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

Page 4 - 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.


Nice response, and I tend to agree. However, you've hit on what I think is a weakness in this article and much of the general analysis on these gaming oriented sites. Increasingly, at least from my own experience, people are multi-tasking while gaming. My 3 boys have all asked for 2nd monitors not for multi-monitor gaming, but because they always want something else going on while they're gaming. They may have a youtube video playthru on the game they're playing streaming or a video from our plex library playing. They also almost ALWAYS have a google hangout or skype group chat going among themselves, their cousins, and friends. There have been times when I've walked in that they have all of the above going on. They haven't gotten into recording yet, but my youngest really wants to be a "Youtuber" when he grows up and so it's only a matter of time.

I think this type of activity is going to be more of the norm if it's not already. The problem is that we don't have answers on how these activities impact gaming performance for different CPUs. It would be nice for Tom's to start doing some benchmarking around common activities WHILE gaming.

Finally, there are a number of activities that gamers DO in the context of gaming that detract from the overall experience if the process takes a long time. Screen loading, installing programs or large mods, zipping large file structures for backup, etc. While Tom's does show a broad array of benchmarks, they are not really put in the context of impact on the overall PC experience. So, one CPU may get you ~5-10 more FPS in a game, but it also may take 5-10 minutes (or 20-30?) longer to zip a full game archive if you need to do that.

Bottom line is (1) benchmarking multi-tasking activities' impact on gaming performance and (2) putting some context around real impact of other computing activities and their impact on the overall gaming experience are areas that would add a lot of valuable information for users designing and building a system. Perhaps it wouldn't change the recommendations at all, but as of now it's probably a question that begs to be answered.
 
The memory support info for the i7-5930K hasn't been updated since it replaced the Ivy Bridge offering. You list it as having DDR3 support, not DDR4.
 
I am talking about gaming/real-time workload performance. The CPU ordering is certainly NOT perfectly OK, there are some glaring flaws.

An FX-9590 is faster than an i7-975 in these workloads, why is the 975 in the top tier while the 9590 is in the same tier with an FX-4350 (which, FYI: has basically the same performance in real time workloads as the i5-2300 found in the tier above).
Click to expand...

The FX-9590 sells without a stock cooler and should automatically be disregarded from this chart by default. It's very presence on the chart is a testament to Don's lack of bias.

 
Another month of copy and paste. Xeon 1231v3 still doesn't receive the honorable mention it deserves. For those gamers, that do not want to overclock, but stream and record their gameplay, it is a great chip for the $$$.
 


Don's "lack of bias" (up for debate) can not be demonstrated by the inclusion of the FX-9590 despite it's lack of an OEM cooler in the chart as long as the chart also includes Intel parts that also do not ship with cooling solutions either.

Boxed Intel 2011 socket CPUs do not include HSFs. That's been the case since Sandy Bridge. Lots of enterprise and enthusiast class CPUs over the years have NOT included cooling solutions, this is pretty normal. The enormous impracticality of the FX-9590 should NOT be somehow represented in a chart that attempts to categorize performance classes.

Heatsink or not, the chart is blatantly wrong in many areas. The FX-9590 is up to 25% faster in real-time workloads than the Nahalem architecture i7's represented in the top tier. I wouldn't recommend an FX-9590 to anyone, it's a factory overclocked gimmick and novelty. This doesn't change the fact that it is a stronger gaming CPU than some of those CPUs found a tier above it in the chart. If Tom's hardware wants to get serious about cleaning up this mess I'd be happy to help in my spare time.
 


 


Perhaps you should look at the Passmark rankings. At least they are easily available.
 
In my experience, passmark is more accurate than any other single source anywhere. If you have the nerve to call it a "joke" then you are nothing more than a hate-big geek-snob who has probably never done any extensive testing or research on the subject.

Take into account the single threaded and regular scores (which are multi-threaded) and you can get a good estimate of overall performance in various workloads. The single threaded performance needs to be weighted more heavily for consideration in real-time workloads like gaming. The main problem with passmark is not passmark, it's those who look at it and don't know how to interpret and translate the scoring to real world applications.

Having done hundreds of hours worth of independent bench-marking myself (using real world workloads), I can say with confidence that there is no existing benchmark "score" system that is more useful or more accurate than passmark for CPUs, provided the reader understands how to properly weight the single vs multi-threaded scores to the workload at hand.

A great example of how the passmark scores are interpreted to be "a joke" by an amateur would be to look at the i7-2600, noting that it scores higher than say, an i5-4690. The i5 haswell is stronger in gaming, but the older i7 has higher combined compute throughput. Without an additional observation of the single threaded scores, and proper weighting in it's effect on the workload in question, the regular (multi-threaded) passmark score can be misleading. The i5 in this case is almost up to 20% faster in workloads that wind up throttled by whatever the maximum throughput on any one thread is.
 
Lol. This article is pathetic and Tom's is obviously paid to promote Intel. I'm seriously not sure how much longer I'll stay here when it is so blatantly and obviously biased.
 


no, they aren't paid shills. As much as I hate to admit it, Intel does have the best processors for gaming, without exception.

I used to love AMD processors. I had all AMD processors from my Athlon 450 until my Athlon X2 6400+, and I had about ten chips in between. AMD simply hasn't kept up since Intel came out with the Core 2. At first, it wasn't by much. AMD processors would use more power and couldn't overclock as high, but were able to keep up at stock. Later, Intel updated their core to Nahalem, and AMD stuck with the same, tired design, and the lead lengthened. Then Sandy Bridge came out, and AMD still had the same basic design, and things got worse.

The only place where they win out now is in the lowest end with their APUs, and then just because of their superior integrated graphics. Their current CPU design just plain sucks. Bulldozer was a flop, and they've only slightly updated it since then. Piledriver and Steamroller did very little to help that, while Intel's improvements just made their lead even bigger. Clock for clock, Haswell is easily 40% more efficient than Steamroller. When power usage is figured in for calculations per clock cycle per watt, it becomes more like 150% more efficient. Figuring in cost, it's just no contest.

I would love it if AMD were keeping up. That would keep costs down and encourage further advancements. As it is, Intel is not likely to see any competition for a long time. I doubt AMD will recover from this.
 
I Have a Phenom II 965(deneb) and I don't believe i am just 2 tiers below a core i7 5960X.... More Tiers for high end cpu's please .. I also love amd but, we need to see truth ..
 
I'd like to point out that the specs for the Core i7-5930K on the last page are wrong. The ram is listed as DDR3 instead of DDR4 as well as the frequency beings off. Socket also doesnt say v3.
 
Don Woligroski@
mistake in article:
in last page you said 4460 is the winner ppd..but in the chart i do not see any 4460?just 4430
by the way thank you very much for wonderful review
 
There should be mention of the other locked i5's, since the prices on them fluctuate so much. Some days you can get a 4460 for less than a 4430. Others I have seen the 4590 within $5-10 of the 4430. The less knowledgeable will not be aware of this fact, and just get the 4430, when they could possibly get a better chip for the same cost, a little less, or for just a few dollars more.
 


Don't like Passmark then search for "Firestrike ranked by CPU" and you will see the top Firestrike runs by CPU or you can enter by GPUs. You might need to have run the free Firestrike and registered first.

 
Can you please add best HTPC cpu/apu for the money? I think it would be super helpful as the number of skus and different tdp realy make it difficult to know if its worth saving a little more on the low end.
 
I don't understand the Performance per Dollar page. Are the bars really the FPS of the games or are they a percentage when using the i7-5930K as a ceiling? If they're a percentage, then what are the FPS the i7-5930K gets?
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom