System Builder Marathon, June 2012: $500 Gaming PC

Page 5 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

pauldh

Illustrious

1) If/when we do mini-ITX, it will be a team-wide trio representing that niche at each price point. I bounced the idea off the team 6 months ago, and none were opposed. If readers got vocal sharing the desire, it would likely happen sooner rather than later.

2) It's hard not to find some fault with sub-$40 cases, but I've been reasonably satisfied with almost every single Rosewill I've ever touched. When funds are limited between case and PSU, you and I both know which deserves attention first.

3) I doubt you'd see me go multi-GPU in an SBM unless given a far higher budget. I can't explain why here and now, but hopefully in the next few weeks. I've been on a long term project that will expose another "killer" besides microstutter or games that do not scale. Data freaks (like myself) should like this monster once live.

Thanks, as always, for your constructive comments.
 

pauldh

Illustrious


:non:

Your build would be a TOTAL FAIL at this system's goals and be a embarrassment agaisnt the $650 at the target gaming resolution. Did you read the article, or just post?

No upgrade path? Are you kidding me? It's ready to pop in a Core i5 or i7 right now, as build!

Plus, After many hundreds of hours with Crossfire on my test bench, pairing it with Llano would be VERY low on my upgrade list.

Besides that, Llano is one of many considerations for next round based on the feedback above. PLEASE, keep the comments civil and constructive.
 

pauldh

Illustrious


First, Thanks for taking the time to discuss. Since you are not alone in wanting to see a Llano build explored, the possibility grows. What motherboard are you using to hit 3.6 and 3.9 GHz?

The problem right now is, playing by the rules I must follow (no combos, rebates), your config can't be done for $500. Dropping to 4GB of DDR3-1866 gives us about $35 for a mobo. No need to go nuts just yet on pricing, as it will be a month or more before that becomes a consideration. But, with 4GB it's looking possible.

It however is not priced up agaisnt a Celeron by any means, which is why I started our return to $500 with the lowest CPU I'd consider using. Moving forward from here I expected more CPU budget, and more balance. Remember, I get one shot at this each quarter, and had to decide with $150 less budget, how would I like to compete against the former build.

Your config in fact the same exact cost as Core i3-2100 + 4GB DDR3-1333, which could then be paired with HD6870. I see reasons for both, but based on our own data, Llano is a hard sell "for gaming". See, I'd like to build BOTH, but it's not possible. Same, I would like to have build a G850 + HD 6870 or GTX 560 build to compete along-side Junes G530 rig.
 
[citation][nom]pauldh[/nom]Put it this way, it is not ideal. Nor is Celeron of course.(Not saying they are equal situations) but remember Don's overclocked Crossfire/FX build how how my i5-2400 + HD 6870 often trumped it at 1920x1080?http://www.tomshardware.com/review [...] 099-5.html[/citation]

FX is worse than Phenom II and the CPU was not the only thing wrong with that build. Even with a 6100, it should have done better than it did. The 6100 would be a bottle-neck and a bad one at that, but not quite as bad as that build had and a Phenom II would be less of a bottle-neck. Overclocked to 4GHz, a Phenom II x4 should be able to handle up to about a 7850 or 6970 without being much of a bottle-neck, if a bottle-neck at all.
 


That could have been part of the problem, but I didn't trust some of those other components either. Catalyst 12.6 is supposed to be AMD's best driver for Crossfire right now and from what I've seen of it, it succeeded in that goal.
 

pauldh

Illustrious

Thanks, will look them over.

Obviosuly to pull this off we'd need to spend as little as possible on the mobo, while retaining acceptable quality and assuring OC potential. We rarely, have budget boards sent for review, but I'd love it if Thomas could take a wack at a few for each platform. It would make things far easier for me. :D

Although I've had tons of fun overclocking AMD, and have written a couple Athlon, Phenom, and Phenom II based overclocking guides, I still havent' had the opportunity to touch Llano yet. My bench is always full, and the time non-existent. Must admit though, early talks of broken USB 3.0 and AHCI, plus the final performance level, left me a bit concerned whether it was all worth it. The SBM could present the perfect opportunity.
 

pauldh

Illustrious

Keep in mind, we use the latest drivers available at the time we start testing each build. You have no idea how frustrating driver consistency has been for my own Crossfire testing. Then add game patch improvements also. Good for gamers, bad for reviewers. Just look how many reruns we've done in Skyrim since release.
 


I admit that even for the possible advantages that a Llano $500 SBM could have, as you mentioned, there sure have been bad reports of motherboard quality problems when it comes to the connectivity and overall stability. It had me not recommending Llano to several people looking for a low end build for minor office work back when Llano was in its early days (granted, that's not long ago at all). I think that at least for the most part, this has improved if you use newer motherboards, but I admit it's been a while since I played with a Llano computer.

I do often enjoy playing with some AMD builds right now because even though they aren't always better than Intel (and usually aren't for most purposes), they do seem to simply be more fun to mess with. For example, for all it's fallacies, I wouldn't trade my Phenom II x6 machine for anything on the Intel side short of an i7 build of much greater quality to at least make the performance different in highly threaded work justify the change.

I'll look into Llano boards a lot more for a while and see if I can come up with anything helpful. While I do that (and I'm assuming that several others will probably be looking at them, maybe yourself included), would you mind spending a minute to comment about the idea I had with the FX-8120? I haven't had an FX build to work with since around early March or late February (can't really recommend FX 8 core CPUs to much of any customers, so they're not a common thing to work with) and I can't really confirm it myself, but there seems to be more than enough proof to say that the FX-8xxx CPUs and maybe the FX-6xxx CPUs too can be useful for gaming (even with their prices) if you make sure that you either disable one core from each module (I don't think that many boards can disable specific cores, but I know that some can and I think that one such board was an Asus, but I don't remember the model number) because this not only leaves the entire module's resources to a single core, but also means that Turbo can be more aggressive (sorry I forgot to mention that earlier).



Understandable, completely understandable. Well, if AMD follows through on their plans for their driver release strategy, maybe drivers won't cause needs for re-testing so often. Also, I wanted to thank you for spending so much of your time here... This is easily one of the most active and helpful conversations with Tom's and it's nice to have someone who can actually do something about these questions listen and respond so well and quickly.
 


I admit that I haven't tried Trifire with any expensive cards and not with any Radeon 7000 cards yet (might try 7750 or 7770 Trifire if I can get it running sometime), but I didn't have driver trouble with Trifire 5770 cards and 6770 2GB cards. I've heard good things about 7970 Trifire from Hard OCP and although I have seen poor reviews of 6870 Trifire and heard a few bad things about 6970 Trifire when they were newer, I don't think that Trifire support is universally bad.
 
[citation][nom]pauldh[/nom]1) If/when we do mini-ITX, it will be a team-wide trio representing that niche at each price point. I bounced the idea off the team 6 months ago, and none were opposed. If readers got vocal sharing the desire, it would likely happen sooner rather than later. 2) It's hard not to find some fault with sub-$40 cases, but I've been reasonably satisfied with almost every single Rosewill I've ever touched. When funds are limited between case and PSU, you and I both know which deserves attention first. 3) I doubt you'd see me go multi-GPU in an SBM unless given a far higher budget. I can't explain why here and now, but hopefully in the next few weeks. I've been on a long term project that will expose another "killer" besides microstutter or games that do not scale. Data freaks (like myself) should like this monster once live.Thanks, as always, for your constructive comments.[/citation]
1. Hopefully this will happen. Pure gamers will probably hate the thermal limitations at the extreme high end, but I think some remarkably capable systems are easily within reach in this form factor.
2. Same. I've built in 6-8 different Rosewill cases, and NONE of them had any flaws worth mentioning. And yes, I'll admit to being a bit of a PSU snob when it comes to quality (but definitely not size).
3. Careful please, I don't want to start drooling. That article is going to be a doozy, I can tell.
---
Anyway, the entire discussion around this build has been very interesting. I think this price point is definitely the most informative. Exposing bottlenecks and other limitations is always useful.
 

demonhorde665

Distinguished
Jul 13, 2008
1,492
0
19,280
[citation][nom]pauldh[/nom]Your build would be a TOTAL FAIL at this system's goals and be a embarrassment agaisnt the $650 at the target gaming resolution. Did you read the article, or just post?No upgrade path? Are you kidding me? It's ready to pop in a Core i5 or i7 right now, as build!Plus, After many hundreds of hours with Crossfire on my test bench, pairing it with Llano would be VERY low on my upgrade list.Besides that, Llano is one of many considerations for next round based on the feedback above. PLEASE, keep the comments civil and constructive.[/citation]

being able to upgrade jsut teh processor is a big differnece from having suport for two vid card dnd two extra ram slots , sure teh a8 APu is not teh msot powerful cpu , but that main board on their build has some severe limitations (only 2 ram slots and only 1 pci-e slot So yet there system is far less upgradable
 

demonhorde665

Distinguished
Jul 13, 2008
1,492
0
19,280
also keep in mind i'm loking at waht i can afford to upgrade , as it stands , i surely can't afford to upgrade ,to an i7 at 250-500 bucks , however , i can afford 30-40 bucks on ram , and 120 , on a second vid card
 
[citation][nom]demonhorde665[/nom]being able to upgrade jsut teh processor is a big differnece from having suport for two vid card dnd two extra ram slots , sure teh a8 APu is not teh msot powerful cpu , but that main board on their build has some severe limitations (only 2 ram slots and only 1 pci-e slot So yet there system is far less upgradable[/citation]

For gaming, a Sandy/Ivy Bridge i5 and i7 are pretty much identical, so knock that down to $180 to $230 for a CPU upgrade or even down to an i3 at $120 to $140 and then sell the Celeron to make up some of the cost of upgrade. On the video card, just replace it and sell the original if you want to upgrade. Sure, it's not as good of a solution as having CF/SLI support, but it's something. Having two RAM slots means that a 2x4GB kit can give 8GB of RAM and a 2x8GB kit can give 16GB of RAM (8GB modules have come a long way in pricing and can be considerable in some cases without paying too much).

If you meant installing a second kit into the 2 spare RAM slots of a 4 slot board, well mixing RAM kits isn't recommended anyway and again, you can sell the current modules and replace them. In all places where I said sell the old parts, putting them in a different computer is obviously also an option.

This price point has the computer being built to perform as well as it can, within reason. Supreme upgrading capacity comes in second place due to that IMO.
 
[citation][nom]tourist[/nom]blaze, thanks for the ccc tip. 12.6 is fantastic !!!!!!!It also officially proved tri-fire works on liano. now ccc says i have three video cards in cross fire. 6530d +6570+6570 No benchmarks but i will test and let you know..Thanks[/citation]

You seriously have trifire working with a Llano IGP? That's something to look into. You're welcome.
 
[citation][nom]tourist[/nom]Yes gpu-id shows all three as well as ccc says crossfire is enabled. I was told it could not be done but when do i ever listen to naysayers. M/b is GIGABYTE GA-A75M-D2H FM1 AMD A75 (Hudson D3) SATA 6Gb/s USB 3.0 HDMI Micro ATX AMD Motherboard and i am assuming it is only running at 4X4 but it is working.This might be a game changer on a 16x16 board[/citation]

Excellent.
 
I like that you stuck to the $500 build requirements!!! First time in a long that Tom's did a true $500 SBM build!! I would have changed a few things, but sticking to the $500 build requirements, REQUIRES some extra thought into what you spend on. With the $650 builds, it's much easier to put your parts in and there isn't quite as much thought needed to put a system together at $650.

