is garbage, read that last SBM then read the Extreme6/GB review.

 


I as well would like to know where to get 4gb 770s for $250...
 


That should actually be x2, or add $390 to the price, and be close to his real-world price.
 


I like the way you think. I first got into F@H because it seemed like a waste to have decent hardware not producing anything while I am not using it.

I am a bit surprised to find out that the Tom's staff is aware of F@H. How about throwing your team a bone and doing a F@H hardware round up or just donate a week of each build's time to the team's points.
 


I have an Z77 Extreme4, which even Tom's rated pretty high, but I don't have real-world experience with the X79 Ext 6, so I'll take your word for it.

We could use your board, but the cost wouldn't be much different. I think the drive performance would make a difference, though: Hybrid vs. WD Green? The EVO and SanDisk are pretty well matched, but I think a benchmark would be necessary to see if the combination of either + hybrid would be any different.

I could have opted for one of many less-expensive cases and put in the BD but I was focused on performance.

I did miscalculate by saying x3 vs. x2 770s, however, for the actual price of your build, I think this would still be a better option; another $117 for 3-way SLI vs 2-way Crossfire.

I also agree with some of the opinions of the CPU, but I didn't want my build to be discounted based simply on using a cheaper CPU. I think a properly OC'd 4930, 3930 or if we moved to Haswell, maybe a 4770K with some luck, might come close in a lot of these tests for significantly cheaper.
 
two r9 290Xs would cost about $1370.88 here in england

although there is VAT and the r9 290X does not appear to have made it over here... (at least all the places i am looking at only have them up for pre-order)
 
Crazy, I'm in the process of building the same rig! except with an 8350 over the i7, 8gb of ram isntead of 16, no optical, no ssd, a bigger case, a better PSU, Noctua nh-d14, and non reference cards...and all of this for less money, hmph..
 
So, you're building far less PC for far less money? Who'd have thought? Thanks for sharing!
Look at my account history, I was here in the forums all along :)

 

It's only peaking at 99% of rated capacity. I've tested similar PSUs to 120% of rated capacity before they tripped the breaker. I haven't fried a high-end PSU.

Moreover, I don't know of anyone who will run this machine at peak. The specific part of the test that's unrealistic at this point is the twelve threads of Prime95 with AVX. It's just way above even a typical Prime95 load. And I had to back down to 11 threads to allow a thread for 3DMark test 1 looping (combined load), otherwise the power results would have been around 75W lower. 3DMark needs that single CPU thread to run the graphics cards at full load.
 


I would think better cooling cpu&case, a better mobo, and better cards would make it perform better and be by your wording "more" PC
 
But it's a lesser CPU, a lesser cooler, and the case was picked for its low-noise design and CPU cooler support. I don't want a big case, just a quiet case that cools well. Let me misquote you:
I can build a noisier PC with a lower-IPC CPU, half as much RAM, no ability to read optical media, and a painfully slow data drive, for less money.
Click to expand...
When we put what you said into that context, my answer makes sense :)

 
[/quotemsg]
I don't know of anyone who will run this machine at peak. The specific part of the test that's unrealistic at this point is the twelve threads of Prime95 with AVX. It's just way above even a typical Prime95 load. [/quotemsg]

Well I will give you that.

It did answer a question I have had about whether your system power was from the supply or from the wall. It seems that you must be using a Kill A Watt type device and just measure the total draw. That isn't obvious in a lot of the reviews.
 
I'm exactly using a Kill A Watt. Don on the other hand has a more complex device with data logging, IIRC.

 
Read all the articles, put in all the contest entries.
also a member of the Tom's Hardware Folding Team.
Glad to have someone else voice the same opinion on Bitcoin/litecoin and similar.. I get it from their point of view- get some hardware, plug it in, let it work to bring in money instead of you. - only time it works enough to be worth the time is if someone else is paying for the hardware, power, or both.

Anyway. -I have an old Cooler Master chassis with the foam inside it ( Sileo)? I think..
and I have been using the corsair builder power supplies to great success.
I would actually like someone to compare the wattage draw to heat output to a cheap space heater. someof the old Prescott or Cyrix (anyone remember them?) would get hot enough to melt the socket.
On topic, I like the Idea of this build, but cant figure any way for it to be practical in a real world sense.. too many overcompensations to make it good at both games and productivity..
 
I believe this build needed more fans and a readjustment of the cooling:

I think putting the radiator in the front was bad idea. You are blowing warm air INTO the computer therfore affecting the GPUs, of which we knew would have more heat issues than the CPU.

I think the optimum setup would have been to put the radiator in the top, buy 2 more case fans: 1 extra front intake (total of 2) and one bottom intake (after you returned the drive cage to its original position and cleaned up your cable management) or to the side panel, which ever works better. You then could have positive air pressure if you adjusted the fan speeds accordingly, more air to the graphics plus you would still get the benefit of exhausting air off the motherboard VRMs.

Also from the look of the picture, cmon, your PCIE cabling is an utter mess, it appears you have blocked the path from intake fans to graphics cards. Rout all the PCIE cables over the GFX cards and around the back of the top card therefore clearing the cards intake fans. This is simple stuff guys.
 


File this under: Don't quit your day job...
 
Great Article! Always love reading the system marathons... Is it possible to include RMAA results for the motherboards in these marathons and in motherboard reviews as well?... It's nice to know at what point in time will onboard audio be enough for me to stop using my 7 year old x-fi xtremegamer>
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom