Any ASUS Prime X370 Pro owners out there ... besides me??

[DaveB]

Currently, I'm more concerned about finding stable income, then I'll worry about credit.

Hey maybe you have to move to where the stable jobs are. After 25 years of full-time employment, all the big more stable corporations started to move out of where I grew up. So, before I got layed off, I put in for a transfer to where I am now and got in my final 23 years to get to retirement. I would have moved on my own, but going through the company got me some relocation costs so it was a good deal. Where I live now, taxes and housing costs were lower so a win, win, win.

 

[Keith Myers]

I'm well aware that not having a credit history can affect being able to buy things like a home or car. Currently, I'm more concerned about finding stable income, then I'll worry about credit.

Anyhow, back on topic... Does anyone know what key differences there is between BIOS 3803 and 3203? Asus mentions AGESA codes, but it seems the 3803 that I'm using has a number that's lower than the AGESA code in 3203. Since Asus doesn't seem to want to publish in depth notes on this, makes it hard to determine what version is best for Ryzen, versus Raven Ridge.

The 3803 BIOS is what would have been published as 1.0.7.1. But since the only difference between 3803 and the previous BIOS' was to incorporate the completely reworked AGESA code to handle the new APU's, ASUS decided to label it 1.0.0.0a. IOW,the first revision of the new AGESA platform. I have seen no benefits in updating one of my Prime Pro's to 3803. No better memory or overclock support. A few things taken away in the BIOS and locked down from prior versions. I am keeping the other Prime Pro on 3203 until some further improvement from other testers that would make another update worthwhile.

 

[Woomack]

If BIOS is adding only APU support then better don't touch it unless you want to use APU. The same as with other brands, there can be little issues with BIOS releases for APU. Just blame AMD for that ( already mentioned why in couple of other threads ).

 

[Thentilian]

Hey maybe you have to move to where the stable jobs are. After 25 years of full-time employment, all the big more stable corporations started to move out of where I grew up. So, before I got layed off, I put in for a transfer to where I am now and got in my final 23 years to get to retirement. I would have moved on my own, but going through the company got me some relocation costs so it was a good deal. Where I live now, taxes and housing costs were lower so a win, win, win.

Can't move until I find something to save money. Stuck here til then.

The 3803 BIOS is what would have been published as 1.0.7.1. But since the only difference between 3803 and the previous BIOS' was to incorporate the completely reworked AGESA code to handle the new APU's, ASUS decided to label it 1.0.0.0a. IOW,the first revision of the new AGESA platform. I have seen no benefits in updating one of my Prime Pro's to 3803. No better memory or overclock support. A few things taken away in the BIOS and locked down from prior versions. I am keeping the other Prime Pro on 3203 until some further improvement from other testers that would make another update worthwhile.

If BIOS is adding only APU support then better don't touch it unless you want to use APU. The same as with other brands, there can be little issues with BIOS releases for APU. Just blame AMD for that ( already mentioned why in couple of other threads ).

Yeah, wasn't sure. I knew they had APU code in there, but was hoping for more stability for Ryzen. It seems to have changed nothing but a few of my OC settings had to be tweaked.

I'm not having issues with 3803, so I'll just leave it be for now.

 

[Keith Myers]

Use the modded BIOS' to get full control

Can't move until I find something to save money. Stuck here til then.





Yeah, wasn't sure. I knew they had APU code in there, but was hoping for more stability for Ryzen. It seems to have changed nothing but a few of my OC settings had to be tweaked.

I'm not having issues with 3803, so I'll just leave it be for now.

The only way to get full control in the BIOS over your chip and motherboard is to use the modded BIOS' that unlock or unhide all the settings. Primarily in the AMD CBS section but also in the memory section. Both Lermite and Reous have modded BIOS. You have to install them with the AFUDOS tool which does come with some dangers. Also you can't update a regular BIOS over the modded BIOS without using the tool again because the modded BIOS don't have the AMD signature.

 

[Keith Myers]

Lost some . . . . gained some ?!

I just swapped my week 7 Ryzen 1700X with a week 37 1700X via the AMD CPU warranty action due the segmentation violation flaw in the early pre-week 25 chips.

Tried to buy a 1800X several times from Amazon but kept getting pre-week 25 chips. That action was to avoid the downtime on the system while awaiting the replacement chip. Finally got AMD to arrange advance replacement at the end of the initial RMA period which was about to expire. So swapped the chips.