Here's my thoughts on the next $500 SBM:
* Go with a bit better CPU (whether it's an AMD or Intel based CPU).
* Don't use a limited use mobo (spend about $75 or so on a decent budget mobo). Intel has the z75 chipset based mobo's that would be a good start for an Intel based system.
* I'd like to see 8 gb's of DDR3, but if you need to squeeze out ~$20 somewhere than 4 gb's will do.
* The PSU selected before is good enough for this price range, so any quality 350-400w PSU will do from Corsair/Antec/Seasonic/PCP&C.
* Personally think that a little more equal CPU/GPU system makes more sense for the overall experience and longevity of the system.

All in all I like the boldness in putting this system together and sticking to the $500 budget! Keep the budget at this price range and you will have plenty of positive discussions (as has been noted on this article) that hasn't been seen for quite some time (in my recollection).
 

pauldh

Illustrious
@ - lunyone - Thanks for the feedback.

Yes, dropping back to $500, I was not going to break budget on the first build. Not a chance. I had that right, but personally set a hard cap this time.

Remember, it was the readers who first suggested we be more flexible, and break budget a bit if it makes sense. That is why so many of the $500 PC's in 2011 ended up in the $513-527 range. I see so many options right now near $500 worthy of exploring, yet would hate to cripple any of them over $10-15 of cheapness.

I could very well get just one SBM look (at the most) at FX-4100, Llano, Pentium 850, etc. So, what is more important, 1) staying under $500 firm, or 2) a build that best allows the platform to shine or can best measures up to the prior efforts? Since the rules allow for breaking budget by a justifiable margin, staying at/below $500 every time is a hard sell.

Also, consider prices change, and my $500 build as built could be $515 when the series goes live. Or a $510 build could be under budget. The ONLY firm price we have to go by is what our build cost us at the time. All other price adjustments can/will change almosy daily. But factor in combos, likely a very similar build would be within budget anyway. So as much as I love a hard budget cap, I need to side with most readers who favor going slightly over to build a better machine. Of course, it’s a slippery slope... drawing the line isn't easy.
 
Going over budget can tend to favor one camp over the other, not that I care which one wins, but this can change things dramatically. I'm not worried about meeting/exceeding previous builds performance for the same budget. I would just like to stick to a firm budget (at ordering time), including any promo codes or even combo deals. This is what I and probably a lot of other builders would do, to save the most and get the most for our $.
 

pauldh

Illustrious
I’m out of time, but so many comments deserve attention.

@ Jtt283 & RedJaron – Feedback noted. Thank you.

@ Blazorthon – Thanks, appreciate any mobo research you are willing to do. Come next SBM, I’ll have a limited time to decide.

Your FX-8120 core/module question is totally worthy of a deeper look, but I’m not sure who among us has the time to pull it off any time soon. I’m committed to other projects outside the SBM, and not sure I could pull off a worthy FX-8120 build at $500. Also a problem, in an SBM all overclocked tests must run the same settings. If we disable cores for games, I’d need to leave them disabled for Apps too. From what I have read (not tested), disabling 4 cores … leaving 4 cores on 4 modules, does outperform FX 4100 (4 cores on 2 modules). I understand frustrations of games/OS not taking advantage of the architectures threading abilities, but otherwise feel overclocked FX can game right now as is. I’d lean towards single-GPU configs of any size, and while low resolution bottlenecks will exist, it more about the experience at the intended resolution/details. But I personally own and have tested so many Phenom II’s that I also have to admit FX didn’t impress me enough to warrant further attention. I’d want to tinker on ALL hardware (and share data) given the time/opportunity, but that’s far from reality.


Sorry, but I need to check out of this thread for a while. Work to do. Thanks ALL for the discussions.
 
