AMD AGESA 1.0.0.3ABA Buggy, Company Pulls it from Motherboard Vendors

Rockenrooster said:
my X370 ASrock only has 16MB EEPROM and it supports ryzen 3000 lol!
I will say ASrock has been killing it with Ryzen.
They will get my money.
Click to expand...

I've bought a lot of ASRock because they had the features I wanted at prices I was willing to pay- and I've had relatively few problems with them. They still have that stigma of being a budget brand, but they also do a lot of things well as well as things that other vendors just won't.
 
Dan_D said:
Fair enough. One point about that example, even Apple screws up from time to time. I remember a rash of iMacs with bad capacitors years after the capacitor issues had been corrected by everyone else in the industry.
Click to expand...
There will always be exceptions to things. But part of Apples secret to success has been hardware control. And they’ve never deviated or at least not for very long.

And that’s partly why they can have such a high market share in cell phones. And also why they have excellent legacy support.

But I digress. I’m not tooting their horn though, they do plenty of things that aren’t perfect by any means.
 
Last edited:
Keljian said:
This so much this




And Linux, and 4 dimms, and latest bioses, and Nvidia cards (reliably) and .(as I said in another thread)..

  • Destiny 2
  • Linux
  • Memory card reader compatibility issues
  • Nvidia whea
  • Ryzen’s vr incompatibilities (vive)
  • Scheduler issues
  • Ram compatibility (including latency). Write latency on single die 3xxx models.
  • Issues with the clock of the infinity fabric
  • Issues with PBO not working properly
  • Sata issues
  • Incompatibilities with soundcards and some other peripheral cards
  • Incomplete ECC implementation
  • Linux compilation issues
  • Revision issues (some things were fixed and no one spoke about them) - consumers didn’t know which to get
  • Random software incompatibilities (Adobe during the last year)
  • Temperature reporting problems
  • Over reliance on AGESA to fix things (which means putting a lot of pressure on motherboard manufacturers to release bios updates)

It will end, this stuff will get better, it just takes time
Click to expand...

Looking at my motherboard, an ASRock X370 Killer, there are a lot of BIOS updates and I may be incorrect but after doing some BIOS upgrades I have to install something within Windows and then upgrade to another bridge BIOS, then upgrade to the latest version. Seems a bit risky and will wait and see how it pans out. Looks like waiting a few months if you're on an older board would be ideal.

The 3700X was a disappointment, hardly any real world improvement over the 2700X so I may just continue using it and not upgrade. I'll see what Intel has next year and maybe the year after that. Doubting AMD will see any substantial gains within 2 years. The good news is the 2700X is a fairly decent CPU as is.
 
All my AMD systems I have built, I built and never looked back. Shit just works same as any other. Intel I use at work.. its nothing impressive not even in the light tasks we do. Might be budget Dell, but if we are going to shit post AMD on the DIY market composed of some many vendors that may or might not be at fault then I am shit posting Intel and their shit hardware.
 
Uvaman2 said:
but if we are going to shit post AMD on the DIY market composed of some many vendors that may or might not be at fault then I am shit posting Intel and their shit hardware.
Click to expand...

You seem to have gotten your feelings intertwined with facts :ROFLMAO:

Uvaman2 said:
So AMD gave them updates to work with, they push it out no validation= AMD sucks?
Click to expand...

Only if you're not paying attention.
 
thebufenator said:
That is not accurate at all.

~12-15% clock per clock performance boost, and much lower power draw + higher clocks.
Click to expand...

Well, you can certainly benchmark that, but for most 'real-world' usecases you're not going to 'feel' the difference. Where you do, you'd need to be butting up against the limitations of the 2700X.
 
IdiotInCharge said:
Well, you can certainly benchmark that, but for most 'real-world' usecases you're not going to 'feel' the difference. Where you do, you'd need to be butting up against the limitations of the 2700X.
Click to expand...

Sure, but the 3700x draws something like 90 watts less at load (provided its running on an x470).

I don't know about you, but my office will notice things like that temp wise
 
thebufenator said:
Sure, but the 3700x draws something like 90 watts less at load (provided its running on an x470).

I don't know about you, but my office will notice things like that temp wise
Click to expand...

No, no, we're handwaving real power usage ;)

But yeah. 90 watts is definitely enough to make a pretty large difference in ambient.
 
IdiotInCharge said:
Well, you can certainly benchmark that, but for most 'real-world' usecases you're not going to 'feel' the difference. Where you do, you'd need to be butting up against the limitations of the 2700X.
Click to expand...

Entirely depends on which application. There is a big difference in highly threaded productivity apps. Heavy clock dependent things like Photoshop benefit from the IPC and clock speed boosts. Gaming above 1080p and general use won’t have a ton of real-world benefits however.
 
Flogger23m said:
Looking at my motherboard, an ASRock X370 Killer, there are a lot of BIOS updates and I may be incorrect but after doing some BIOS upgrades I have to install something within Windows and then upgrade to another bridge BIOS, then upgrade to the latest version. Seems a bit risky and will wait and see how it pans out. Looks like waiting a few months if you're on an older board would be ideal.

The 3700X was a disappointment, hardly any real world improvement over the 2700X so I may just continue using it and not upgrade. I'll see what Intel has next year and maybe the year after that. Doubting AMD will see any substantial gains within 2 years. The good news is the 2700X is a fairly decent CPU as is.
Click to expand...

If "real world" is browsing facebook and playing fortnite, then you won't notice much difference between a i3 and a i9 lol.

The 3700x is a great upgrade to the 2700x: 2x FP 15% IPC + higher clocks, lower power.
Dude WTH?
 
IdiotInCharge said:
Well, you can certainly benchmark that, but for most 'real-world' usecases you're not going to 'feel' the difference.
Click to expand...

I don't know what you're 'feeling,' but you are not supposed to insert the CPU or other parts of the computer into your body. To each their own, I guess. ;) Personally, I buy computer parts based on cost and how quickly they do the things I want them to do.

The Intel fan club's biggest complaint with Ryzen was it's lower IPC/clocks. Now that AMD has largely closed that gap it doesn't matter?
 
DeathFromBelow said:
The Intel fan club's biggest complaint with Ryzen was it's lower IPC/clocks. Now that AMD has largely closed that gap it doesn't matter?
Click to expand...

Depends. When considering the 'appeal' of a CPU you have to consider the potential market- anyone with an 8600 or better that rates gaming as their most intensive desktop task is really not going to be incentivized to look closely at any Ryzen CPU. If they have less than an 8600, then maybe, depending on their workload.

But this gets beyond benchmarking and toward what people actually do and toward what upgrade cycles look like.

For most, throwing more than six cores into a socket is going to do all of squat, and for those that already have higher-clocking six-core CPUs, at best, they won't see a performance downgrade for what they do most by swapping in Ryzen for more cores. They also won't see an upgrade.

And again, unless they were pushing the limits of what they have, and most aren't, then they're not going to see a difference at all.
 
DeathFromBelow said:
The Intel fan club's biggest complaint with Ryzen was it's lower IPC/clocks. Now that AMD has largely closed that gap it doesn't matter?
Click to expand...

Individually IPC or Clocks don't mean squat. It's the combination that matters. The per thread performance.

With the FX chips AMD could deliver the clocks but was way way behind on the IPC front.

I think everyone assumed that if AMD improved IPC they would hit the same clocks they did before, but that obviously was not the case.

Now AMD is actually leading in IPC in most workloads, but the clocks are holding them back.

The end result is that they are just shy of Intel on a per thread basis. On the flip side they absolutely crush Intel in many thread capable loads, but as has been pointed out, most people don't care about many threaded loads. in their daily work they don't push their CPU's at all. The only time they do is with certain games, and most of those see no benefit from adding cores somewhere afelter 4-6 of them.

Either way it is moot to me. I game at 4k. I'm not going to see a difference between most modern CPU's in games as I am going to be GPU limited.

The only people who will really notice a difference are the 1080p 240hz crowd, and they are an extreme minority.
 
Last edited:
tenor.gif
 
  • Like
Reactions: blkt
like this
IdiotInCharge said:
I've bought a lot of ASRock because they had the features I wanted at prices I was willing to pay- and I've had relatively few problems with them. They still have that stigma of being a budget brand, but they also do a lot of things well as well as things that other vendors just won't.
Click to expand...

They have the stigma of being a budget brand... because they are a budget brand.

I got an AB350M-Pro4 mainboard early last year, and ripped the 2nd x16 PCIe (the one with the x8 # of pins) slot off of the board by accident removing the video card. I didn't work at it or anything, it just came off. Apparently on mine, there is a single plastic stud at the back of the socket that holds it to the mainboard, and that stud somehow got cracked or broken leaving NO support. So when I pulled the card out, the tension on the slot connector was greater than the tension of the x8 PCIe number of pins holding it on the board. I've NEVER had this happen on any of the MSI, Gigabyte or Asus boards I've used.

Fortunately NewEgg RMA'd it. I sold it off as soon as I got it back and got something else instead. They had their chance, they blew it.
 
mvmiller12 said:
They have the stigma of being a budget brand... because they are a budget brand.

I got an AB350M-Pro4 mainboard early last year, and ripped the 2nd x16 PCIe (the one with the x8 # of pins) slot off of the board by accident removing the video card. I didn't work at it or anything, it just came off. Apparently on mine, there is a single plastic stud at the back of the socket that holds it to the mainboard, and that stud somehow got cracked or broken leaving NO support. So when I pulled the card out, the tension on the slot connector was greater than the tension of the x8 PCIe number of pins holding it on the board. I've NEVER had this happen on any of the MSI, Gigabyte or Asus boards I've used.

