System Builder Marathon, Sept. '09: $2,500 Performance PC

Page 3 - 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.
Well I have a 850hx and love it, whether it is silver or gold..... (you all are bickering over what, 2-3%?!)

Interesting direction Tom's went on this build, I was interested to see how they would make an AMD box $2500 haha. The answer was 2 ssd's, 2 hdd's, a rather expensive case, and a blu ray burner. So what I took from this, is if you are building a gaming box using AMD platform and have a large budget, you can get ALL the bells and whistles!

Nice system and I will hope to win it as well 😉
 
Great job. I'm glad to see you guys branching out lately. Anyone can slap together some i7 with a GTX 295 and then call it "high end", but you guys are really doing some interesting builds that no one has seen much on yet. Keep it up! While it may incite some fanboyish rage in some people, it keeps others reading. Also, the person who wins that may very well die of a heart attack before it reaches them! :/
 
A very nice build...This seems like alot more value for $2500 than just buying the most expensive CPU and Graphics Cards. Well DOne.
 
you know for the top of the line models you guys should have gone with a core i7, its just your own reviews say its better for gaming which is kinda odd to me why you dident use one, and dont give me that budget reason you guys blew a bunch of money on ssd drives which yes help with boot time and load times but in terms of fps dont really make that big a difference. go back to the drawing board and learn.
 
How come there wasnt a direct comparison between this and last times $2500 system build? I would have liked to see the AMD Intel comparison. Also, why no 5870s?

Double also, according to your older article that benchmarked the bandwidth of the SB750 chipset 2x Intel SSDs would be overkill due to the fact that the chipset maxes out around 300MB/s. One reason I don't feel like adding another SSD to my system.

Good article.
 
[citation][nom]omegashiva[/nom]you know for the top of the line models you guys should have gone with a core i7, its just your own reviews say its better for gaming which is kinda odd to me why you dident use one, and dont give me that budget reason you guys blew a bunch of money on ssd drives which yes help with boot time and load times but in terms of fps dont really make that big a difference. go back to the drawing board and learn.[/citation]

Let me guess,you're one of the guys who demanded SSD's last time and are now looking for a fight? You could put a bag over your head and fight for air, but otherwise read the article to find reasons.

[citation][nom]raptor550[/nom]How come there wasnt a direct comparison between this and last times $2500 system build?[/citation]

The last $2500 build was a micro-mini system that couldn't be overclocked effectively and had to be undervolted just to run at stock speed.
 
Interesting build. Given the benchmarks lately coming out from the HD5870 I don't think I'd go multi-GPU. The place the 5870 really shines is in minimum framerates, it decimates all of the multi-GPU cards in that area and isn't far behind them in peak framerates. Makes it more playable and enjoyable in my opinion. After all, who's going to notice that your peak fps is 95 instead of 90, but everyone notices when it spikes down to 15 fps.
 
I suggest to re-purpose this SBM (or create a new category) not only to collect all possible subjective bells and whistles in a subjectively "cool" enclosure, but also provide end-user comfort by producing the lowest noise pollution as practical and within reach of an enthusiast (e.g. no liquid Helium cooling).

Swiftech kit is a good start, but it was used not to reduce the noise output, but to increase overclocking capability of the system. A the same time, three dual-slot graphics cards probably make the system sound like a 747 during take-off at 100 yards. That's just not acceptable when we're talking about $2500 build.

I argue that people who build $2500 systems look forward not only to raw performance, but also comfort of use.
 
i am surprised by several things in this build;
* three cards in crossfire instead of two slightly more powerful ones,
* 8gb ram instead of four (i thought thw only recommended 4gb - it's a pleasure seeing them recommend 8gb) -- and the ddr3-1333 instead of a higher speed (i realize their choice is a great o'clocker),
* a raid1 storage drive, I'd have probably choose a single drive mounted in the front of the case in a nice hot-swap type bay
* spending more $ on a psu than a cpu, and
* the high cpu cooler expense (not to mention being surprised they picked a h2o cooled solution)...

...all in all, i'm going to study the performance results for quite awhile to try to understand their choices and why they work so well.
 
actually it was my first ever post mostly because went i see something completely off the wall dim witted i feel it necessary to comment. now as for solid state hard drives yes i could see why a person would use one for loading there os for the faster boot time, this makes sense. but if your trying to build a gaming rig on a budget its a simple cost vs performance ratio. which i shouldn't need to tell you the ssd's dont hold up the cost would have been better spent on a core i7 or i5 even. its just by your own benchmarks amd is out classed for the moment though i like the rest of the build very much 3 way crossfire
 
[citation][nom]xizel[/nom]i thought this was an AMD build, why is 20% of the budget still going to INTEL?[/citation]

Because AMD divested its interest in SPANSION instead of developing its own line of high-end SSDs?

