Fulgurant :
I supplied worst-case numbers taking the overclock's load scaling into account, and still couldn't come up with a slam-dunk case for the i5 3570k on a price/performance basis.
We're talking about an overclock (for the 3350p) at full load from 3.1 GHz to 3.5 GHz (3.7 GHz single-threaded). That's nothing special, true. But a ~13% overclock at full load is certainly worth mentioning. It's certainly not logical to dismiss the full-time 400 MHz overclock while simultaneously insisting that the extra 200 MHz of load-variable speed is deathlessly important.
With all due respect, the bolded statement is flatly asinine. If you think Tom's has been running a benchmark suite exclusively or even primarily comprised of single-threaded games, then you're still living in 2006. And if you really think an i5 3350P even at stock speeds can't run modern games extremely well, then you've lost all perspective. Have you read the article that goes with this comment thread?
usually whatever follows "with all due respect" is an insult. see that holds true here.
Listen. i'm not trolling you, nor am i saying what you think i'm saying, we're simply talking past eachother, and you're getting more and more agitated thinking i'm saying things i'm not. I'm not knocking an i5's performance. In the grand scheme of things ANY i5 is a high performing cpu.
I think what started this misunderstanding is i didn't quote the comment i was responding to, which of course left my initial comment without the necessary context (especially since the innitial comment wasn't directed at the post directly above mine).
I was commenting on people's general discussion about cheap intel mbs and cheap non-k chips being a better buy then a higher end settup. There was a great deal of confusion in the comment, with some weird expectation that this low end cpu would match an overclocked K series or something. Elsewhere in the forums, not too much sooner, i was chatting with someone who was trying and failing to overclock their non-k intel... and i was getting frustrated becuase he seemed to be quoting a Tom's article expressing how awesome the i5 (non-k) will performed overclocked.
In general, turbo overclocking is what i'd call "extremely light" overclocking. The way Tom's portrays (in their i5-3350p reviews for example) the performance of these chips is very idealized. And in my experience well outside the range of what an end user might experience.
Lets put to bed the whole "turbo overclocking" thing... cause we all know it produces an imperceptible change in performance for the most part.
Lord, if a PhII x4 965 is generally inperceptibly slower then an i5-3570k, no one will be able to tell the difference between a stock i5-3350p and a turbo overclocked one.
My general stance on this issue is actually the opposite you think it is. I don't think the end user will ever be able to tell the difference between one quad core or another... be it a core2duo quad core, a Phenom II x4, an FX4300, a i5-3350p or an i5-3570k. Generally in 95% of day to day operation there will be almost no way to tell them appart without a stopwatch.
which leaves us with the performance end of the scale under heavy loads... where more cores, or higher clocks will rule... and a turbo overclocked i5-3350p will be noticeably slower then an
actual overclocked i5-3570k.
as a result i think the general conversation about cheap motherboards and impressive performance in locked chips is sorta funny. because there really hasn't been a great deal of cpu inovation in the past 5 years that the end user can detect, appart from high end high cpu usage aps.
and if you're buying/building a new pc and not building it for the "other" 5% of the time where all that extra power IS evident, then you have no reason (unless you're on something worse then a core2due) to upgrade.