System Builder Marathon, Q1 2013: $1000 Performance PC
Tags:
- System Builder
- Performance
Last response: in Reviews comments
Crashman
February 28, 2013 3:00:04 AM
With his budget cut in half, Thomas is trying to take the value crown by building the highest-end PC in this quarter's System Builder Marathon for $1,000. Can he out-do the value of Paul's $600 config or Don's $800 effort using solid-state storage?
System Builder Marathon, Q1 2013: $1000 Performance PC : Read more
System Builder Marathon, Q1 2013: $1000 Performance PC : Read more
More about : system builder marathon 2013 1000 performance
samuelohagan
February 28, 2013 3:13:22 AM
iknowhowtofixit
February 28, 2013 3:19:13 AM
Related resources
- System Builder Marathon, Q1 2014? - Forum
- System Builder Marathon, August 2012: $1000 Enthusiast PC - Forum
- why is the system builder marathon 2013 based on mini itx plat form? - Forum
- Looking at System Builder Marathon (or any other PC) for New Build - Forum
- Help! System Builder Marathon, Q2 2014: Our Budget Gaming PC - Forum
mayankleoboy1
February 28, 2013 3:20:29 AM
Chairman Ray
February 28, 2013 3:20:56 AM
samuelohagan
February 28, 2013 3:23:36 AM
lightofhonor
February 28, 2013 3:31:35 AM
dscudella
February 28, 2013 3:32:15 AM
Anonymous
February 28, 2013 3:40:39 AM
Crashman
February 28, 2013 3:41:40 AM
mayankleoboy1The only thing i could have tried to change in this build was to get a 64GB SSD , get a 500GB HDD , and if i have any money left, try to get a HD7950.Otherwise, not much wriggle room here. Nice build!
No point, We Like Our Games To Load Fast!!!Chairman RayI'm not sure if having an SSD without an HDD is an attractive option for most people. 240GB goes by pretty fast.
Programs can eat most of a 240GB drive up. In fact, I just shrunk a copy of the test suite, with OS and games, to 132 GB. Using the drive performance measurement to reflect program load times means loading all the programs on the SSD. And that explains why SSD capacity wasn't sacrificed to make more room in the budget for an HDD.
lightofhonorIt actually plays Skyrim WORSE than the $800 build...
Nobody knows why, but the $800 PC did use a newer GPU driver.dscudellaIt's the same PC as the $800 build except they just swapped out the HDD for an SSD. Dissappointing.
So it's going to lose the Day 4 Value Roundup, right?Score
13
mayankleoboy1
February 28, 2013 3:41:52 AM
Score
7
mayankleoboy1
February 28, 2013 3:47:39 AM
coonday
February 28, 2013 3:47:42 AM
Crashman
February 28, 2013 3:58:47 AM
mayankleoboy1Looking from another perspective, these two builds, with two different builders, with $200 difference, just show(again) how much better price/performance wise are Intel CPU's and AMD GPU's.
Builder 2 copied Builder 1's power supply. Builder 1 copied builder 2's RAM. Both were stuck with the same video card due to availability issues
mayankleoboy1I see your point, but I'd rather see slower game loads and better FPS , than faster game loads and lower FPS. And, the OS is accelerated in both cases anyway.
Performance difference between Tahiti and Tahiti-LE is fairly small, right? But the price difference isn't small, is it? $1000 PC builder simply hoped for a better overclock, even though that didn't work out.coondayYou guys must really want me to buy that 7870 Myst Edition. I'm actually hoping to see 2 of those cards in a crossfire configuration on the $2000 build.
Nobody wants you to buy the Myst Edition. Tahiti-LE is a good value, if you like that argument pick your favorite vender. Vender choice wasn't available on purchase day.But hey, I'm on board with the 7870 Myst Edition CrossFire suggestion...I'll see if we can make it happen!
Score
3
J_E_D_70
February 28, 2013 4:02:15 AM
roltzje
February 28, 2013 4:38:12 AM
Im going o defend their choice here, this isnt "Build the best gaming pc for the money" its to build the best performance PC for the money. And having a large capacity SSD along with mid-upper end CPU/GPU seems good to me. If they wanted to build a pure gaming PC the extra $200 would just be spent on the GPU.
I think theres something to be said about the value at above $1000 though.. past this price range, people really start caring about having a nice case, nice cooler, etc that are more than just performance but aesthetics too.
