Well ive got my ram 100% stable at 3600, 16,16,16,1T 1.5v on the taichi, this board is unbelievable. im now pushing for 3733mhz.

to achieve this leave advanced boot training to auto, and slightly raise the soc voltage.

ignore the above, although i managed 3600 ram in the end i had to bump the voltage up to 1.535, and performance gains at that speed were at best 1% so its not worth it, im now overclocking with bclk and the rams at 3520mhz at 1.43v with slighly tighter timings, seems on ryzen tighter timings is more important than clock speeds with ram.

Anyway ill change what i said about BCLK overclocking being useless, it aint on this board! damn its unreal. in fact i would go as far as to say it beats multiplier overclocking for speed, for the sme clock speed im getting way more gflops in intel burn test avx, and the whole board runns hell of a lot cooler!

I think you can see how much i like this motherboard, at last i have a board with no shitty uefi, no giimicks a great bclk chip and monster performance! damn its ace. the built in wifi module on this is the best ive ever used its a top end intel bluetooth wifi dual antenna combo, even with the top lid down on my thermaltake tower 900 case (its metal mesh too) i get full signal strength from 2 large rooms away from the router about 35 feet

also note on this board bclk overclocking still alows the cpu to idle! even if you change the vcore voltage manually.

using the regular p-state overclocking at 3.8ghz im getting the same gflops in IBT as using bclk at 3.55ghz! and as i say the whole board stays really cold. now i know why these boards are expensive. ill also say, since i used the bclk method for overclocking, the temps in the uefi have gone down by a massive 12c, and boot times have drastically decreased, i can get in to windows 10 in 6 seconds, before was 10 seconds.

Zitat von POE_UK im Beitrag #375

---

Zitat von ket im Beitrag #365

---

Updated Realtek ALC1220 driver to 8328, see post #3

Probably last audio driver update I'll do on-board audio rarely has any issues so this latest driver update should ensure any kinks there might of been are now worked out.

---

Ill give this a try on my taichi, theres no issues with the bog standard driver, but its just that, bog standard, hopefully the skin is different on this driver, thats all im looking for really, a different skin over the standard boring realtek one.

---

Zitat von POE_UK im Beitrag #374

---

Ket i dont know where i got that from i mean 3.8 hahaha,

anyhoo - theres only one thing that bugs me about the taichi board, must be the only am4 board that doesnt have a bios logo editor software!

the realtek codec software is the bog standard one with realtek branding, but i tell ya the audio quality on this board is epic.

---

Zitat von POE_UK im Beitrag #376

---

Well ive got my ram 100% stable at 3600, 16,16,16,1T 1.5v on the taichi, this board is unbelievable. im now pushing for 3733mhz.

to achieve this leave advanced boot training to auto, and slightly raise the soc voltage.

ignore the above, although i managed 3600 ram in the end i had to bump the voltage up to 1.535, and performance gains at that speed were at best 1% so its not worth it, im now overclocking with bclk and the rams at 3520mhz at 1.43v with slighly tighter timings, seems on ryzen tighter timings is more important than clock speeds with ram.

Anyway ill change what i said about BCLK overclocking being useless, it aint on this board! damn its unreal. in fact i would go as far as to say it beats multiplier overclocking for speed, for the sme clock speed im getting way more gflops in intel burn test avx, and the whole board runns hell of a lot cooler!

I think you can see how much i like this motherboard, at last i have a board with no shitty uefi, no giimicks a great bclk chip and monster performance! damn its ace. the built in wifi module on this is the best ive ever used its a top end intel bluetooth wifi dual antenna combo, even with the top lid down on my thermaltake tower 900 case (its metal mesh too) i get full signal strength from 2 large rooms away from the router about 35 feet

also note on this board bclk overclocking still alows the cpu to idle! even if you change the vcore voltage manually.

using the regular p-state overclocking at 3.8ghz im getting the same gflops in IBT as using bclk at 3.55ghz! and as i say the whole board stays really cold. now i know why these boards are expensive. ill also say, since i used the bclk method for overclocking, the temps in the uefi have gone down by a massive 12c, and boot times have drastically decreased, i can get in to windows 10 in 6 seconds, before was 10 seconds.

---

The strixx isnt bad, but it isnt great either, for starters there is no usb 3.1 and its not really an outstanding overclocker. its just a pretty standard performing board that performs slightly less than the gigabyte k7

asrocks bios 3.20A gets me up to 182Gflops on stock cooling, with intel burn test, i managed 173 on the asus and 174 on the gigabyte.

Strix does have USB 3.1 Gen1 and 2, something was seriously wrong with your Strix. Clicky.

Was something to do with usb 3.1 crapping out when overclocking if i remember right, and the heatsinks got red hot. but i had this board a while ago it "might" have those issues sorted now. its not a bad board but no brilliant either. i expected more with the price of it. from the point of view of someone who pays cash for hardware not test samples i was disappointed ive yet to find any board that matches the asrock, even though the bundled items are very minimal and sparse, the new v3.20A update really made this great board greater! if your only goal is overclocking its the one to go for no doubt about that, all the moneys gone on power delivery and stability, theres only 1 thing that i dont like about the taichi, the vrm heatsinks look like coffin's haha. apart from that its been flawless even beating asus's most expensive 370 board. i also found the suppplied cd to the strixx and hero to be nothing but bloatware and heavy on resources that when run, keeps the cpu out of p-state2 at all times making the voltages fluctuate.

Another limitation of the asus is it has no boot training, meaning the ram will always boot at 1.2v and force constant boot looping, meaning from cold boots there will be a problem with 1.35v overclocked ram forcing jedec values to load. so you will have to load the ram overclock everytime you boot up with trident z

Hello again!

Turned out that "solution" provided by GIGABYTE support isn't effective (recording works right after Windows booted up and breaks few minutes later when I want to actually record something, even with totally all power-saving options disabled. I also installed fresh Windows 7 to see if it changes something, but nope, no difference) and after 3 weeks of mailing to them, they are still deaf to my requests to add PCI Subsystem Setting menu to new BIOS release or even look for this workaround that @ket said about. I only got brief answer from them "Thank you for additional information. We try to solve this issue ( if possible) in new BIOS relase." and that was two weeks ago. I starting to lose hope that they will do something to fix this.

@ket I made screenshot of every page in BIOS (don't worry that current version I have is F22e - CPU microcode update only). So if you could look if it's possible to expose these options, I would be very grateful.

M8 they wont fix anything they're the worst company on earth for end users.

Zitat von tyuper im Beitrag #381

---

Hello again!

Turned out that "solution" provided by GIGABYTE support isn't effective (recording works right after Windows booted up and breaks few minutes later when I want to actually record something, even with totally all power-saving options disabled. I also installed fresh Windows 7 to see if it changes something, but nope, no difference) and after 3 weeks of mailing to them, they are still deaf to my requests to add PCI Subsystem Setting menu to new BIOS release or even look for this workaround that @ket said about. I only got brief answer from them "Thank you for additional information. We try to solve this issue ( if possible) in new BIOS relase." and that was two weeks ago. I starting to lose hope that they will do something to fix this.

@ket I made screenshot of every page in BIOS (don't worry that current version I have is F22e - CPU microcode update only). So if you could look if it's possible to expose these options, I would be very grateful.

---

I should've been explain that earlier, my apologies.

From TechARP BIOS Optimization Guide

Zitat

---

PCI Latency Timer controls how long a PCI device can hold the PCI bus before another device takes over. The longer the latency, the longer the PCI device can retain control of the bus before handing it over to another PCI device.

---

There are reports from users with X-Fi cards which were claiming that changing PCI Latency Timer default value 32 to higher 64 or even 96 resulted in dissappearance of issues like stuttering, sound cutting etc. Of course that was on motherboards with native PCI bus implementation (before Sandy Bridge and Z68 chipset era) and I'm really curious (and hopeful) about result on non-native PCI bus implementation, with 3rd party (ASMedia ASM1083) PCIe-to-PCI bridge chip in operation.

Anyone been able to stabilize Voltage on Aorus X370 Gaming-K5??? I'm on F4 BIOS currently doing some testing... I've tried F5a & F10 with no success... Specs are down below... Even if I set the voltage in BIOS or Ryzen Master it doesn't stick... It keeps droping to ~0.400...