[citation][nom]pauldh[/nom]@ - lunyone - Thanks for the feedback. Yes, dropping back to $500, I was not going to break budget on the first build. Not a chance. I had that right, but personally set a hard cap this time.Remember, it was the readers who first suggested we be more flexible, and break budget a bit if it makes sense. That is why so many of the $500 PC's in 2011 ended up in the $513-527 range. I see so many options right now near $500 worthy of exploring, yet would hate to cripple any of them over $10-15 of cheapness. I could very well get just one SBM look (at the most) at FX-4100, Llano, Pentium 850, etc. So, what is more important, 1) staying under $500 firm, or 2) a build that best allows the platform to shine or can best measures up to the prior efforts? Since the rules allow for breaking budget by a justifiable margin, staying at/below $500 every time is a hard sell.Also, consider prices change, and my $500 build as built could be $515 when the series goes live. Or a $510 build could be under budget. The ONLY firm price we have to go by is what our build cost us at the time. All other price adjustments can/will change almosy daily. But factor in combos, likely a very similar build would be within budget anyway. So as much as I love a hard budget cap, I need to side with most readers who favor going slightly over to build a better machine. Of course, it’s a slippery slope... drawing the line isn't easy.[/citation]

Most people making a build at this budget would go a little over if they had to to get a much better machine in some ways than a slightly cheaper one. I'd draw the line at one of three places :$515, $520, or $525, just to get a nice number to look for. Doing more than $525 would be a greater than 5% increase over budget and I would prefer that we not be that lenient with how far we stretch budgets.

Like you said, it's a slippery slope and with a $500 budget, sometimes you need to stop and think about just how much of an increase a price jump really is because for the $500 machine, every $1 increase, by percentage, is twice as great as a $1K machine and four times as great as a $2K machine. When we realize that going like $30 over would be like going $120 over in the $2K machine, I think we can more clearly see that it's pretty much unacceptable, unless there was some huge gain for doing it.
 
[citation][nom]pauldh[/nom]I’m out of time, but so many comments deserve attention.@ Jtt283 & RedJaron – Feedback noted. Thank you. @ Blazorthon – Thanks, appreciate any mobo research you are willing to do. Come next SBM, I’ll have a limited time to decide. Your FX-8120 core/module question is totally worthy of a deeper look, but I’m not sure who among us has the time to pull it off any time soon. I’m committed to other projects outside the SBM, and not sure I could pull off a worthy FX-8120 build at $500. Also a problem, in an SBM all overclocked tests must run the same settings. If we disable cores for games, I’d need to leave them disabled for Apps too. From what I have read (not tested), disabling 4 cores … leaving 4 cores on 4 modules, does outperform FX 4100 (4 cores on 2 modules). I understand frustrations of games/OS not taking advantage of the architectures threading abilities, but otherwise feel overclocked FX can game right now as is. I’d lean towards single-GPU configs of any size, and while low resolution bottlenecks will exist, it more about the experience at the intended resolution/details. But I personally own and have tested so many Phenom II’s that I also have to admit FX didn’t impress me enough to warrant further attention. I’d want to tinker on ALL hardware (and share data) given the time/opportunity, but that’s far from reality.Sorry, but I need to check out of this thread for a while. Work to do. Thanks ALL for the discussions.[/citation]

I was thinking that the FX build I suggested be done with the $1K budget and also test out the feasibility of Radeon 7770 2GB CF, not be in a $500 build. I also own a Phenom II x6 and just couldn't trade it for an FX at this time, but it's not in a gaming computer, it's my VM hosting machine.
 

calguyhunk

Distinguished
Jul 6, 2010
1,456
0
19,460
If somebody on the streets had told me that a 560ti will be not just matched but inded bottlenecked by a 'Celeron' I would have prolly just laughed him off as an idiot had it not been for the facts and figures presented by Paul ;)

Un-freakin'-believable . To think then what dual cores can still do in a biga$$ PC in this day and age when they're making quad-core cell-phones :O
 
Status
Not open for further replies.