I appear to have lost in the silicon lottery in one area and gained in another. I can now run 3950Mhz at 1.35V BOINC stable where I earlier could only run 3900Mhz at 1.38V BOINC stable. That is the gain.

The loss is that I can't run my previous 3333Mhz CL14 stable memory clock anymore. I tried every trick in the memory tuning book and I could not get back to 3333. Had to settle for 3200Mhz CL14 for BOINC stable.

 

[Bluefalcon13]

So I think I missed the boat or something.... Ended up updating my bios to 3803, ram runs great @ 3000 instead of 2933 (was on 090x I think...)

Went to set my OC back up, and low and behold, only multiplier settings were the FID/DID... No more CPU ratio. How do these work?

 

[mackerel]

Went to set my OC back up, and low and behold, only multiplier settings were the FID/DID... No more CPU ratio. How do these work?

Effective multiplier is related to bigger number divided by smaller number. From memory, the smaller number could be left at 8, and just tinker with the higher number. If you're like me and don't want to do the calculations, the bios shows the resulting clock at the top.

 

[Keith Myers]

So I think I missed the boat or something.... Ended up updating my bios to 3803, ram runs great @ 3000 instead of 2933 (was on 090x I think...)

Went to set my OC back up, and low and behold, only multiplier settings were the FID/DID... No more CPU ratio. How do these work?

Are you sure it is not there?? It still is in mine. I have never used the FID/DID ratios since I have never used the P-states. I just run all cores flat out overclocked at all times. I just set the multiplier to 39.5 in Custom CPU Core Ratio [Auto]

Only when you select Manual does the FID/DID ratio input boxes appear. This is in BIOS 3803. Same as every BIOS previously. Wonder when we get a new one? Been a while.

 

[Bluefalcon13]

Are you sure it is not there?? It still is in mine. I have never used the FID/DID ratios since I have never used the P-states. I just run all cores flat out overclocked at all times. I just set the multiplier to 39.5 in Custom CPU Core Ratio [Auto]

Only when you select Manual does the FID/DID ratio input boxes appear. This is in BIOS 3803. Same as every BIOS previously. Wonder when we get a new one? Been a while.

I just want you to know, you are absolutely correct, and I fell off the short bus as a kid [emoji23]

 

[mackerel]

I just want you to know, you are absolutely correct, and I fell off the short bus as a kid [emoji23]

I must have fallen off at the same time. Mine is on 3803 and when overclocking it last week for the Superposition comp, I also only saw the FID/DID values.

 

[Bluefalcon13]

I must have fallen off at the same time. Mine is on 3803 and when overclocking it last week for the Superposition comp, I also only saw the FID/DID values.

Oh it's there... Donno if you have checked yet. Decided to lean on her a bit tonight... Trying for a 3.9ghz before I head to bed [emoji23]

 

[mackerel]

I'm not going to revisit it today, but I have a mini-vacation soon making next weekend a long weekend, and will see how far I get with OC there. Think I got my 1700 to 3.9 before but it was silly volts needed and cooling was marginal...

 

[Bluefalcon13]

I'm not going to revisit it today, but I have a mini-vacation soon making next weekend a long weekend, and will see how far I get with OC there. Think I got my 1700 to 3.9 before but it was silly volts needed and cooling was marginal...

Well.. I failed. Gave up after a while. Seems my ram ISN'T stable at DOCP 2933 or 3000 anymore. :/

Up to 3.8ghz now. Currently testing stability for 3.8 @ 1.2475V and LLC @ 1. SOC is at 1.1V, and it's LLC is at Auto.

 

[Keith Myers]

Try backing down the Vsoc voltage from Auto. The 3803 BIOS likes a lot less SoC voltage for even the highest memory clocks. I'm manually inputting 1.025V for my 3333Mz CL14 memory clocks. The memory voltage also got dropped from 1.39V to 1.36V. I also am using only 12mV offset for my 3.95Ghz cpu core clock instead of the 18mV offset I was previously using for BIOS 3203 I upgraded from.

 

[Bluefalcon13]

Try backing down the Vsoc voltage from Auto. The 3803 BIOS likes a lot less SoC voltage for even the highest memory clocks. I'm manually inputting 1.025V for my 3333Mz CL14 memory clocks. The memory voltage also got dropped from 1.39V to 1.36V. I also am using only 12mV offset for my 3.95Ghz cpu core clock instead of the 18mV offset I was previously using for BIOS 3203 I upgraded from.

Thanks. I'll give it a whirl later this week. Two midterms to prep for [emoji50]

 

[Aballarde]

Hi there Keith Myers

I just purchased one ASUS PRIME X370- PRO. Installed all ASUS drivers and make some basic adjustments in Bios. I build this pc mostly for game and some programs testing, but taking in mind users with tight budget. So I gathered all and come out with a pretty decent rig in a twelve hundred bucks budget.

ASUS PRIME X370-PRO
AMD RYZEN 3 @ 1200 3.21 MHz
AMD RADEON RX460 (2@ 2gb RAM each) cross fire
TEAM T-FORCE DARK DDR4 16 Gb (2x8)
900 WATTS PSU
2 1T WESTERN DIGITAL DRIVES
AIR COOLING x 4 200mm each
RAIDMAX MID SIZE TOWER CASE
MICROSOFT WINDOWS 10 X64 BIT

 

[Keith Myers]

Congratz on the new build. You should get years of service out of it. I speculate how much the first Gen processors will come down in price when the Gen 2 processors hit the market in April. There have already been some single day fire-sale prices on the 1800X. That said, I have come to even more appreciate the economics of a Ryzen build now that I have completed my first Intel build in 20 plus years. I was gifted a Intel X99 workstation motherboard and spent a fortune getting the rest of the pieces to finish it and make it functional. I use my computers for distributed computing, so no gaming, but tons of gpus. I'm a bit chagrined in seeing the actual cpu compute performance of the Intel system and it has half the floating point performance of my 1800X and at similar core clocks. The memory is clocked mostly the same too. I had heard for years that Intel always cleaned house on AMD for compute performance. I'm not finding that is the case.

Getting ready to rebuild my two oldest FX systems with the new Ryzen+ processors and X470 motherboards in April. Can't wait.

 

[Thentilian]

I just swapped my week 7 Ryzen 1700X with a week 37 1700X via the AMD CPU warranty action due the segmentation violation flaw in the early pre-week 25 chips.

Tried to buy a 1800X several times from Amazon but kept getting pre-week 25 chips. That action was to avoid the downtime on the system while awaiting the replacement chip. Finally got AMD to arrange advance replacement at the end of the initial RMA period which was about to expire. So swapped the chips.

I appear to have lost in the silicon lottery in one area and gained in another. I can now run 3950Mhz at 1.35V BOINC stable where I earlier could only run 3900Mhz at 1.38V BOINC stable. That is the gain.

The loss is that I can't run my previous 3333Mhz CL14 stable memory clock anymore. I tried every trick in the memory tuning book and I could not get back to 3333. Had to settle for 3200Mhz CL14 for BOINC stable.

Just saw this post. I stopped getting thread updates on this page for some reason. Anyhow, how do you find out the week the chip is made? And does this affect all the CPU's pre-week 25? I got my R5 1600 back in August, so it might be a pre-week 25 chip.

If I recall correctly, this flaw was only affecting Linux users anyways right?

 

[Keith Myers]

Just saw this post. I stopped getting thread updates on this page for some reason. Anyhow, how do you find out the week the chip is made? And does this affect all the CPU's pre-week 25? I got my R5 1600 back in August, so it might be a pre-week 25 chip.

If I recall correctly, this flaw was only affecting Linux users anyways right?

You have to look at the batch code number printed on the IHS. Like my first 1700X was Batch 1707 or week 7 of 2017. If I remember right, week 25 was second week of July. Only way to be sure is look which means pulling the heat sink off and scrubbing off the heat sink compound. The replacement cpu got its picture taken with the cell phone so I know what week it was made without having to disassemble.

The consensus I have read is that the segmentation violation is NOT affecting ONLY Linux users. They are just the likely ones to trip up the cpu because they would be the ones heavily loading the cpu with multi-thread workloads like compiling kernels. If you are similarly heavily loading the cpu with multi-threaded work in the Windows environment, you could have the seg fault too. I am that case in fact since I run the Windows system just as hard as the Linux system with distributed computing on all cores 24/7. I had segfaults on my workload tasks. The tasks write out the error dump down to the registers values and it is pretty easy to see the condition that threw the upset. So I took advantage of the RMA offered.