Fortunately NewEgg RMA'd it. I sold it off as soon as I got it back and got something else instead. They had their chance, they blew it.
Click to expand...

I think they have some good and some bad boards.

The only ASRock board I have had any personal experience with has been my Asrock H270M-ITX/ac.

I only bought it because it was the only mini ITX board I could find for my i3-7100 which had dual on board Intel NIC's. It's been a pretty impressive little board for my pfSense router.

I've also heard good things about Asrock Rack products.

I have also heard plenty of bad experiences.

My conclusion is that their quality is inconsistent. But then again, that can be said for even the best brands. I briefly used an Asus MicroATX Haswell board for a HTPC project. Hated it so much I returned it. In general Asus boards are great though.
 
All that said, just about all motherboard vendors today have boards across the entire price range. (Except Biostar. They only sell junk)

The quality ones cost more than the budget ones. Quality motherboards have cost $350+ now for almost 10 years.

Most complaints I see are from people who buy the cheap $129 boards.

At some point you have to admit that you kind of get what you pay for.
 
illli said:
Between this and MSI cutting corners trying to shoehorn bios updates into their mobos... B/c they only used 16-megabyte EEPROMs, I'm thinking my next choice will be ASRock
Click to expand...

Every motherboard manufacturer cuts corners somewhere. ASRock does it too. Their PCB's are pretty much the bare minimum thickness they can get by with on most models. On cheaper models, their component selection is the technological equivalent of mystery meat. Is it good? Maybe...... What is it? I have no idea. Each manufacturer picks and chooses where they can save money. Some things, every manufacturer is guilty of doing. All of them use cheap audio CODECs on most models as an example of what I mean. Most quit using socketed BIOS ROMs although I think ASRock still does it and ASUS still offers this on a few select models. Most haven't done it for years.

Dan said:
shit man, i need to play destiny 2...
Click to expand...

I did testing for the 3900X review but couldn't get it to run on that CPU which is now a known issue. Not having enough time to troubleshoot it I dropped it thinking it was something weird going on with my particular test system and moved on. I found out a day or two later that this was an issue with all Ryzen 3000 series CPU's. I'm getting the itch to buy a new CPU as my Threadripper doesn't cut it for D2 @ 4K. Which right now is the game I play the most.
 
Keljian said:
Dan_D

Vive issues- they actually refunded a whole lot of people who bought vives because of this issue: https://community.viveport.com/t5/T...ncompatible-with-RYZEN-cpus/td-p/23285/page/7

Linux issues:

Sata issues:
(There are more, just not digging now)

ECC: http://www.hardwarecanucks.com/foru...ws/75030-ecc-memory-amds-ryzen-deep-dive.html
Click to expand...

Like I said on the ECC front, its up to the motherboard manufacturers to validate and enable the feature. AMD has been pretty up front about this.
 
Dan said:
shit man, i need to play destiny 2...
Click to expand...

Dan_D said:
I did testing for the 3900X review but couldn't get it to run on that CPU which is now a known issue. Not having enough time to troubleshoot it I dropped it thinking it was something weird going on with my particular test system and moved on. I found out a day or two later that this was an issue with all Ryzen 3000 series CPU's. I'm getting the itch to buy a new CPU as my Threadripper doesn't cut it for D2 @ 4K. Which right now is the game I play the most.
Click to expand...

Dans, I have to be honest.

I don't remember ever hearing of Destiny 2 or the Destiny series at all until the Ryzen 3000 reviews mentioned that it wasn't working :p
 
Zarathustra[H] said:
Dans, I have to be honest.

I don't remember ever hearing of Destiny 2 or the Destiny series at all until the Ryzen 3000 reviews mentioned that it wasn't working :p
Click to expand...

The first one was on consoles only. The 2nd one has had a surge in popularity recently and a bunch of changes an an expansion are coming in September I think. I've been an avid player of the game since the free weekend last Thanksgiving.

Mchart said:
It makes absolutely no sense to me why anyone who actually *needs* ECC would even buy a standard consumer line-up RYZEN setup.
Click to expand...

I think this is exactly the same attitude motherboard manufacturers have on the subject. It's also why I think people bitching about ECC not working on gaming motherboards aren't being very reasonable.
 
Dan_D said:
I think this is exactly the same attitude motherboard manufacturers have on the subject. It's also why I think people bitching about ECC not working on gaming motherboards aren't being very reasonable.
Click to expand...

I think it depends. If you pay extra for a model with official ECC support (Asus Pro WS x570-Ace, I'm looking at you) then it had better work properly.

