PDA

View Full Version : Z370-E 15. december BIOS 0605 & 5605



Carbonicdk
12-16-2017, 04:36 PM
Please share your experiences and issues with these bios versions. Most likely identical but with different settings with the beta driver having all cores synced (although you can't tell due to the bad descriptions and naming schemes).

Version 5605 Beta Version
2017/12/158.21 MBytes
ROG STRIX Z370-E GAMING BIOS 5605
Sync all core
Sufficient processor cooling is required for better performance.

Version 0605
2017/12/158.21 MBytes
ROG STRIX Z370-E GAMING BIOS 0605
Improve system stability

I'll share myself once I am done testing.

Tbolt214th
12-16-2017, 06:13 PM
It still wont XMP my Corsair Vengeance CMK32GX4M4B3600C16 4x8-Kit. All sorts of BSOD and error msg from Bios that my ME-downgrade had failed, meSpi,.. something. Same as before, not even a latency till 20-20-20-38 will let them boot any OS, up to 1.40v neither. VCCIO VCCSA both upped by Auto 1.3 and 1.35 ( maybe too high for Z370/8700K ?? )

A bit dissapointing :( My Z270 Prime-A ran them just fine with XMP, tho not from the beginning on, that needed a few bios updates as well, so there is hope.

Sys:
Asus ROG Strix 370 Gaming-E, 8700k, Asus Poseidon GTX1080-Ti, Corsair AXi-1200, DIY-HighEnd watercooling

AntonioL
12-16-2017, 11:04 PM
Please share your experiences and issues with these bios versions. Most likely identical but with different settings with the beta driver having all cores synced (although you can't tell due to the bad descriptions and naming schemes).

Version 5605 Beta Version
2017/12/158.21 MBytes
ROG STRIX Z370-E GAMING BIOS 5605
Sync all core
Sufficient processor cooling is required for better performance.

Version 0605
2017/12/158.21 MBytes
ROG STRIX Z370-E GAMING BIOS 0605
Improve system stability

I'll share myself once I am done testing.

For the Z370-F, this is what you can read under this new 0605 (I wonder what is the logic for the numbering) BIOS:
ROG STRIX Z370-F GAMING BIOS 0605
Enable Intel(R) Speed Shift Technology.

Speed Shift Technology is supposed to be the replacement technology for Intel Speed Step Technology (EIST), that allows the CPU to adjust its frequency according to what the OS requests based on the load. With this new approach introduced with Skylake and Windows 10, the CPU can set its frequency itself, determining the right value for the need. The advantage is said to be the speed of the frequency switch (1ms vs 25ms says Intel) and the unparking of cores (35ms vs 100ms), making the system more reactive. I have never read any proof of the reality of this assertion, but if it is properly managed and does not induce any issue (even more audio popping for instance...), it cannot be bad.

Maybe someone from Asus (Mr Raja, please) could confirm ?

I would have tested by myself, but the current BIOS version (0420) of my Z370-F runs fine (except for audio crackling) and I am afraid to lose stability (such as RAM XMP compatibility). I am not sure it is possible to revert to an older BIOS version, some people on this forum indicate that no.

jologskyblues
12-17-2017, 06:34 AM
I also have a Z370-F and speed-shift and XMP works fine. I'm not having any sound issues with neither the onboard sound nor my USB DAC. The system feels more snappy and responsive as expected even with the balanced power profile selected.

BTW, speed shift can still be disabled in the UEFI BIOS settings if that's what you prefer. The new update just gives us the option to use it if we want to.

Carbonicdk
12-17-2017, 12:50 PM
So far I've not seen much improvement in this BIOS release. My G.Skill F4-3600C16Q-32GTZR kit's XMP settings still don't allow the computer to boot and run stable unless I ramp back the mhz to 3000mhz from 3600mhz.
Also still seems hard to get a stable overclock that doesn't have weird voltage spiking.

Tbolt214th
12-17-2017, 01:03 PM
So far I've not seen much improvement in this BIOS release. My G.Skill F4-3600C16Q-32GTZR kit's XMP settings still don't allow the computer to boot and run stable unless I ramp back the mhz to 3000mhz from 3600mhz.
Also still seems hard to get a stable overclock that doesn't have weird voltage spiking.

It seems the boards dont like the CL16 settings at all. Fastest I can get stable in all tests is 5.0/3200CL16-16-16-36-2.

All 3600/CL16 in QVL have 16-18-18-36, none is a straight 16-16-16-36. Maybe the Z370/8thgen-IMC dont like to go higher, curious to see if that can be fixed.

My latency at 5.0/3200 is ~50ns. My Z270 using the exact same kit at XMP is at 40.8ns Memory Latency ( Aida64 Benchmark ).

