System Builder Marathon, Q2 2013: System Value Compared

[Crashman]

Our top three builders struggled to fit big performance into three tiny ITX-based platforms at three budget levels. Will the more expensive machines once again justify their higher prices through dramatically-improved performance? Read on for more.

System Builder Marathon, Q2 2013: System Value Compared : Read more

 

[swordrage]

Finally some builds that cost nearly the same in my country India. Thanks..

 

[manitoublack]

Great to see M-ITX in a SBM. The days of needing a full sized ATX are mostly over for 90% of people. M-ATO or M-ITX is the way forward.

 

[slomo4sho]

The extra $1200 from the $1300 doesn't add much value in this form factor.

 

[DarkSable]

Sigh. This is why I really, really dislike the system builder marathons; they do nothing but perpetuate the fallacies that already are far too common.

Someone looking at just this article, which isn't that unlikely, would be lead to believe that an i7 is something that an "ultimate" gaming computer has, that an expensive motherboard helps, and that a $2500 PC is going to be far better than a $1500 one.

 

[slomo4sho]

Sigh. This is why I really, really dislike the system builder marathons; they do nothing but perpetuate the fallacies that already are far too common.

Someone looking at just this article, which isn't that unlikely, would be lead to believe that an i7 is something that an "ultimate" gaming computer has, that an expensive motherboard helps, and that a $2500 PC is going to be far better than a $1500 one.

They really should include performance per dollar figures in this writeup.

 

[DarkSable]

They really should include performance per dollar figures in this writeup.

For the parts, or for the computers themselves? Either would be nice, actually.

One thing that would go a long way is stressing how wonky their testing is - most people reading this as advice for building a computer are going to be building a gaming computer purely, rendering 70% of the test bench pointless.

 

[slicedtoad]

I still don't like the bf3 benchmarks. They in no way represent the online experience and really, people that play bf3 spend at least 95% of their time on mp. I realize it's nearly impossible to generate a fair benchmark for online play but the current benchmarks are very misleading.

And I'm not griping at tom's, all review sites seem to do this. There should be some way to create a better benchmark. Maybe host a custom server and load it up with scripted "players" or something.

 

[allanitomwesh]

Cheaper is better basically Where's that $400 rig?

 

[Achoo22]

most people reading this as advice for building a computer are going to be building a gaming computer purely, rendering 70% of the test bench pointless.

I feel like they've modified the benchmarking suite to favor AMD as much as possible.

 

[slomo4sho]

I feel like they've modified the benchmarking suite to favor AMD as much as possible.

And when was the last time an AMD CPU made it into a SBM? Modifying benchmarks to favor a product that is never showcased is a moot point.

 

[Amdlova]

I got in mine micro atx 3 intel cpu... i7 3770k i3 3225 and now i have here the i5 3470 and what i can say. 3770k IS fast but hot... i3 3225, its ok can run everthing on max. when u pull 107 mhz on fsb he wake up and get better fps even on gta 4, u don't need a cooler to run the i3. and for now i got here the i5 3470 this have the best valuei can run this at 4.0ghz and i don't get my case hot. good speed barely touch the 3770k on applications. the mostly demanding i miss its about zip and rar files... 3470 epic win. 180usd

 

[de5_Roy]

Sigh. This is why I really, really dislike the system builder marathons; they do nothing but perpetuate the fallacies that already are far too common.

Someone looking at just this article, which isn't that unlikely, would be lead to believe that an i7 is something that an "ultimate" gaming computer has, that an expensive motherboard helps, and that a $2500 PC is going to be far better than a $1500 one.

They really should include performance per dollar figures in this writeup.

were you looking for something other than the performance per dollar charts present in the last page?
if you're looking for perf/$$ for individual componentes, look into the component reviews. sbm has figures for the whole pc only, because the whole pc is being tested.

 

[jee_are]

They really should include performance per dollar figures in this writeup.

Isn't that exactly what the last two graphs are all about?

 

[cscott_it]

@DarkSable

Did you even read the article? At all - or did you just flip through charts? And then not all of the charts, just some of them. In every SMB they always talk about diminishing returns and sweet spots - ALWAYS. And nearly every time the lowest end wins the price/performance category. I've seen the mid rig win a few times, but that was only when they were doing off-the-wall rigs.

 

[ojas]

You know what, @DarkSable read the last 2 pages.

Anyway, that's not what i had to say.

What i had to say was, always looking at perf/$ is sort of narrow minded too.

If someone wanted a minimum of 50 fps maxed out at 1600x900 and above, they'd be looking at the $1300 build.

If someone wanted the best of everything, they'd look at the $2500 build.

I personally look to spend that much that gives me around 50-60 fps minimum maxed out (with at least 4xAA) in all games that i play, at the resolution that i play on. Any additional funds go into other things, like storage, power, cooling, the case, etc.

I'm a big believer in an all round rig. If i'm spending money, i don't want to regret small sacrifices later, and i do that a lot (regret small sacrifices).

 

[Onus]

...
.
.
.
I personally look to spend that much that gives me around 50-60 fps minimum maxed out (with at least 4xAA) in all games that i play, at the resolution that i play on. Any additional funds go into other things, like storage, power, cooling, the case, etc.

I'm a big believer in an all round rig. If i'm spending money, i don't want to regret small sacrifices later, and i do that a lot (regret small sacrifices).

