Question about 2X4 vs 8 pin CPU power cable compatability

[VinnyVincent]

So, the motherboard looks like this:

and the plug from my PSU looks like this:

Notice how on the right side of the plug, they're all semi-round, where on the MoBo that side is a mixture of square and semi-round?
Will that cause any issues? I'm asking because it hooked up and seems to be running fine, however I just upgraded from a B250 MoBo with the four pin connector, to a z270 with the 8 pin connector...My B250 was doing an 1188 @ 70c on the intel XTU benchmark, and my first test with the Z270 was only a 700 @ 80c. In spite of the higher temperature, it was only getting up to 65-70W TDP where the B250 would reach 91W.
Even with a slight overclock on the z270, it's only benching 1213 @ 80c. Something seems off for sure. I'm also getting random freezes for a few seconds when switching between programs, which never used to happen.
I tried a fresh install of windows and that made no difference.

Any ideas what may be causing this?

 

[Snipergod87]

Well that is the correct power connector for that port, if it wasn't your system would literally be fried, so the problem lies elsewhere.

 

[VinnyVincent]

Thanks for the confirmation.
I'd blame the issue on the MoBo, but this is actually my second Z270 I've tried with the exact same issue. I got the first one back in Jan. as an open box and returned it assuming I got a bad board. I ended up just going back to using my b250 for a while.
Recently the ASUS TUF 270's went on sale which prompted me to pick one up, so I could do some overclocking...hooked it up and it is behaving just like the other z270 I returned.
One would assume that out of the box, the z270 would perform at least on par with a b250.

What do you think could be causing it?

 

[Snipergod87]

What clocks speeds are you getting when doing the test's? Some baord by default have MCE turned on, and other's off, this will greatly affect multi threaded benchmarks.
(MCE=Multi Core Enhancement, which set all cores to run at max turbo speed which it typically only allowed for workloads using 1-2 cores)

 

[VinnyVincent]

What clocks speeds are you getting when doing the test's? Some baord by default have MCE turned on, and other's off, this will greatly affect multi threaded benchmarks.
(MCE=Multi Core Enhancement, which set all cores to run at max turbo speed which it typically only allowed for workloads using 1-2 cores)

The max core frequency on the b250 was reading 4.31 and on the z270 it's reading 4.51.

One other thing I notice is that it seems to be less stable...meaning when I look at the CPU utilization on the XTU test with the B250, it had a nice, smooth , predictable spike pattern. On the z270 it's kind of rough/jittery and all over the place.
]

 

[VinnyVincent]

Essentially it looks like even though the z270 is reaching a higher max frequency, it's spending more time at lower frequencies during the test, due to the difference in the spike patterns. I could post some screens shots of both tests later today when I get home from work.

 

[zyh1987]

whats the model for b250 and z270. Also your cpu.

AFAIK tuf z270 mk2 is worse than a lot of the b250 board. Mk1 is looks amazing and I think it’s not bad.

 

[VinnyVincent]

whats the model for b250 and z270. Also your cpu.

AFAIK tuf z270 mk2 is worse than a lot of the b250 board. Mk1 is looks amazing and I think it’s not bad.

The B250 is an ASRock B250M-HDV
http://

and the z270 is an ASUS TUF Z270 Mark 2
http://


FWIW I had a Gigabyte GA-Z270-HD3 back in Jan. and was having the exact same issue(lower performance, higher temps)
http://

 

[zyh1987]

whats the model for b250 and z270. Also your cpu.

AFAIK tuf z270 mk2 is worse than a lot of the b250 board. Mk1 is looks amazing and I think it’s not bad.

The B250 is an ASRock B250M-HDV
http://

and the z270 is an ASUS TUF Z270 Mark 2
http://


FWIW I had a Gigabyte GA-Z270-HD3 back in Jan. and was having the exact same issue(lower performance, higher temps)
http://

are you trying to run a 7700k or something?

 

[VinnyVincent]

Yes it’s a 7700k. Sorry forgot to mention that

 

[zyh1987]

What clocks speeds are you getting when doing the test's? Some baord by default have MCE turned on, and other's off, this will greatly affect multi threaded benchmarks.
(MCE=Multi Core Enhancement, which set all cores to run at max turbo speed which it typically only allowed for workloads using 1-2 cores)

The max core frequency on the b250 was reading 4.31 and on the z270 it's reading 4.51.

One other thing I notice is that it seems to be less stable...meaning when I look at the CPU utilization on the XTU test with the B250, it had a nice, smooth , predictable spike pattern. On the z270 it's kind of rough/jittery and all over the place.
]

what cooling solution do you have?

higher voltage may cause higher power consumption, meaning that your vrm needs some cooling.

with an AIO, there is very little airflow hence, it cause uneven power delivery.