P.S. Any workaround for Win10 CFU forcing Core-Parking???

Zitat von @JSebastian im Beitrag #385

---

Anyone been able to stabilize Voltage on Aorus X370 Gaming-K5??? I'm on F4 BIOS currently doing some testing... I've tried F5a & F10 with no success... Specs are down below... Even if I set the voltage in BIOS or Ryzen Master it doesn't stick... It keeps droping to ~0.400...

P.S. Any workaround for Win10 CFU forcing Core-Parking???

---

Zitat von tyuper im Beitrag #384

---

I should've been explain that earlier, my apologies.

From TechARP BIOS Optimization Guide

Zitat

---

PCI Latency Timer controls how long a PCI device can hold the PCI bus before another device takes over. The longer the latency, the longer the PCI device can retain control of the bus before handing it over to another PCI device.

---

There are reports from users with X-Fi cards which were claiming that changing PCI Latency Timer default value 32 to higher 64 or even 96 resulted in dissappearance of issues like stuttering, sound cutting etc. Of course that was on motherboards with native PCI bus implementation (before Sandy Bridge and Z68 chipset era) and I'm really curious (and hopeful) about result on non-native PCI bus implementation, with 3rd party (ASMedia ASM1083) PCIe-to-PCI bridge chip in operation.

---

I'm aware of the function it's my fault for not clarifying what I meant, I was tired when I made the post. I was saying if you wanted the latency decreased its already at its lowest. Trying to edit the firmware is a no-go anyway sadly, every tool I have crashes when trying to save any changes. The whole structure of the firmware in fact is a complete mess I'm surprised the Gigabyte firmware works at all.

Zitat von ket im Beitrag #386

---

I'm aware of the function it's my fault for not clarifying what I meant, I was tired when I made the post. I was saying if you wanted the latency decreased its already at its lowest. Trying to edit the firmware is a no-go anyway sadly, every tool I have crashes when trying to save any changes. The whole structure of the firmware in fact is a complete mess I'm surprised the Gigabyte firmware works at all.

---

Zitat von ket im Beitrag #386

---

[quote=@JSebastian|p47196]Anyone been able to stabilize Voltage on Aorus X370 Gaming-K5??? I'm on F4 BIOS currently doing some testing... I've tried F5a & F10 with no success... Specs are down below... Even if I set the voltage in BIOS or Ryzen Master it doesn't stick... It keeps droping to ~0.400...

P.S. Any workaround for Win10 CFU forcing Core-Parking???

---

Zitat von ket im Beitrag #386

---

Sounds like you are using offset voltage which allows the CPU to reduce frequency and voltage when not under load. If I remember correctly simply setting the high performance power profile in windows will disable core parking.

---

Zitat von @JSebastian im Beitrag #388

---

Keyboard and Mouse USB receiver crash the USB ports sometimes too...

---

Zitat

---

Fault bucket , type 0
Event Name: LiveKernelEvent
Response: Not available
Cab Id: 0

Problem signature:
P1: 144
P2: 100d
P3: fffff888075e9970
P4: 0
P5: 0
P6: 10_0_16299
P7: 0_0
P8: 256_1
P9:
P10:

Attached files:
\\?\C:\WINDOWS\LiveKernelReports\USBXHCI\USBXHCI-20180201-0148.dmp
\\?\C:\WINDOWS\TEMP\WER-235791625-0.sysdata.xml
\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERF024.tmp.WERInternalMetadata.xml
\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERF024.tmp.csv
\\?\C:\ProgramData\Microsoft\Windows\WER\Temp\WERF055.tmp.txt

These files may be available here:
C:\ProgramData\Microsoft\Windows\WER\ReportQueue\Kernel_144_f89244eec9dc8531bf7d9cece3aff43e287c97d_00000000_cab_2009f053

Analysis symbol:
Rechecking for solution: 0
Report Id: e0022968-8340-44b2-bd72-3a078c9239e1
Report Status: 100

---

http://download.gigabyte.cn/FileList/BIOS/?C=M;O=D

New updates...


Update AGESA 1.0.0.0a for new upcoming processors
Supports GPU overclockable on Raven Ridge processors

🤣😂🤣😂