Gadgetory

welcome back to harbor unboxed I'm your host Steve and today we're doing a little bit of core i7 70 100x overclocking just a little bit I do realize everyone is height for thread Ripper right now but we're still a few weeks away from how to show you any benchmarks so let's focus our attention on something we can show you that said I discussed the 1920 X in 1950 XO the new thread ripper parts in yesterday's video so if you miss that check it out they do mighty impressive in the meantime though I slip a few intel core X tests something together the way so let's do that alright so three weeks ago now I certain see if it was possible to keep the new Intel Core I 970 100x cooler 4.7 gigahertz without having addy lead the $1,000 chip in the end I found that Thermaltake specific RL 360 custom lube liquid cooling kit was able to keep the semi 100x in check and temps hovered around 73 degrees under load which is acceptable certainly not a bad result for a 10 core CPU clocked at 4.7 gigahertz though it did push total system consumption to an insane 375 watts under load so yeah overclocked is spat out a score of 2500 21 points which at the time I deemed quite impressive that said though aimed at probably a little less impressed as there 16 core 1950 X produces a score in excess of 3,000 points so that kind of puts a damper on my 70 100x overclock anyway a number of you have demanded I do the same and overclock the much more affordable six core Core i7 700 X so that's what I've done recently we saw how the 700 X compares to the 7700 K in 30 games and even overclocked it wasn't very impressive for the most part as disappointing as the gaming performance is productivity is a little more inspiring in order to stabilize the 4.7 gigahertz overclock on the 700 X we required a V core of 1.25 volts and this led to a low temperature of 68 degrees so just five degrees less than the ten core part mind you I was using slightly less voltage as well I should also note that when stressing the FPU the temperature spiked up towards 100 degrees and throttling was detected anyway with the temperatures in check for the most part I moved on to run a few quick tests first up around Cinebench r15 as this has become a fairly standard benchmark for measuring CPU performance here the overclocked 700 X produced a multi-threaded score of fourteen hundred and eighty points with a single thread score of 191 points and that's only about a 10% improvement over the stock result next up I ran blender and here the render was completed in twenty three point four seconds which is faster than a stock core i7 sixty nine hundred K but again only a 10% improvement over the out-of-the-box performance last up I checked performance using the corona benchmark and here the test took 144 seconds and as you guessed it that's a 10% improvement over the stock performance so if that's pretty disappointing given we are squeezing at least 18% more frequency from the CPU I saw no evidence of throttling in my tests apart from that one FPU stress test and the vrm temps were well within acceptable limits most telling though was the fact of power consumption increased by an insane 52% from 236 Watts seen previously during the cinema r15 multi-threaded test to an eye-watering 358 watts that's not really much better than thrown in 75 watts the 10 core part gobbled up lowering the voltage further led to crashes so the only way to shed those what if you will is to lower the frequency as well anyway at this point it's hard to tell if the disappointing performance is down to an issue with the x-29 motherboard that I use for testing or if it's just the CPU itself in any case I can't imagine the results are too far from the truth and at this point I have tried two different external motherboards of admittedly they admittedly they were both as rock boards so it would be good to test a board from a competing brand anyway last week when I did conduct my 30 game benchmark comparing the 77 okay and see 100x both stock and overclocked I found that on average the seminar X enjoyed just a 5% boost in performance from the overclock when gaming that said though the 70 similar case or even less of a gain though for the most part that CPU is probably being either GPU limited or there was a frame cap so the game was limiting performance that way in any case it's really starting to look like overclocking your skylake xcp you simply isn't worth the added heat and power draw and well that's where I'm going to end this rather short video what they don't like to be benchmarked marathons you