 

[VinnyVincent]

What clocks speeds are you getting when doing the test's? Some baord by default have MCE turned on, and other's off, this will greatly affect multi threaded benchmarks.
(MCE=Multi Core Enhancement, which set all cores to run at max turbo speed which it typically only allowed for workloads using 1-2 cores)

The max core frequency on the b250 was reading 4.31 and on the z270 it's reading 4.51.

One other thing I notice is that it seems to be less stable...meaning when I look at the CPU utilization on the XTU test with the B250, it had a nice, smooth , predictable spike pattern. On the z270 it's kind of rough/jittery and all over the place.
]

what cooling solution do you have?

higher voltage may cause higher power consumption, meaning that your vrm needs some cooling.

with an AIO, there is very little airflow hence, it cause uneven power delivery.

It is an AIO cooler. The case has really good air flow, though. There's two 140mm fans in front, the radiator fan, which blows out of the top of the case, plus an additional 120mm 65cfm exhaust fan, which I believe is right next to the VRM.
There's not much heat coming from the GPU either since it is just a GTX1060 it's pretty low power/heat.(not to mention it's pretty much idling during the CPU bench test)
As far as voltage goes, IIRC according to bios it's running at just slightly over 1.1.

Is there a way to check for temps on peripherals like that? I'd really like to get this figured out as it doesn't even really make sense to keep using this board if my b250 performs better...the fact that it's running a full 10c+ hotter during any given task, without really performing any better is especially annoying.

 

[VinnyVincent]

I do notice that the VRM is right next to where my radiator is located at the top of the case....

 

[zyh1987]

I do notice that the VRM is right next to where my radiator is located at the top of the case....

vrm on the top is usually for onboard graphics, the actually vrm is three parallel to ram slot, plus the left most vrm.

however, the io cover is really not helping at all. can you try to reverse the radiator fan downwards, that should forces the air into the cpu area.

 

[VinnyVincent]

I do notice that the VRM is right next to where my radiator is located at the top of the case....

vrm on the top is usually for onboard graphics, the actually vrm is three parallel to ram slot, plus the left most vrm.

however, the io cover is really not helping at all. can you try to reverse the radiator fan downwards, that should forces the air into the cpu area.

I can do even better than that. I have an open fan bay right above that area(it's only a 120mm radiator) along with an extra 120mm fan I can put in a reverse direction right there.
I'll play around with it when I get home from work later and post up some pics with benchmark results.

 

[zyh1987]

Asus is known to put these massive io shroud (they do look good), not too much issue with their high end boards, but vrm on lower end boards tends to get hot.


going off topic here:
from what i heard, this was when asus started to destory their TUF brand.
originally, tuf brand is a cut down *real* rog board, less oc feature, solid component. (check the layout on z270 tuf mk1 and you will see what i mean)
now, most tuf boards from what i see is a cut down prime series with black paint.

 

[VinnyVincent]

Asus is known to put these massive io shroud (they do look good), not too much issue with their high end boards, but vrm on lower end boards tends to get hot.


going off topic here:
from what i heard, this was when asus started to destory their TUF brand.
originally, tuf brand is a cut down *real* rog board, less oc feature, solid component. (check the layout on z270 tuf mk1 and you will see what i mean)
now, most tuf boards from what i see is a cut down prime series with black paint.

Okay so I didn't even bother installing the other fan. I located the VRM and it's got a massive heat sink on it.

I did several bench marks and after touching the heat sink, it's barely even getting warm. Maybe just a bit over body temperature.
I also downloaded speed fan and none of the sensors were getting very hot.

Here are some bench marks to compare:
This one is with the B250 board:
Pay close attention to the CPU utilization chart.

Compare that to the chart on the z270:

I also did a UserBench to compare to other z270 boards, which says the CPU is performing below average at the 38th percentile. However, I'm running stock clock speeds so I'm thinking that's mostly why. http://

 

[VinnyVincent]

One thing I'm noticing is the max core frequency during idle.(red on the chart) Do you notice on the B250 that it's pinned at 4.32ghz even when it's idling before and after the test...but on the z270 the idle core frequency is getting down to 1ghz or so when it's idling before and after the test.
What could be causing that and do you think that may be the issue?

Oh and I forgot to include the one that shows the actual benchmark numbers for the z270 on XTU...

 

[zyh1987]

One thing I'm noticing is the max core frequency during idle.(red on the chart) Do you notice on the B250 that it's pinned at 4.32ghz even when it's idling before and after the test...but on the z270 the idle core frequency is getting down to 1ghz or so when it's idling before and after the test.
What could be causing that and do you think that may be the issue?

Oh and I forgot to include the one that shows the actual benchmark numbers for the z270 on XTU...

the difference behavior could be related power saving feature in the bios.

it's hard to tell, as your b250 and z270 are from different company. do you have similar voltage when running b250 vs z270? try manually set vcore on both and see if that changes the temperature or not.