I think theres something to be said about the value at above $1000 though.. past this price range, people really start caring about having a nice case, nice cooler, etc that are more than just performance but aesthetics too.
Score
5
rahulkadukar
February 28, 2013 4:45:11 AM
ta152h
February 28, 2013 5:42:58 AM
I think these would be more interesting if they pitted AMD against Intel/NVIDIA at the same price target.
Most likely they'd end up with a similar ugly case that no one would really want, possibly the same memory and hard disk, but the heart of the system would always be different.
Besides, people love rivalries. Sure, AMD processors blow in absolute performance, but they're cheap, and maybe the video card can save the day against the evil Intel/NVIDIA empires. It's a lot more interesting than testing two essentially identical machines, except for the hard disk.
Do it for different price ranges, and it might even be more competitive. $500, $750, and $1000 might not all have the same winner.
Most likely they'd end up with a similar ugly case that no one would really want, possibly the same memory and hard disk, but the heart of the system would always be different.
Besides, people love rivalries. Sure, AMD processors blow in absolute performance, but they're cheap, and maybe the video card can save the day against the evil Intel/NVIDIA empires. It's a lot more interesting than testing two essentially identical machines, except for the hard disk.
Do it for different price ranges, and it might even be more competitive. $500, $750, and $1000 might not all have the same winner.
Score
0
dudewitbow
February 28, 2013 5:46:45 AM
Crashman
February 28, 2013 5:51:14 AM
mayankleoboy1
February 28, 2013 5:57:20 AM
Crashman
February 28, 2013 6:24:08 AM
mayankleoboy1Lets do a "Worse of" builds, just for shites and gigglesPair a AMD CPU with a Nvidia GPU. So its expensive and may not perform as well.(I will be extremely happy if this build performs well)
I believe Nvidia has lower CPU overhead. So, if you're going AMD for the CPU, you'd might be onto something with the Nvidia GPU.Score
4
Yargnit
February 28, 2013 7:24:13 AM
This would make a much more interesting value comparison compared to the $600/800 build's if you'd stuck to a HDD and put the extra $ into GPU performance. Then we'd really get to see where the sweet spot lies. Burning the extra $ on a SSD basically ensures one of the cheaper builds will win.
Honestly I'd like to see all system builder marathon machines tested with the same hard drive configuration and be purely about CPU/GPU combinations. Everyone basically knows what they're getting by adding/subtracting a SSD at this point. They do nothing but skew value comparison's.
Honestly I'd like to see all system builder marathon machines tested with the same hard drive configuration and be purely about CPU/GPU combinations. Everyone basically knows what they're getting by adding/subtracting a SSD at this point. They do nothing but skew value comparison's.
Score
6
silverblue
February 28, 2013 7:54:35 AM
pauldh
February 28, 2013 8:08:22 AM
CrashmanI believe Nvidia has lower CPU overhead. So, if you're going AMD for the CPU, you'd might be onto something with the Nvidia GPU.
I don't know Thomas, that is possible. But this fluctuates by driver maturity. A year ago, I know for a fact SLI had way less CPU overhead. When earlier, back in 2009, Nvidia was behind both single GPU and dual. At one point AMD's driver team made huge CrossFire improvements, in terms of CPU scaling, rendering a ton of my data, dated and useless.
From the cheap seats I have strong suspicions NVidia driver's now have greater thread dependancy, at least for single GPUs. So Pentiums will compete better when paired with Radeon is my hunch. Since there were other factors clouding this belief, I'm now looking into it.
Score
4
pauldh
February 28, 2013 8:15:51 AM
sanilmahambre
February 28, 2013 8:45:22 AM
stoogie
February 28, 2013 9:05:45 AM
Crashman
February 28, 2013 9:41:22 AM
Anonymous
February 28, 2013 10:25:52 AM
de5_Roy
February 28, 2013 11:08:55 AM
... looks like intel cpus, amd gpus and asrock made a clean sweep this quarter. so did xigmatek, to a lesser extent. personally, i didn't like any of the motherboards. i've seen the extreme4 and 6, and those look weak, kinda flimsy despite being very feature-rich.
keeping the roomy ssd will likely help score against the $800 build while keeping rest of the scores similar. interesting.
will there be a surprise $800-1000 amd-build-for-teh-lulz? like with fx6300/8320, cm hyper 212 evo, asrock(a-gain) 970 mobo, and so on. i think the amd fanboys would be pacified a bit.
keeping the roomy ssd will likely help score against the $800 build while keeping rest of the scores similar. interesting.
will there be a surprise $800-1000 amd-build-for-teh-lulz? like with fx6300/8320, cm hyper 212 evo, asrock(a-gain) 970 mobo, and so on. i think the amd fanboys would be pacified a bit.
Score
0
silverblue
February 28, 2013 11:17:44 AM
mayankleoboy1
February 28, 2013 11:53:03 AM
Score
1
ojas
February 28, 2013 12:01:23 PM
Intel and AMD work well together. The irony.
I'm wondering, could you guys look into this stuff:
1) An article that compares GPU performance on a range of processors from Intel and AMD from the Core 2 era, by using a 680 first and then a 7970.
2) Thomas said he'd like to repeat the Cross-Fire scaling article (FX vs Core i7) with an SLI config...
3) $600, $900, $1200 builds for next time's SBM
4) A10-5800K vs FX-8350 vs Core i3 3225 vs Core i5 3570 vs Core i7 3770 (yes, all non-K Intels) value comparison, with games capped to 60 fps (using vsync), with frame time analysis
5) Same as (4) but this time running a virtualized Windows environment
6) I think time for a Linux Distro v12.10 vs Win 7 vs Win 8 vs Mac OS X 10.8 performance comparison?
7) Memory bandwidth (and the impact it has had) and total power consumption (over the benchmark's duration) comparison from the Core 2/Phenom II days to today, use only 4C/4T processors.
8) Crysis 3 benchmark?
9) Adobe Photoshop CS6 Open CL comparison: AMD vs Nvidia across multiple price points
10) actually i think (7) should be divided into two, i'd actually like to see the performance impact of an external memory controller, in terms of latency, bandwidth and frame-times. Or something like that. External vs internal, basically.
I think that's all for now
I'm wondering, could you guys look into this stuff:
1) An article that compares GPU performance on a range of processors from Intel and AMD from the Core 2 era, by using a 680 first and then a 7970.
2) Thomas said he'd like to repeat the Cross-Fire scaling article (FX vs Core i7) with an SLI config...
3) $600, $900, $1200 builds for next time's SBM
4) A10-5800K vs FX-8350 vs Core i3 3225 vs Core i5 3570 vs Core i7 3770 (yes, all non-K Intels) value comparison, with games capped to 60 fps (using vsync), with frame time analysis
5) Same as (4) but this time running a virtualized Windows environment
6) I think time for a Linux Distro v12.10 vs Win 7 vs Win 8 vs Mac OS X 10.8 performance comparison?
7) Memory bandwidth (and the impact it has had) and total power consumption (over the benchmark's duration) comparison from the Core 2/Phenom II days to today, use only 4C/4T processors.
8) Crysis 3 benchmark?
9) Adobe Photoshop CS6 Open CL comparison: AMD vs Nvidia across multiple price points
10) actually i think (7) should be divided into two, i'd actually like to see the performance impact of an external memory controller, in terms of latency, bandwidth and frame-times. Or something like that. External vs internal, basically.
I think that's all for now
Score
8
I did not care for this build. It was made to perform in benchmarks, but I do not think is a realistic representation of what someone would actually build. Other than lacking a hard drive, it is, however, what someone might end up with if they started with a lesser build (e.g. the $800 PC) and added a SSD to it.
Score
-7
SuperVeloce
February 28, 2013 12:26:03 PM
mayankleoboy1
February 28, 2013 12:50:34 PM
aebome
February 28, 2013 1:00:47 PM
What's with the case selection here (and even on the previous two builds)? I know it doesn't add anything to the performance metrics (even though it does affect thermals/noise) but I'd like to see some thought go into the case, considering that is what you see the whole time you have the computer, and it what you build in. Just being cheap isn't enough. At this price range $5-10 makes a difference in case choices, and probably won't affect performance component selection.
For $1000 build why wasn't the Corsair 200R used, for example? It looks a lot nicer and less like 9 yr old gamer than the Rosewill chosen. It's not like it is a huge price difference. At least the Xigmatek's chosen in the prior builds are less eye searing, although for the same amount of money, I'd go with the NZXT Source 210: better cable management, and cleaner looks.
For $1000 build why wasn't the Corsair 200R used, for example? It looks a lot nicer and less like 9 yr old gamer than the Rosewill chosen. It's not like it is a huge price difference. At least the Xigmatek's chosen in the prior builds are less eye searing, although for the same amount of money, I'd go with the NZXT Source 210: better cable management, and cleaner looks.
Score
2
hqarmstrong
February 28, 2013 1:07:39 PM
onusI did not care for this build. It was made to perform in benchmarks, but I do not think is a realistic representation of what someone would actually build. Other than lacking a hard drive, it is, however, what someone might end up with if they started with a lesser build (e.g. the $800 PC) and added a SSD to it.
For what it's worth this is quite similar to the system I've built over the past two weeks, based largely on following articles and forum threads right here on TH. (very much appreciated)
Prior to these recent batch of Q1 articles I purchased the following:
$190 - I5 3570k (price matched at a local Fry's Electronics to a sale on microcenter.com)
$110 - ASRock Z77 Pro4 (with 8GB of free Corsair RAM - Newegg)
$238 - Sapphire 7870 XT (reviews I read commended the card's superior cooling, superbiiz.com)
$150 - Samsung 840 250GB SSD (w/ $100 restaurants.com giftcard, on cnet's deals page)
$55 - XFX Pro 550W PSU (w/ $20 MIR - Newegg w/ a promo code)
I tossed all that gear in a second-hand Lian Li PC-B25F along with a Corsair H60 I bought used for $20.
It's been close to 8 years since I built a new system, so I'm bloody stoked.
Score
2
Novuake
February 28, 2013 1:10:59 PM
de5_Roy
February 28, 2013 1:12:24 PM
ojas said:
Intel and AMD work well together. The irony.
I'm wondering, could you guys look into this stuff:
1) An article that compares GPU performance on a range of processors from Intel and AMD from the Core 2 era, by using a 680 first and then a 7970.
2) Thomas said he'd like to repeat the Cross-Fire scaling article (FX vs Core i7) with an SLI config...
3) $600, $900, $1200 builds for next time's SBM
4) A10-5800K vs FX-8350 vs Core i3 3225 vs Core i5 3570 vs Core i7 3770 (yes, all non-K Intels) value comparison, with games capped to 60 fps (using vsync), with frame time analysis
5) Same as (4) but this time running a virtualized Windows environment
6) I think time for a Linux Distro v12.10 vs Win 7 vs Win 8 vs Mac OS X 10.8 performance comparison?
7) Memory bandwidth (and the impact it has had) and total power consumption (over the benchmark's duration) comparison from the Core 2/Phenom II days to today, use only 4C/4T processors.
8) Crysis 3 benchmark?
9) Adobe Photoshop CS6 Open CL comparison: AMD vs Nvidia across multiple price points
10) actually i think (7) should be divided into two, i'd actually like to see the performance impact of an external memory controller, in terms of latency, bandwidth and frame-times. Or something like that. External vs internal, basically.
I think that's all for now
1) a Lot of work.
3) how about $600, $800 and a $1000 amd-powered sbm builds with the same goal as this quarter and compare that data to this quarter's and figure out once and for all which build offers better bang for buck overall and for specific tasks such as 1080p gaming. prices change all the time, next quarter, who knows, amd might have an advantage.
4) capping fps introduces an artificial gpu bottleneck which will in turn give the illusion of all the cpus performing similarly.
7) would show amd cpus in a very bad light.
reviewers might need to bring out task-specific performance per dollar comparisons to lessen the impact of worse performance per watts.8) Ditto. i'd like to see how toms bench cpus for crysis 3. i've been seeing some crysis 3 benches being paraded around like the time techspot showed bf3 sp benches showing amd and intel cpus performing similarly making amd fanboys very very happy....until 2011's sub $200 gaming cpu roundup came out.
Score
0
aebome
February 28, 2013 1:13:24 PM
cknobman
February 28, 2013 1:35:36 PM
Although a interesting approach I dont like limiting the "high" end build to $1000. It really defeats the purpose of making a high end build to begin with and while the value and performance per dollar is nice you loose out on intangibles that make a high end build nice.
I am not willing to cut these just to save a few bucks (on a high end build):
Dedicated HDD storage drive in addition to an SSD boot drive.
Blue Ray drive
Hyperthreading (yes I video encode a lot so it provides a benefit I want)
Better air cooling
GPU - the chosen card is fine but in a high end build I would opt for a little more power under the hood.
To sum it up: dont skimp so much on the high end builds any more it encroaches too much on the mid level build and sacrifices too many intangibles that I expect to see in a high end machine.
I am not willing to cut these just to save a few bucks (on a high end build):
Dedicated HDD storage drive in addition to an SSD boot drive.
Blue Ray drive
Hyperthreading (yes I video encode a lot so it provides a benefit I want)
Better air cooling
GPU - the chosen card is fine but in a high end build I would opt for a little more power under the hood.
To sum it up: dont skimp so much on the high end builds any more it encroaches too much on the mid level build and sacrifices too many intangibles that I expect to see in a high end machine.
Score
0
ARICH5
February 28, 2013 1:48:15 PM
Score
2
BSMonitor
February 28, 2013 2:18:15 PM
MasterMace
February 28, 2013 2:26:50 PM
ibjeepr
February 28, 2013 2:34:47 PM
Nice job on the build but it's clear the $800 price point is the way to go.
I've always found the price points to be at $500-$800-$1200-$1500 and up. As someone else mentioned, the $1200 price point is to get you to the next tear of video cards and a little fancier looks.
Overal, very close to my build. I splurged on a little extra ram, power supply and a fancy case (first time) but I even used the Redbone on my cheaper i3-3220 build. I'm also still trying to decide how much I want to spend and how long I want to wait to buy a new GPU but have been staring at the Sapphire 5870 XT.
I've only overclocked the CPU to 4.0 at this point.
ASrock Z77 Extreme4
i5-3570K
CM 212 EVO
G-skill Ripjaws 12gb 1600 9C
Sapphire 5870 Vapor-X OC
Samsung 830 128Gb; WD Blue 320GB (from old system)
Seasonic X850
Antec 900 USB 3.0
I've always found the price points to be at $500-$800-$1200-$1500 and up. As someone else mentioned, the $1200 price point is to get you to the next tear of video cards and a little fancier looks.
Overal, very close to my build. I splurged on a little extra ram, power supply and a fancy case (first time) but I even used the Redbone on my cheaper i3-3220 build. I'm also still trying to decide how much I want to spend and how long I want to wait to buy a new GPU but have been staring at the Sapphire 5870 XT.
I've only overclocked the CPU to 4.0 at this point.
ASrock Z77 Extreme4
i5-3570K
CM 212 EVO
G-skill Ripjaws 12gb 1600 9C
Sapphire 5870 Vapor-X OC
Samsung 830 128Gb; WD Blue 320GB (from old system)
Seasonic X850
Antec 900 USB 3.0
Score
2
spidey81
February 28, 2013 2:41:35 PM
I have this same motherboard with a 2700k. There's a known "flaw" with the way Asrock reports and/or applies voltage to the cpu. The only way I've found to verify this is by using a multimeter to read the voltage across one of the caps feeding the cpu from the back of the board. Under load I've seen nearly .1V difference from what CPU-Z is reading and what is actually being fed to the processor. So under load you may actually be seeing closer to 1.4V than 1.3V. Just a heads up to anyone using or thinking of getting this, or most other Asrock Z77 motherboard.
Score
4
TeraMedia
February 28, 2013 3:13:20 PM
Compared to the other two systems, this one was disappointing. With $20 budget left over, the case could have been better. Rather than a 240 GB SSD, the GPU could have been made more interesting - perhaps a bump-up to a 7950, or maybe some SLI / Xfire solution using cheaper cards with a combined price at ~300. Not necessarily because it would be definitively better. Just because it's possible, and to shake things up a bit.
Score
0
Yuka
February 28, 2013 3:14:10 PM
- 1 / 2
- 2
- Newest
Related resources
- SolvedAround 1000€ August 2013 System: AMD FX-8350 vs Intel Core i5 - First Time Builder Forum
- New $1000 gaming system or System Builder Marathon, Q4 2012: $1,000 Forum
- System Builder Marathon: Price/Performance Forum
- System Builder Marathon: Price/Performance Forum
- With Tom's System Builder Marathon in mind: Build a Infinite Budget PC Forum
- System Builder Marathon, May '09: $1,300 Enthusiast PC Forum
- System Builder Marathon: Sub-$4000 PC Forum
- More resources
!