This. You don't build a PC without specific purpose(s), and without performance targets. And, you cannot judge a build without considering the purpose(s) for which it was built. The SBM PCs are built to compete in certain benchmarks (and to encourage lots of discussion, hopefully intelligent). Most people don't build that way (which is no slight at the SBMs; they are consistently one of my favorite features on this site). My primary PC has a card reader, and one of those 5-1/4" drawers, and a pair of drives for storage in RAID1; you'll never find those in a SBM, nor would I ever call for them. The SBM provides interesting performance and general build data points, and does not claim to be a "build this" instructional article. I do remember a "Build a $500 Gaming PC" article some years ago (featuring a Pentium 805D) which very likely influenced the SBMs, but was written very differently. "Build a ________ PC" would indeed make another interesting and useful instructional series, perhaps one every 2-3 months, NOT always focused on gaming. I'd suggest every other one be devoted [primarily] to something other than gaming, such as HTPC, Home Office, "typical" office, CAD, database, etc. These need not be given away, but could be used as instructional articles for people looking to build. Flesh them out by publishing videos of each actual build, such as on YouTube.

 

[RedJaron]

Are the CPU temps on page 13 accurate? Don's build stays essentially room temp while idling? Paul's only goes up 2 ºC at load? Impressive.

 

[lowguppy]

Any chance you could benchmark the Tiki for comparison?

 

[RedJaron]

...
.
.
.
I personally look to spend that much that gives me around 50-60 fps minimum maxed out (with at least 4xAA) in all games that i play, at the resolution that i play on. Any additional funds go into other things, like storage, power, cooling, the case, etc.

I'm a big believer in an all round rig. If i'm spending money, i don't want to regret small sacrifices later, and i do that a lot (regret small sacrifices).

This. You don't build a PC without specific purpose(s), and without performance targets. And, you cannot judge a build without considering the purpose(s) for which it was built. The SBM PCs are built to compete in certain benchmarks (and to encourage lots of discussion, hopefully intelligent). Most people don't build that way (which is no slight at the SBMs; they are consistently one of my favorite features on this site).
...
I do remember a "Build a $500 Gaming PC" article some years ago (featuring a Pentium 805D) which very likely influenced the SBMs, but was written very differently. "Build a ________ PC" would indeed make another interesting and useful instructional series, perhaps one every 2-3 months, NOT always focused on gaming. I'd suggest every other one be devoted [primarily] to something other than gaming, such as HTPC, Home Office, "typical" office, CAD, database, etc. These need not be given away, but could be used as instructional articles for people looking to build. Flesh them out by publishing videos of each actual build, such as on YouTube.

This hearkens to Paul's comment a few days ago how so many people think a certain budget range automatically denotes certain components, or how you can't claim a certain computer category name if you don't have certain components. I feel like Barbosa in that the code is more about guidelines than actual rules. Everyone can and should put their own spin on a build for their own purposes.

I like the idea of an instructional series of how to build machines specialized to a certain task. I remember Tom's used to have a standard configuration area where people could submit build lists for a lot of computer types like Sub $500 Intel, Sub $500 AMD, HTPC, Home & Office, Professional Design, Mid-range Gaming, All-Out Gaming, etc. Do those still happen? Have I just lost track of them due to the site changes?

 

[slomo4sho]

were you looking for something other than the performance per dollar charts present in the last page?
if you're looking for perf/$$ for individual componentes, look into the component reviews. sbm has figures for the whole pc only, because the whole pc is being tested.

Isn't that exactly what the last two graphs are all about?

Those graphs use the $650 build as the baseline for comparison. This isn't a true performance per dollar representation.

 

[de5_Roy]

Those graphs use the $650 build as the baseline for comparison. This isn't a true performance per dollar representation.

then what is the true perf/dollar representation?

 

[slomo4sho]

then what is the true perf/dollar representation?

Performance/cost

Tom's uses the arbitrary formula (performance/performance of $650 build)/(cost/$650). These figures are useless outside of this current article whereas the performance/cost calculations can be used universally.

 

[RedJaron]

Those graphs use the $650 build as the baseline for comparison. This isn't a true performance per dollar representation.

Maybe so, but considering the number of variables ( CPU, RAM, GPU, storage, and the overclock amount for each, ) performance per dollar becomes a very difficult thing to lock down and present. At least at the whole system level. To get full representation, you'd need to test every GPU with every combo of CPU, RAM, and possibly mboard ( mboards themselves don't directly affect game perf, but their varying OC abilities do. ) Each of those combos would also need to be checked at stock clocks and OC.

In the interest of time, you could cut down the number of combos to only use some of the most common tiers of components. Find the top OC mboard and RAM for each CPU, and use those for all the tests. Then just use just the top model FX4000, FX6000, FX8000, i3, i5, and i7 and a reference model of each main GPU model. Even cutting it down to that would still be very time consuming.

It'd actually be a pretty interesting research article, investigating how each GPU is boosted or limited by a given CPU, much RAM overclocking affects game performance depending on platform and graphics, etc. We can already extrapolate some of this info from the current GPU and CPU reviews. The numbers are given for the reviewed model in comparison to others in the similar range, so applying those to the MSRP could give a somewhat usable baseline.

 

[de5_Roy]

Performance/cost
Tom's uses the arbitrary formula (performance/performance of $650 build)/(cost/$650). These figures are useless outside of this current article whereas the performance/cost calculations can be used universally.

i think directly dividing performance by cost makes it easy to misrepresent if a baseline isn't present. sbm uses cheapest machine as a baseline. kinda like measuring without a scale... i think. even a universal measure will need a baseline and even that measurement will be useless outside the scope of the measurement. i hope that made some sense.