Best Gaming CPUs For The Money: January 2012 (Archive)

Page 48 - 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.
It's a pretty clear consensus that the "Best CPUs for the Money" format needs to change. Much more helpful, I think, and one that would spark more intellectual dialogue rather than the typical AMD vs. Intel rant, would be a format that highlights each of the major CPUs (there aren't that many), and why you would purchase that CPU. Each of the CPUs have their strengths & weaknesses. What are they, and in what use-case scenarios would each one give you the best Value for your objectives.

This format will also allow you to bring into account platform decisions that fold very much into the overall system value and price. Motherboard options, whether or not you'll need an aftermarket cooler, memory required, etc all play into the cost and performance of the core system.
 


Well this would be true if you hadn't misread the title.
Its: Best Gaming CPUs For The Money
So it's pretty clear that the factor for this list is: Gaming.
And for this kind of use, the list is solid.
 


Actually, it could use some work. FX 8320 belongs on the list. The i5 3350p should drop off and make room for The i5 4430. At $185, it is close enough for the $180 mark. The price is close enough that it would be silly to choose Ivy over Haswell. Only ivy chips I would consider now are a 3570k if it is a really good deal, and probably a 1230 v2 for the non overclocking crowd that wants i7 performance.
 


Well, I have been following the article pretty closely for about a year and a half, so no, I didn't misread the title. Let's just debate the merits.

Yes, the focus is "Gaming". So even with Gaming, the list is not that solid. The best choices are going to be more nuanced. Just having what is the "Best" is actually rather shallow. A better question is what CPU is going to give you the best value for your style of gaming. And that question works well at both ends of the spectrum.

I would categorize styles and levels of gaming and then go through each of the current CPU options (not EVERY CPU, but every group (i.e. Pentium G, I3, I5, I7, I7E, FX4, FX6, FX8, A Series, Athlon). Then you have styles of play and your goals. In general, you have older titles or MMOs that will perform better with single-threaded performance, but the latest big titles optimize better for multi-threaded. Then you have the casual gamer who plays most of the popular Steam games with one or two major titles. The choice of processor would be different for each of these styles of play.

As far as capabilities, you have enty level 1080p (or sub-1080p if you are entry-level playing the latest titles). You have full 1080p capability on all titles. You have 1440p, 5760x1080, and now 4k. So what are you going to need at a minimum at each level.

This actually helps out at both ends of the spectrum. For an entry level rig, looking at styles of play as well as desired form factor would tell you whether you want a Pentium G, 750 or 760k, or even now a Kaveri based rig.

In the mid-range it will give you some advice around when to choose an I-3 or FX-6300, or whether you will need to step up to even the most budget friendly I5.

At the high range, Tom's can finally give credit to the GAMING use case scenarios where having a 2011 based platform will provide additional benefits.

And finally, Tom's hardware readers do more than just "Game". So, even though the title of the article is "Best GAMING CPUs for the Money", I'm suggesting that perhaps that needs to change as well.

If you fold in some of the activities that Tom's Hardware readers are also likely to want to accomplish with their PCs, then that would add an additional dimension to the article that is currently severely lacking.

Of course there are reasons to keep it the same. I'm just suggesting the format is getting a little stale and that the CPU decision is more nuanced than identifying a single "BEST" based on average FPS among 3 stale titles. And keeping up with the comments and dialogue on this forum, I would suggest I'm not the only one who thinks there are serious flaws with this format.
 


I agree with these suggestions. Apart from one.
Well since hyper threading doesn't help a lot at gaming I think being able to overclock can provide more performance. That's why they suggest 4670K.
Also not only 4430 but I can usually find 4440 around $180, most of the times cheaper than 4430 (dunno why...). So with 4440/4430, the 3350p should be replaced...
 

Many models have exactly the same MSRP from Intel's so the different prices on the market are little more than an effect of demand and availability. The 4430 is priced the exact same $182 as the 4440 so the 4440 may very well be cheaper simply because most distributors and retailers are only bothering to stock the 4440 to spare themselves the trouble, cost and risk of warehousing the 4430.

You see the same thing occur all over Intel's product lines... 100MHz increments are just much too fine to make sense at 3-4 GHz and this is even more ridiculous when many of those models are priced exactly the same at the source.
 


41xx, not good for gaming, and the motherboard they give for free is of the lowest quality. youd be better off with the pentium g chip they offer for $50, and getting a motherboard that you can eventually put an i5/i7 in.
 


The 4130 is about on par with a PhII 965, so in multithreaded titles, it might just pull ahead of a Pentium G. The Pentium G was removed from the best of list a long time ago for this fact. Before the 750k came around and the supply dwindled, Athlon II X4 was given the nod as a best of over the Pentium G. FX 4130 actually was on the list for a little while. It was bumped off this month probably due to the low cost of the FX 6300.

FX 6300 and the Gigabyte GA-78LMT-USB3 would not be a horrible budget combo. I know many don't like that board, but it definitely a better board than some of the cheap boards that don't even bother with heatsinks for the VRM's.
 


Um... no... Bulldozer was unquestionably slower then the Phenom II... almost 25% slower per core/clock. Sure they clocked up better, but with a 25% deficit it needed to be clocked a LOT faster then a PhII x4. A stock 965be (3.4ghz) was as fast as a bulldozer fx4130 at 4.25ghz... unless you had a c2 PHii you'd hit 3.8ghz on your overclock without trying, which meant the bulldozer would have to clock up to 4.8ghz to match it.

There is a pretty good reason why people called bulldozer "faildozer".

Piledriver on the otherhand closed that IPC gap pretty significantly, and even overclocked better then bulldozer... so while on a clock per clock and core per core basis PhII tended to be a little faster then piledriver, piledrivers higher average achievable clock speeds usually made it walk away from the phII in most benches.

 
THG, in their previous best of, said the 4130 was on par with a PhII 965 at stock. Clock for clock, yes it is definitely slower. We are talking stock clocks here, nothing more. Out of the box running their stock rates, the two are about even. I agree with the Faildozer sentiment on here. I was so disappointed in it that I bought a 3570k when my i5 2400 went to my file server.
 
I have found the below linked articles MUCH more helpful than this monthly "Best Gaming CPUs for the Money" article. They are a great reference. I just wish they would have joined the two analysis for each game, instead of just at the end. It's good to bring them up in two side-by-side tabs and step through both.

http://www.tomshardware.com/reviews/ivy-bridge-wolfdale-yorkfield-comparison,3487.html
http://www.tomshardware.com/reviews/piledriver-k10-cpu-overclocking,3584.html

According to the testing in the 2nd article, FX beats Phenom II in general for gaming.
 


Excellent advice! My closest Micro Center is well over 400 miles away. Do you offer free pickup and delivery services?

I guess this is only an option for those select few who live with 50 miles of one of the 23 Micro Center retail locations in the world (all of which are in the US and require that you live near roughly 10 greater metropolitan areas).

There is a reason Micro Center pricing is never used in reviews and general recommendations...
 
Microcenter takes itself out of competition for just that reason. No shipping does more harm when you're trying to reach potential buyers. As much as I'd like to pay them a visit, I still have to justify going 2 hrs out of my way for any of their deals. The money spent in gas could have been used to buy from New Egg, whom does deliver to my location.
 


Plus you have to figure Microcenters are usually located in places with high sales taxes.

I actually use them for CPUs, where the savings are usually $30 or more. My sister lives 5 mins away from one. I just have her buy the CPU, un-box it, fold the cpu into a couple sheets of paper, slip it into a 1st class envelope, and mail it (yes, snail mail). I haven't had one damaged yet.

Perhaps there is something in setting up a Microcenter buying service..... Right...
 


Actually Microcenter has 3.5% tax. at least near my house. It says it right on the store itself.
 


Thanks, I regretted that last one right after I submitted it...

I had no idea MSI had started making stuff that was low quality, especially to the point that it is worth avoiding. That's a really sweet spreadsheet. Is there a similar one for Intel socket 1155 and/or 1150 boards?
 
I have not seen one for Intel boards, but I'd probably assume similar component quality at the same level of budget. I do need to point out that high-end MSI boards can be quite nice. I got to review their Z77A-GD65 Gaming board, and it was sweet. I purposefully aimed an IR thermometer at the VRMs, and I believe 38C was as high as I saw.
 
Intel boards generally aren't as much trouble, because non-Z chipsets only need to push 77W at most (no OCing), and Z-series boards still have lower power chips, even when OCed.

LGA2011 is a different thing, but you tend to have a higher cost of entry; there's no cheap boards.
 
That makes sense. I have blown a VRM on a decent Intel board though, but not by overclocking. It popped at stock during a gaming session. Either it was a fluke, of the presence of a HD7870 nearby didn't let it cool off enough, despite airflow through the case.
 
Wow that's surprising! I can't imagine a 7870 putting out enough heat to pop a VRM... but stranger things have happened. My first guess would be that you somehow got a lemon. I'm also interested to hear what make/model board that was.
 
Status
Not open for further replies.