To me, it looks such, that the 8thgen IMC is not as tolerant as the 7thGen/Z270 combo is.

Carbonicdk
12-17-2017, 03:10 PM
It seems the boards dont like the CL16 settings at all. Fastest I can get stable in all tests is 5.0/3200CL16-16-16-36-2.

All 3600/CL16 in QVL have 16-18-18-36, none is a straight 16-16-16-36. Maybe the Z370/8thgen-IMC dont like to go higher, curious to see if that can be fixed.

My latency at 5.0/3200 is ~50ns. My Z270 using the exact same kit at XMP is at 40.8ns Memory Latency ( Aida64 Benchmark ).

To me, it looks such, that the 8thgen IMC is not as tolerant as the 7thGen/Z270 combo is.

Just tried 18-18-18-38 2ns at both 3600 and 3200mhz. 3600 didn't boot and 3200mhz crashed in Windows after 2 minutes.

mikex999
12-18-2017, 12:19 AM
still nothing more added to qvl. DDR4 support is still ridicoulus

Raja@ASUS
12-18-2017, 09:08 AM
still nothing more added to qvl. DDR4 support is still ridicoulus

New kits are seldom added to the QVL post launch timeframe. Not ideal, but nothing new, either.

mikex999
12-18-2017, 10:24 AM
New kits are seldom added to the QVL post launch timeframe. Not ideal, but nothing new, either.
was excpecting more ddr 4 support. expecially for the 4000mhz versions

freighters
12-18-2017, 12:12 PM
Version change description is simply insulting.

Carbonicdk
12-19-2017, 01:33 PM
was excpecting more ddr 4 support. expecially for the 4000mhz versions

If the board has trouble actually doing 4000mhz that's not surpricing.

mikex999
12-19-2017, 11:53 PM
If the board has trouble actually doing 4000mhz that's not surpricing.

why? :)

janos666
12-20-2017, 08:18 AM
Does anybody know the actual difference between these synchronously updated 5*** and 0*** version chains (besides the "beta" label hanging on the 5*** chain)? The official change log really doesn't help (it's pretty useless even withing the same chain, let alone in differentiating between the two of them).

Raja@ASUS
12-20-2017, 08:54 AM
Does anybody know the actual difference between these synchronously updated 5*** and 0*** version chains (besides the "beta" label hanging on the 5*** chain)? The official change log really doesn't help (it's pretty useless even withing the same chain, let alone in differentiating between the two of them).

The builds with the 5 prefix have ASUS MCE enabled by default. No other difference.

janos666
12-20-2017, 09:56 AM
The builds with the 5 prefix have ASUS MCE enabled by default. No other difference.

Ah! Thanks for the explanation. The officially published release notes make more sense now. I was confused because the feature is available with both variants and I failed to notice the difference in the defaults when I tried the "beta".

Tbolt214th
12-20-2017, 10:42 AM
New kits are seldom added to the QVL post launch timeframe. Not ideal, but nothing new, either.

The kit I am talking about is a year old and ran fine at XMP with my Asus Prime Z270-A, tho, as said earlier, it also took a few Bios updates on that board to support that specific RAM Kit flawless...but it is NOT new and a well kown and reviewed kit. Something I hoped would run OUT-OF-THE-BOX. I guess those times are gone and the product has to ripe with the customer. Not nice tbh.

To me, honestly speaking, this is due to the fact that AMD has stirred up the market with Ryzen and Intel is now pushing harder & faster than they should, the result is clearly visible, many new boards with many new chipsets flood the market and the dev teams struggle to support them all in a timely nice manner.

I dont blame Asus as #1, they are a victim of that circumstance as well, too many new products in too little time, the result is on your desk, with BSODs and XMP errors.

Carbonicdk
12-20-2017, 09:26 PM
The kit I am talking about is a year old and ran fine at XMP with my Asus Prime Z270-A, tho, as said earlier, it also took a few Bios updates on that board to support that specific RAM Kit flawless...but it is NOT new and a well kown and reviewed kit. Something I hoped would run OUT-OF-THE-BOX. I guess those times are gone and the product has to ripe with the customer. Not nice tbh.

To me, honestly speaking, this is due to the fact that AMD has stirred up the market with Ryzen and Intel is now pushing harder & faster than they should, the result is clearly visible, many new boards with many new chipsets flood the market and the dev teams struggle to support them all in a timely nice manner.

I dont blame Asus as #1, they are a victim of that circumstance as well, too many new products in too little time, the result is on your desk, with BSODs and XMP errors.

My memory works fine on the Z170 platform as well. It's just sad, really. Would be nice with some official comments about the memory stability issues and some solutions that might help for it and some timeframes.

Raja@ASUS
12-21-2017, 11:09 AM
My memory works fine on the Z170 platform as well. It's just sad, really. Would be nice with some official comments about the memory stability issues and some solutions that might help for it and some timeframes.


Your kit has been tested by one of the engineers and it checks out fine on our side. What happens when you reduce the SA and IO voltages below 1.225V? Try tuning from 1.10V upwards, in slow steps.

Carbonicdk
12-21-2017, 03:10 PM
Your kit has been tested by one of the engineers and it checks out fine on our side. What happens when you reduce the SA and IO voltages below 1.225V? Try tuning from 1.10V upwards, in slow steps.

Been trying a bunch of things, setting both IO+SA to:
1.1
1.11
1.13
1.15
1.16
1.17
1.18
1.20
1.22
1.23
and then IO+SA combinations like:
1.32 - 1.17
1.30 - 1.17
1.22 - 1.15
1.23 - 1.13

all resulting in failure to boot followed by the safe boot or immidiate crash on the load screen.

Have also tried memory voltages from 1.35v to 1.40v and loose timings like 20-20-20-40. Lowering the frequency to 3400mzh makes it possible to enter Windows where it soon crashes and 3200mhz crashes during gaming after like 20 minutes. Only 3000mhz is stable.

In short, I've never had this many issues getting a motherboard to play nice with memory before.

theWanderer521
12-26-2017, 12:17 AM
ROG Strix Z370-E / i7 8700k
Hi new member here. Upgraded my BIOS to 5605 but I have CPU temperatures fluctuating up until 85c just doing normal desktop things (browsing etc.).
PC isn't overclocked. Just built this one few weeks ago. All BIOS settings are in default.

Am I able to downgrade or Is it safe to downgrade to 0419 or from the old BIOS version included in the support DVD?

Or should I try to fix my BIOS settings on this version? Can anyone point me on any particular setting?

Thank you so much.

Raja@ASUS
12-26-2017, 03:58 PM
ROG Strix Z370-E / i7 8700k
Hi new member here. Upgraded my BIOS to 5605 but I have CPU temperatures fluctuating up until 85c just doing normal desktop things (browsing etc.).
PC isn't overclocked. Just built this one few weeks ago. All BIOS settings are in default.

Am I able to downgrade or Is it safe to downgrade to 0419 or from the old BIOS version included in the support DVD?

Or should I try to fix my BIOS settings on this version? Can anyone point me on any particular setting?

Thank you so much.

Use version 0605 instead of 5605, or disable ASUS Multicore Enhance in UEFI.

AntonioL
12-26-2017, 06:45 PM
Hi, have updated my z370-F to this latest 0605 BIOS, and this new version activates indeed Intel Speed Shift.
Coming from 0429, I had to completely change the voltage settings, but I think most differences in voltage management have been introduced in the previous version. Asus MCE no longer appears to boost all cores to max turbo ratio, applying a "per number of loaded cores" logic and following Intel specs. The cache frequency now is set accordingly to this same variable, decreasing to 40 when all cores are in use rather than 44, producing far less heat for hardly any drop in performance. I had no XMP stability issue either.

But with this new BIOS, something very strange is happening: OCCT now sometimes fail at stress launch with the message "an error was detected" giving no clue of what made it halt. If it does not stop during the first seconds, it will run without any problem for a long time. The error message seems to mainly occur when less than 12 threads are used, and also more often just after reboot
Other test softs run fine: IXTU, Realbench, Prime95, Linpack, Memtest... No system crash so far neither.
I have tried a lot of things: raising vcore, chosing manual vcore, reducing vdroop through LLC, setting all cores at 43, setting uncore at 40, switching off XMP... but nothing improves OCCT probability of failure when starting stressing the CPU.

Does anyone have an idea, please, of what could happen ? As temps are ok, would it be a wrong calculation detected by OCCT due to the sudden transition of cores from idle to busy ? Is there a way to solve this faulty computing ?

Thanks in advance for any help.*

theWanderer521
01-07-2018, 12:57 AM
Maybe due to the recent Meltdown and Spectre issues ~ ASUS has released another BIOS update (ROG STRIX Z370-E GAMING BIOS 0606) It says:

- "1. Update CPU Microcode

- Improve system compatibility and stability"

Anyone already tried it? Would love to hear your thoughts...thanks!

(EDIT)

Also after Windows Update (KB4056892) gui got affected - ASUS AI Suite 3 - "The server threw an exception error"

https://rog.asus.com/forum/showthread.php?98821-ASUS-AI-Suite-3-quot-The-server-threw-an-exception-error-quot

Hoping for ASUS to roll in a fix :)