Probably in the Intel reviewers guide. Maybe not as "benchmark with MCE ON", but as, "Reset CMOS to default before benchmarking". Which... viola, with compliant board makers, gives you BONUS (bogus, really) BENCHMARKS. MOAR SCORE!
Probably in the Intel reviewers guide. Maybe not as "benchmark with MCE ON", but as, "Reset CMOS to default before benchmarking". Which... viola, with compliant board makers, gives you BONUS (bogus, really) BENCHMARKS. MOAR SCORE!I find the amount of reviews published demonstrating 'out of box' performance with MCE enabled quite astounding. why has this suddenly become OK?
Also probably the main reason for the runaway power consumption numbers, resulting in the thought line that CFL is somehow inefficient. MCE seems to jack up the voltage a lot.I find the amount of reviews published demonstrating 'out of box' performance with MCE enabled quite astounding. why has this suddenly become OK?
For me, this is my number one reason I don't like MCE. The voltage is too aggressive. I have seen some screenshots of it enabled running 4.7GHz on all cores at ~1.4v...when it could probably do it at <1.2v.MCE seems to jack up the voltage a lot.
Also probably the main reason for the runaway power consumption numbers, resulting in the thought line that CFL is somehow inefficient. MCE seems to jack up the voltage a lot.
Probably in the Intel reviewers guide. Maybe not as "benchmark with MCE ON", but as, "Reset CMOS to default before benchmarking". Which... viola, with compliant board makers, gives you BONUS (bogus, really) BENCHMARKS. MOAR SCORE!
Probably in the Intel reviewers guide. Maybe not as "benchmark with MCE ON", but as, "Reset CMOS to default before benchmarking". Which... viola, with compliant board makers, gives you BONUS (bogus, really) BENCHMARKS. MOAR SCORE!
On the topic of Multi-core enhancement, what precisely is its effectiveness on Coffee Lake non-K processors? In Skylake and Kaby Lake, setting was ineffective, so I extrapolated the same for Coffee Lake. TechPowerUp did not get MCE to work on ASRock and Core i5-8400, but GIGABYTE USA representative replied on YouTube that MCE is effective on non-K processors. Any other reviews, including non-English, that cover this topic?
I am on the hunt for confirmation on the same issue. There is a user on another forum getting a 8700 Monday that is going to test MCE.On the topic of Multi-core enhancement, what precisely is its effectiveness on Coffee Lake non-K processors? In Skylake and Kaby Lake, setting was ineffective, so I extrapolated the same for Coffee Lake. TechPowerUp did not get MCE to work on ASRock and Core i5-8400, but GIGABYTE USA representative replied on YouTube that MCE is effective on non-K processors. Any other reviews, including non-English, that cover this topic?
Good info. Gigabyte rep made it sound otherwise.I have an i5 8400 in ASUS Strix z370-E and MCE doesn't seem to do anything aside from jacking up the voltages. Tried various settings including manual adjustment, but turbo behaviour remains the same.
That's not exactly the subject here but ok, let's compare (based on newegg prices); in BF1 multiplayer at 1080p, an 8400 is 23% faster than a 1600X, and its 99th percentile is 10% higher
.
Are you able to manually increase max turbo multiplier for 6C operation to values higher than 38x?I have an i5 8400 in ASUS Strix z370-E and MCE doesn't seem to do anything aside from jacking up the voltages. Tried various settings including manual adjustment, but turbo behaviour remains the same.
Reviewers should clearly state that once they leave MCE enabled, they are reviewing the board, not the CPU. Also, MCE enabled tests belong to the overclocking section of the review.I find the amount of reviews published demonstrating 'out of box' performance with MCE enabled quite astounding. why has this suddenly become OK?
https://nl.hardware.info/reviews/75...n-round-up-17-keer-coffee-lake-stroomverbruik
What as massive difference in power consumption between the boards, caused by the Multicore enhancement and may be other things. It's a pity no reviewer tried out some bios tweaks, for example disabling the MCE on Asus boards (looks like auto is default on them).
I find the amount of reviews published demonstrating 'out of box' performance with MCE enabled quite astounding. why has this suddenly become OK?
It's not good to pass off results with MCE on (either performance or power consumption) as "stock" because that's simply not what it is -- it's an overclock. But I do think reviewers should have results with MCE both on and off, since many users will turn on MCE.
Not Intel at all. It's the MB makers that are trying to make their MB looks "faster".
It is definitely a feature that should be off by default. Power consumption jumped something like 40% for a 10% performance boost.
Stock Settings should really be Stock.
Not Intel at all. It's the MB makers that are trying to make their MB looks "faster".
.
Asus overclock both the core and the L3 cache, HFR made some nice graphs :
That s about it but Intel is not that innocent...
http://www.hardware.fr/articles/970-17/attention-aux-overclockings-automatiques.html
Asus overclock both the core and the L3 cache, HFR made some nice graphs :
Are you able to manually increase max turbo multiplier for 6C operation to values higher than 38x?
If so, then the next logical step is to check for the Max Current and Short Term / Long Term power settings and increase those as well. The Max Current is probably a must to see the effect under load, but increasing the TDP might also help.
Could you post his result in this thread?I am on the hunt for confirmation on the same issue. There is a user on another forum getting a 8700 Monday that is going to test MCE.
Sorry to hear that. I'll probably be able to test myself on a 8700 + MSI mobo towards the end of next week, see if there's more we can try.Thanks, I tried this but turbo behaviour remains the same - perhaps I'm missing something. I can manually set each core to x40 and adjust settings as suggested, however there is no change.
I am considering getting a 8700k and Z370 motherboard, as I don't really need the extra PCIE lanes so much from the 5930k anymore. My question is, how much more performance would I get out of a 8700k, vs a 5930k @4.5 GHz? I would be ocing the 8700k as well, same cooler likely. Worth it perhaps?
Also, what motherboards look good, and where to get a 8700k? :/
Those are some epic gains from Uncore OC.
Is there any review that does the following:
1) Using DDR4 CL14
2) Uncore @ 4.4 - 4.5
3) AVX offset ~5
4) 4.9-5.1Ghz top clock on sensible non auto voltages according to capabilities of their chip. Not the auto 1.4 volts, but rather "if my chip requires 1.275 for 4.9 and 1.35 for 5Ghz".
Perf, power numbers in games, Cinebench would be lovely for memory/uncore/clock scaling would be amazing. That would be lovely real world OC investigation, not MCE drama/soap opera.
Just downloaded and ran, got 1306.What's your CB score at 5930@4.5 ?