Anyone running a professional application is going to be leery of overclocking (not semi-professional but real professional). Gamers will overclock but people working and running NAS or database searches are more than not going to run the chip at stock (and if they are going to be running the chip at full load then power consumption is going to come into play). At stock we are using 75 watts more than the i7 and getting similar performance (within 5%). If that is running 24/7 then assuming electricity is 11 cents a kilowatt (0.075*24*365*0.11 =$72 a year in additional electricity consumption from the cpu alone--ignoring additional losses from the power supply(efficiency is a percentage of output) and additional costs to cool the case--AC). Now look at the overclocked power usage (generally going to be about 10% faster than the i7) and suppose power is twice the price (as it is in many countries). I use this argument to illustrate why anyone who is going to run professional applications 24/7 probably woudn't overclock and are concerned about cost over the lifetime of the machine. For the casual professional (not someone who has a rendering project on 24/7) the 8350 is a good deal and buy but if power consumption is coming into play then things change.
"Do you not still understand that people who buy FX chips use them for both work and gaming. Did you see the benchmarks I gave to you? What do you believe NAS, C-ray, NAS, BZIP, 3DMAX, database search, JTR... benchmarks are for?"
They are going to care about extractable performance.
Intel is a penny pincher who likes to nickle and dime their buyers with motherboard changes every couple years.
I said multiple times FX overclocks well. To put it in perspective what clocks is the average air-cooling enthusiast going to use. That is what matters the most. No one is getting 5.2 ghz with ease. 8350 tops out at 4.8 with an air cooler.
They can probably get about 4.7-4.8 on a good chip under air (from stock 4.0 (turbo 4.1)) that is an overclock of 17% (4.8/4.1). The 3770k can get around 4.4-4.5 ghz on a good chip (from a 3.8 turbo (regular 3.5)) on four cores. That is an overclock of 16-18%, which is roughly the same. The FX can hit higher clocks but the gain that both chips gain from overclocking is roughly the same (FX does not scale as well with frequency because the cache runs at a constant speed).
They got their i7 chip to 4.9 ghz (obviously an outlier).
http://techreport.com/review/22833/ivy-bridge-on-air-the-core-i7-3770k-overclocked-on-four-motherboards
Toms hardware review (bolded is mine)
Using a 1.375 V CPU voltage and a 1.175 V northbridge voltage, I was able to get FX-8350 running stably at 4.8 GHz under full load. In the screen capture above, I'm running a single-threaded test to spin the chip up, but the highlighted maximum temperature is where our benchmark suite peaked.
The FX-8350 wanted to go even faster, but the key here is a voltage setting low enough that you avoid hitting 70 degrees Celsius. At that point, the thermal monitor starts cycling cores to throttle down (evidenced in the image above), keeping the chip from getting any hotter and negatively impacting performance. So long as I didn’t trigger any threaded workloads, I was even able to run benchmarks as high as 5.125 GHz (requiring a 1.4375 V CPU voltage and 1.2 V northbridge setting).
Their system builder (not all chips are created equal --FX requires good cooling).
This is the first time anyone at Tom's Hardware has tried his hand at overclocking a retail FX-8350. And, after reading Chris' experience taking his sample from AMD up above 5 GHz, I was looking forward to something similar. It turns out that I was being far too ambitious, though. Xigmatek's Loki doesn't have the headroom to keep the 125 W processor cool beyond its stock clock rates. Beyond performance, thermals are probably AMD's biggest disadvantage in this comparison. We really would need to spend a lot more on cooling to achieve any sort of meaningful overclock.
Regardless of the processor or northbridge voltages we used, we couldn't exceed 4.63 GHz. "Fair enough," I first thought. "If I disable Turbo Core and lock the chip in at 4.6 GHz, I should still see a reasonable speed-up." But a Prime95-induced load quickly demonstrated instability as the FX-8350 shot up over 80 degrees.
It seems as though I had underestimated the FX's ability to generate copious heat, and failed to budget enough for cooling. Even at the stock 1.35 V setting, and with the clock rate dialed in to the processor's peak Turbo Core frequency of 4.3 GHz, Prime95 caused the chip to falter. Simply nudging clock rate, without touching the voltage, results in a significant temperature increase. For example, operating at 4 GHz yields a maximum 60-degree reading, but 4.2 GHz sees that number jump to 70 degrees. Interestingly, I didn't see any throttling, as Chris did when his sample crested 70 degrees. Here's the thing, though: while his Tj. Max was reported as 70 degrees, the retail processors are capped at 90, though the chip is clearly unstable well before it gets that hot.
The best I could achieve with this build's heat sink was 4.33 GHz, forced by dropping the voltage to 1.3375 V, turning off Turbo Core, and increasing the multiplier. Prime95 didn't crash, and the temperature stayed under 75 degrees. We're hesitant to call this a bad sample when the cooler is seemingly barely adequate. Should we choose an FX in the future, we'll need to cut back elsewhere on our budget to leave more room for a higher-end air or closed-loop liquid solution.
Anandtech review
AMD's FX architecture was designed for very high clock speeds. With Piledriver we're able to see some of that expressed in overclocking headroom. All of these chips should be good for close to 5GHz depending on your luck of the draw and cooling. For all of these overclocking tests I used AMD's branded closed loop liquid cooler which debuted back with the original FX launch. I didn't have enough time to go through every chip so I picked the FX-8350 and FX-4300 to show the range of overclocks that may be possible. In my case the FX-4300 hit 5GHz with minimal effort, while the FX-8350 topped out at 4.8GHz (I could hit 5GHz but it wasn't stable through all of our tests). Both of these overclocks were achieved with no more than 10% additional core voltage and by simple multiplier adjustments (hooray for unlocked everything). The increase in performance is substantial:
Tech report
When you're overclocking a CPU that starts out at 125W, you're gonna need some decent cooling. AMD recommends the big-ass FX water cooler we used to overlocked the FX-8150, but being incredibly lazy, I figured the Thermaltake Frio OCK pictured above, which was already mounted on the CPU, ought to suffice. After all, the radiator is just as large as the water cooler's, and the thing is rated to dissipate up to 240W. Also, I swear to you, there is plenty of room—more than an inch of clearance—between the CPU fan and the video card, even though it doesn't look like it in the picture above. Turns out the Frio OCK kept CPU temperatures in the mid 50° C range, even at full tilt, so I think it did its job well enough.
Trouble is, I didn't quite get the results I'd hoped. As usual, I logged my attempts at various settings as I went, and I've reproduced my notes below. I tested stability using a multithreaded Prime95 torture test. Notice that I took a very simple approach, only raising the voltage for the CPU itself, not for the VRMs or anything else. Perhaps that was the reason my attempts went like so:
4.8GHz, 1.475V - reboot
4.7GHz, 1.4875V - lock
4.6GHz, 1.525V - errors on multiple threads
4.6GHz, 1.5375V - errors with temps ~55C
4.6GHZ, 1.5375V, Turbo fan - stable with temps ~53.5C, eventually locked
4.6GHZ, 1.5375V, manual fan, 100% duty cycle at 50C - lock
4.6GHZ, 1.55V, manual fan, 100% duty cycle at 50C - crashes, temps ~54.6C
4.4GHz, 1.55V - ok
4.5GHz, 1.55V - ok, ~57C, 305W
4.5GHz, 1.475V - errors
4.5GHz, 1.525V - errors
4.5GHz, 1.5375V - OK, ~56C
At the end of the process, I could only squeeze an additional 500MHz out of the FX-8350 at 1.5375V, one notch down from the max voltage exposed in the Overdrive utility. AMD told reviewers to expect something closer to 5GHz, so apparently either I've failed or this particular chip just isn't very cooperative.
I disabled Turbo Core for my initial overclocking attempts, but once I'd established a solid base clock, I was able to grab a little more speed by creating a Turbo Core profile that ranged up to 4.8GHz at 1.55V. Here's how a pair of our benchmarks ran on the overclocked FX-8350.
From openbenchmarking
In the end, we were able to take the FX-8350 up to a stable 4.7GHz. Unfortunately, due to time constraints and an incompatibility with AMD OverDrive and our test-bed’s motherboard, we don’t have accurate temperature data to share at this point. But considering how easy it was to take our CPU to 4.7GHz, we suspect that higher clocks will easily be possible with more exotic cooling and more aggressive voltage tweaking.
Under air, clocks top out at about 4.8 ghz on a good chip for the 8350 and 4.4-4.5 for the 3770k. The % overclocks are similar.
Intel's stupid decision was probably decided by the marketing/accounting team to maximize profits. Or possibly the engineers were told to bring the costs down to $x per cpu and cutting the heat transfer material was the cheapest and easiest way to do that (considering as a percent of the market, few people overclock). Anyway, to say the ivy bridge "architecture" is poor is incorrect, rather the ivy bridge "implementation" is poor.
Edit:
"You cannot pretend to select a few biased tests {*} showing a 15% difference and say that the "fx8350 lagged behind", whereas say that the i7 was "very, very close" on tests where the FX was 30-70% faster."
What biased tests? The FX isn't 30-70% faster on two of the three tests, it was pretty much margin of error (<5%). The other test the FX was significantly faster because for some unusual reason hyperthreading wasn't being used (which is unusual but a fair victory to the 8350). I'm saying that we must look at all the tests. You are essentially showing me three tests from a sample where the FX is basically tied or beating the i7-3770k. What about other tests in that review? Where is the link?
Facebook
Twitter
Instagram