System Builder Marathon Q1 2014: System Value Compared
-
Page 1:Three Strong Systems Face Off
-
Page 2:Benchmark And Overclocking Configurations
-
Page 3:Results: 3DMark And PCMark
-
Page 4:Results: SiSoftware Sandra
-
Page 5:Results: Battlefield 4
-
Page 6:Results: Arma 3
-
Page 7:Results: Grid 2
-
Page 8:Results: Far Cry 3
-
Page 9:Results: Audio And Video Encoding
-
Page 10:Results: Adobe Creative Suite
-
Page 11:Results: Productivity
-
Page 12:Results: File Compression
-
Page 13:Power Consumption And Heat
-
Page 14:Overall Performance And Efficiency
-
Page 15:Picking A Performance-Value Winner
Results: 3DMark And PCMark
Futuremark’s graphics benchmark heavily favors the $2400 machine’s two GeForce GTX 780 cards, producing a chart of nearly perfect performance scaling between five of the six tested configurations. The $750 machine doesn’t live up to its scaling expectations when Paul overclocks it, but that's probably because it's constrained by a multiplier-locked CPU.
That same $750 machine outperforms Don’s $1600 PC in PCMark’s “Work” test, at least at baseline frequencies. At first I was unable to explain Don's loss. But after reading his write-up, I figured out that he was running his baseline numbers with his memory underclocked to DDR3-1333.
PCMark’s storage suite uses traces recorded from a number of productivity, content creation, and entertainment apps to emphasize real-world differences between drives. That's why a lot of SSDs will appear incredibly similar in this particular metric.
This is the only synthetic benchmark used in our value scoring, which is going to give the two SSD-equipped boxes an advantage over the $750 machine.
- Three Strong Systems Face Off
- Benchmark And Overclocking Configurations
- Results: 3DMark And PCMark
- Results: SiSoftware Sandra
- Results: Battlefield 4
- Results: Arma 3
- Results: Grid 2
- Results: Far Cry 3
- Results: Audio And Video Encoding
- Results: Adobe Creative Suite
- Results: Productivity
- Results: File Compression
- Power Consumption And Heat
- Overall Performance And Efficiency
- Picking A Performance-Value Winner
A couple of thoughts:
1. Is it true that gaming only gets 20-30% of the overall performance weighting? If so, I'd suggest moving that up to at least 50%. Even though I'm one for a balanced system, I still think the amount of $$ you're spending on the graphics capability should be reflected more in the overall performance rating.
2. There is quite a bit of back-and-forth on the impact of power consumption on these forums, and I think this competition would be a great place to factor in and raise awareness on the real costs of power consumption. I understand that electricity costs and system use varies greatly. But I would add in another value analysis incorporating the present value of 3 years' electricity costs using a discount rate (simple excel function 'pv'). Of course you would have to make some broad assumptions around average power, hours of use per day, whether or not you idle the system 24x7, and electricity cost, but I think the present value of 3 year electricity cost would make a decent impact on your value calculations across these systems.
I would suggest making a forum poll regarding this. From my steam library I can tell that I've been playing games on average 2.25 hours a day over the last 5 years, and since the bulk of my games are on steam (1400+) it's not much below my real usage. However despite being a gamer I find myself having the computer either idle or playing youtube etc for an amount equal to the time I'm playing games. Assuming I'm around average, with regards to gaming:idling ratio, that'd mean perhaps 3 hours gaming and 3 hours idle/video playback a day on average.
i didn't agree with any of the motherboard choices in this quarter's build. despite that, all three builds were very interesting in terms of performance, choice of parts and builds.
my takeaway from this is that ddr3 1333 is not the baseline for cheap system memory anymore. 2x 4GB ddr3 1600 and higher, especially 2x 8GB ddr3 1866 or higher memory is optimum for performance. don not enabling xmp for the $1600 pc's stock performance analysis helped me understand this.
haswell i7 ramped power and heat really high after 4.4ghz and bit more voltage. both were higher than don's previous o.c. of i5 4670k with asus z87-a board. i still don't like asrock. can you guys compare o.c. clockrates, temperature and voltage of haswell cpus used in sbm in q4 to see which combo was better? imo it'd help with how haswell behaves in real pcs instead of open test benches or test pcs.
I'll repeat again my request for non-linear value assessments for FPS>60 in games. I'm not saying it's invisible, but once play is smooth, the subjective experience is not going to get much better. This will make Paul's machine even better in the value analysis.
I'm really not sure what I would do with any of these. They're all built with some nice parts, but my existing PCs are meeting my needs quite nicely.
1. $2400 PC: Massive overkill for my needs, but there are things about it I like. I might build this with just one of the graphics cards, then pull the RAID Array out of my "Omega" system and add that, then donate the remainder of Omega for use as a server to a group that needs one. The second graphics card I'd probably give away to a Tom's reader who is not in the US.
2. $1600 PC: Not sure; I'd probably do some mixing and matching with parts from "Phoenix," likely ending up donating most of this one too (except for the graphics card).
3. $750 PC: I'd pull the graphics card for a HD6850, add a SSD, and build it for my Mom. She's not a gamer, but would no doubt appreciate the speedup from her older AM3 machine for editing her photos and other media.