[citation][nom]bitterman0[/nom]I suggest to re-purpose this SBM (or create a new category) not only to collect all possible subjective bells and whistles in a subjectively "cool" enclosure, but also provide end-user comfort by producing the lowest noise pollution as practical and within reach of an enthusiast (e.g. no liquid Helium cooling).Swiftech kit is a good start, but it was used not to reduce the noise output, but to increase overclocking capability of the system. A the same time, three dual-slot graphics cards probably make the system sound like a 747 during take-off at 100 yards. That's just not acceptable when we're talking about $2500 build. I argue that people who build $2500 systems look forward not only to raw performance, but also comfort of use.[/citation]

The system runs cool at 50% system fans when at or near stock speed. Then it's quiet. Even the graphics cards are relatively quiet before being overclocked...at least until you get into a really tough gaming situation.
 
[citation][nom]omegashiva[/nom]you know for the top of the line models you guys should have gone with a core i7, its just your own reviews say its better for gaming which is kinda odd to me why you dident use one, and dont give me that budget reason you guys blew a bunch of money on ssd drives which yes help with boot time and load times but in terms of fps dont really make that big a difference. go back to the drawing board and learn.[/citation]
reading just isn't your strong suit is it? if you had read the article you'd know why there was no i5/i7.
 
[citation][nom]Shadow703793[/nom]FINALLY! You guys decide to pick up a good WCing kit.[/citation]

We've used the same kit in past SBMs

[citation][nom]Shadow703793[/nom]Btw, WHY did you get the Apogee Driver!?!? It's old (but better than the H50,etc cr@p)by todays WCing standard. Should have gotten this: http://www.jab-tech.com/Swiftech-H [...] -4324.htmlIt's a bit more expensive, but worth it imo.[/citation]

Easier installation, no case drilling required, etc.
 
[citation][nom]Crashman[/nom]The system runs cool at 50% system fans when at or near stock speed. Then it's quiet. Even the graphics cards are relatively quiet before being overclocked...at least until you get into a really tough gaming situation.[/citation]

With all due respect, this explanation sounds a bit... misleading. According to MSI HD 4890 Cyclone SOC 1 GB GDDR5 - Fan Noise, 38.4 dbA is the figure obtained for the MSI HD 4890 Cyclone SOC (should closely resemble what was used for the SBM) under load. While a single card does not produce enough noise to make it downright unbearable for everybody (I'd still argue that this is too noisy to begin with), three cards most certainly will.

Alternative cooling solutions should be considered.
 
[citation][nom]bitterman0[/nom]With all due respect, this explanation sounds a bit... misleading. According to MSI HD 4890 Cyclone SOC 1 GB GDDR5 - Fan Noise, 38.4 dbA is the figure obtained for the MSI HD 4890 Cyclone SOC (should closely resemble what was used for the SBM) under load. While a single card does not produce enough noise to make it downright unbearable for everybody (I'd still argue that this is too noisy to begin with), three cards most certainly will.Alternative cooling solutions should be considered.[/citation]
I've got to say I can't consider 50% fan speed near silent as crashman does either. At anything over 28% fan speed with the stock ati cooler setup it's audible, and it's growing in volume a lot faster than rotations increase.
 
I notice that in many reviews on this site some comment is made about anything over 50 avg FPS being indistinguishable from 50 fps avg. I have always thought that speeds up to 80-100 fps were noticeably different than lower FPS and that minimum was a more relevant measure since a steep drop, even if short, really stands out and can impact control if FPS gets into the low twenties. I find that when playing action games in first person perspective, I prefer to turn down AA and shadow res and get a frame rate that won't drop below 80 avg 45 min (70 min if I can get it is nice).
If you want to see the difference here is a simple test that I did (although normally I turn vsync off, I used it here to eliminate tearing as a source of visual artifacts).
I had Call of Duty: United Offensive (old game easy to get stable high fps) on a CRT set to 1600x1200 @100 hz, 4xAA, 16xAF, vsync on, all quality settings maxed. Using the in-game console, I set fps max to 110 and used FRAPS to confirm that the game was displaying at 100 fps. I then set the maxfps to 55 so that the game would be displayed at 50fps. When walking around the entire game world takes on a sight shudder. On busy textures this shudder becomes pronounced. Try strafing or slowly turning while looking at bricks or pebbled ground. I find this effect to be more noticeable at high resolutions and high texture resolutions unlike say aliasing which is less noticeable when other rendering settings are higher.
If some people don't see a difference between 50 and 100 fps then that is fine, but when reviewing products I think an enthusiast site should take those who do see the difference into account when making recommendations.
 
i think everybody could see the difference between 50 and 100 fps. the problem is (i'm guessing) 90% of us have LCDs that don't go above 60hz at the native resolution.

The thing that varies from person to person is probably their threshold to low fps. For me I can take 30fps easy. I would say that the best price/performance is right on this individual threshold. Would be interesting to see statistics on what people consider playable fps :)
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom