System Builder Marathon, Q2 2014: Our Budget Gaming PC

Page 3 - 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.

Understood. This would be something to work out in the lab before bringing it to the hard-budget space of the SBM.

A quick glance at Newegg prices right now:
750K - $80
760K - $90
A88X mboard - ~$90
8GB 2400 Team RAM - $70 (2133, 1866, and 1600 right around the same price. )
FX-4300 - $110
990FX mboard - ~$130

MBoard price can vary widely, I just picked a price on the higher side of middle to get a good board without going crazy. So assuming RAM prices don't go crazy again, and using stock cooling, that's an Athlon platform w/ speedy RAM for $250 while the FX will be $310 or more. If you can approach that 4300 performance for $60 less, that would be pretty interesting in the budget system segment.
 


Yet all of it is pretty irrelevant now that you can overclock chips on non z87/z97 chipsets like on the Asus boards and can grab a G3258 for 80 bucks since it's faster than all of them (at least if you want a budget gaming rig).

 

Sure, if you're content with a two-thread chip.
 


If it's faster, who cares how many cores it is, or any specs for that matter. It beat the 750k on every single game they tested it on. It uses less power too and it's arguably cheaper and can now be overclocked on non zXX chipsets. There is quite literally no reason to buy AMD currently for gaming rigs.

You could argue that if you were making a rendering machine or media transcoding box that the 8350 might be better than an i5 because it's cheaper and pretty quick at those tasks but other than this there is quite literally no other reason to buy AMD CPUs anymore.
 
Of course, the G3258 was not available when this quarter's SBM was planned. I would not be surprised if we see this chip in September though.
You might PM Damric for some of his findings on the 750K and memory access.
 

"Faster" in what terms? I'm not suggesting no one wants a high efficiency 2C/2T chip. However you're suggesting no budget gamer wants a 4T capable chip.

If you revisit that review, you'll see a few points where the G3258 may have posted the higher average framerate, but it also suffered a much lower minimum and noticeable lag spikes. We're finally hitting the point where quad-cores are becoming more relevant in gaming. Sure, most can get by lovely on two threads and probably will for the next year, maybe two. But you also have more than a few titles now that will benefit from four cores, and that number is growing.

If you're building a budget machine now and you're hoping to get three or more years out of it, you may want to stop and think about your core count. Sure, a lot of people are just fine with the G3258 and likely will be for years to come ( and they'll have the benefit that if the CPU really starts to drag, they can change in out for an i5 if they really need to. ) But for some others, going that route now may be short-term gain / long-term loss.

TL : DR - It's rather foolish to be making blanket statements saying one CPU is the best option for everyone in a given market.
 


Faster as in it wins in every single game they tested. In the frame variance I noticed you ignored all the ones that the poor little 2 core wins, and also forgot to mention that the times were so low that it's not really a problem in their tested config anyway. That poor little 2 core also wins in a few of the application tests that generally would go to the quad core like the Adobe products, and only loses a few of these. All while being slightly cheaper, and using less power and also being completely upgradable to an i3, i5, or i7 where the AMD is stuck where it is. It really makes no sense to do the 750k anymore, I don't know why you would ever defend it when the facts clearly speak for themselves. Hopefully AMD either severely cuts the price on their chips, or maybe comes out with something new that's worth purchasing again, it sure would be nice to see a good CPU price or performance war since the CPU scene has been pretty stagnant for a few years.
 

jlwtech

Honorable
Mar 8, 2012
58
0
10,630


Paul, thanks for the detailed response.
I didn't know the parts were ordered that long ago. I was thinking 2 weeks at most, but clearly, I assumed wrong.
So much can happen in 6 weeks.... ( I recently saw $100 price drop in open-box r9-290's, over a 3-4 week span. Just got the Sapphire Tri-X OC 290 for $304, 3 days ago!)

Is there any way you can shorten the time between ordering parts and posting the review?
Or, maybe do the SBM in 2 pieces? (1st part of SBM lists the parts/prices, on the day the parts are ordered. 2nd part is the review, 6 weeks later. Maybe even do a little reader voting....)
 

mitcoes16

Distinguished
Jul 6, 2007
21
0
18,510
I missed some comments I was able to read before I did log in and being 60 i only was able to read about 30 not working the more comments button
 

mapesdhs

Distinguished


You can switch to the UK site instead for an easier interface, or toggle the oldest vs. newest first
to prioritise what you want to see.

Ian.

 

emmveegee

Honorable
Apr 30, 2012
13
0
10,510
Thank you for integrating the Arma 3 feedback into your tests. Doing quality research is something not every site can boast, but you're definitely on the ball. Keep up the good work!
 

pauldh

Illustrious

Unfortunately, probably not. It's far more involved than you may imagine. Sure, each of us acting on our own could pull off his own build/story in two weeks. But as a group, and in a series, it take far longer. Once it falls in our laps (the builders), we must decide on a theme and budgets, and get the parts ordered. Schedules aren't in sync either. Nobody is sitting idle; all must finish up whatever other work is already on the table. Some of that may be very time sensitive, too. We can't start testing if something has our bench/priority. Finally, once we have a realistic ETA for all individual stories, the series needs to claim a week on the editorial calendar to publish in succession. First they need to go through edits, too. So in the end, at times it can become a group-crunch to pull off.
 

pauldh

Illustrious
I need to sign out for a while, but will take any further comments into consideration for next quarter.

Regarding this Athlon 750K, it certainly has room to go above 4.2 GHz. Properly cooled, I wouldn't be surprised to hear the lucky winner tunes stability beyond 4.4 GHz. I've encountered the throttling issues mentioned with my (lab's) 750K. They were solved over time, but the chip still topped out at 4.3 GHz at the maximum voltage I cared to provide. Interestingly, Chris hit the same fully-stable 4.3 GHz limit in that G3258 review, with a different 750K, on a different test platform (and mobo model).

To add to the Pentium discussion, I'll definitely consider one next time. From what I have seen fully tuned G3258 can beat 4.5 GHz 760K in BF4 MP. Out of a huge sampling of heavy hitters, the 760K only claimed honors in Watch Dogs, where it's performance was still sub par anyway. Watch Dogs is tough to touch on a budget.

Thanks for the feedback and discussions!

 

This very definitely bodes well for the Q3 SBM; I sense another excellent data point on the way!
 

Thanks for that, Tourist, so I didn't have to write it. It's as though he just wants to argue with me on general principle this last week. I too find the new Pentium very compelling. I'm sure many a budget tweaker will have a lot of fun with it. I also would like to see AMD challenge in the performance segment again, but they've publicly declared they're not touching that for the foreseeable future.

But, Traciatim, making broad, sweeping generalizations and applying them to every imaginable contingency is completely asinine.
 
I also would like to see AMD challenge in the performance segment again, but they've publicly declared they're not touching that for the foreseeable future.

But, Traciatim, making broad, sweeping generalizations and applying them to every imaginable contingency is completely asinine.

Got a link to this public declaration from AMD ?
 


Where in the link you provided does it say AMD have no plans to challenge in the performance processor market?
I cant see it .

And while Im sure performance processors are not their focus ATM I dont think they will abandon the top end of the market
 

Oh, for the love of . . .

If you want an explicit quote saying, "No, we're abandoning the desktop market," you won't find it. They didn't say they're giving up on desktop, only that their focus was going to be on mobile devices ( though not smartphones, ) low-power SoC, and heterogeneous computing. If your primary focus is in those groups, it means it is NOT in high-performance desktop processing. Now, it is possible that HSA could overlap that high-performance sector if it got enough software support so that an APU would perform similar to the i7 through superior parallelization ( which is something I wouldn't mind seeing. ) But right now, that's not the case and it doesn't look like it's going to be in the immediate future.

Here's a direct quote from Mike Silverman, an official AMD spokesman: "we will all need to let go of the old 'AMD versus Intel' mindset, because it won't be about that anymore." My translation: they're not trying to directly compete against Intel's strengths anymore ( at least for the next few years. )

When was the last FX architecture released? Late 2012, after that statement I mentioned was given. AMD then decided against Steamroller FX chips this year. So since that announcement was given, the only new CPUs we've seen from AMD are APUs and APU-based Athlon chips. When is the FX line going to see an update? Right now it's looking like that won't happen until late 2015 with Excavator. Now, maybe Excavator gets AMD's IPC closer to Intel's. That would be great. But by the time that happens, it will have been three years since a new FX architecture was released. Three years. Hardly sounds like they're trying trying to compete in that market.
 
So just to summarize : AMD never said they werent going to make performance processors anymore so you were incorrect , BUT since they havent updated FX they probably should have said that and even though you were completely wrong you were actually right?
 

Nice try, but I never claimed they were completely abandoning that market segment. But please, feel free to keep lying about it and putting words in my mouth. It makes you look dumber each time you repeat it.



Nor did I claim or say, "What AMD really meant was..." I only offered my own take and opinion on their public statements. I don't even know where you're trying to pull that from. Actually, I probably do, but I try to keep my comments kid friendly, so I won't go there.

So, just to summarize: I never said what you claim I did, so you were incorrect. BUT since you haven't updated your lies, and even though you're completely wrong and lying, you want to say you're right?
 
Status
Not open for further replies.