If you are just popping ECC modules in a gaming board without official ECC support and hoping it will work, then yes, you are being unreasonable if you complain.
 
Rockenrooster said:
If "real world" is browsing facebook and playing fortnite, then you won't notice much difference between a i3 and a i9 lol.

The 3700x is a great upgrade to the 2700x: 2x FP 15% IPC + higher clocks, lower power.
Dude WTH?
Click to expand...

Might want to check some benchmarks. In practically everything the difference is negligible. Also, the higher clocks aren't much higher at all. 2700X is 3.7 boosting to 4.3GHZ, the 3700x is 3.6 boosting to 4.4GHZ. Not exactly a huge jump there.

If you're coming from an i3 I'm sure it would be a nice upgrade, but going from a 2700X to a 3700X isn't worthwhile unless you use the one or two applications that may benefit from it.
 
Zarathustra[H] said:
I think it depends. If you pay extra for a model with official ECC support (Asus Pro WS x570-Ace, I'm looking at you) then it had better work properly.

If you are just popping ECC modules in a gaming board without official ECC support and hoping it will work, then yes, you are being unreasonable if you complain.
Click to expand...

Well, I wasn't talking about motherboards like the ASUS Pro WS or ASRock X99-WS, etc. There are motherboards that are more workstation focused, but still designed for enthusiasts. They are in kind of a crossover realm but ECC is part of the feature set from the start. But buying an ASUS Maximus XI APEX and bitching that it doesn't have ECC makes no sense to me.
 
Zarathustra[H] said:
I think they have some good and some bad boards.

The only ASRock board I have had any personal experience with has been my Asrock H270M-ITX/ac.

I only bought it because it was the only mini ITX board I could find for my i3-7100 which had dual on board Intel NIC's. It's been a pretty impressive little board for my pfSense router.

I've also heard good things about Asrock Rack products.

I have also heard plenty of bad experiences.

My conclusion is that their quality is inconsistent. But then again, that can be said for even the best brands. I briefly used an Asus MicroATX Haswell board for a HTPC project. Hated it so much I returned it. In general Asus boards are great though.
Click to expand...

My home server has been running on an asrock b85m-pro4 for about 6-7 years, thing doesn’t miss a beat
 
Keljian said:
Dan_D

Vive issues- they actually refunded a whole lot of people who bought vives because of this issue: https://community.viveport.com/t5/T...ncompatible-with-RYZEN-cpus/td-p/23285/page/7

Linux issues:

Sata issues:
(There are more, just not digging now)

ECC: http://www.hardwarecanucks.com/foru...ws/75030-ecc-memory-amds-ryzen-deep-dive.html
Click to expand...


I had that same issue with that same model SSD. It was annoying, but I would not consider it to be a "general AMD SATA incompatibility error that AMD needs to correct" as it seems to only be an issue with that specific model. What it NEEDS is a firmware update (the drive, that is) that addresses this issue. It would only be an "AMD" issue if a range of units had this problem. That drive is now the OS drive in my Dell PowerEdge R515 connected in place of the SATA DVD-ROM drive it came with, running Windows Server 2012 R2 as a Plex and storage server.

Edit: Just read your linked article regarding ECC. The CPU ECC was working perfectly fine. The single motherboard (Asrock Taichi x370) that they tested did not issue a halt on error. And this was tested only shortly after the Ryzen 1000 series launch; it is known that all of the launch motherboard BIOS's were all pretty buggy and it took several revisions to settle down. Since there are AM4 motherboards that explicitly support ECC and the function definitely works correctly on them, the issue was the mainboard and then possibly just the BIOS. If it were the chips, the ECC function COULD NOT function at all. Someone should revisit this on those boards so we can see if it was a hardware issue with early boards or just their firmware.

Again, you can't call either of these "issues" endemic to either the CPU or the platform when the sample size is 1 out of a bunch. I mean, I'm sure we could cherry pick little one-off things that don't work quite right (or at all) on specific Intel equipment as well.
 
Last edited:
NickM said:
How's that Broadwell E overclocking working out for ya? :LOL:
Running 1809 or newer? Checked what your actual clock is versus what you set in BIOS?
BTW, that bug can possibly kill a Broadwell E IMO (setting overclock including voltage bump, but current microcode FROM INTEL disabling anything higher than stock-I don't think they like cranked voltage at stock clocks)
Click to expand...

Let us not forget the greatness that is Avoton SoC's. My server board's chip could just oh, randomly die.
 
mvmiller12 said:
I had that same issue with that same model SSD. It was annoying, but I would not consider it to be a "general AMD SATA incompatibility error that AMD needs to correct" as it seems to only be an issue with that specific model. What it NEEDS is a firmware update (the drive, that is) that addresses this issue. It would only be an "AMD" issue if a range of units had this problem. That drive is now the OS drive in my Dell PowerEdge R515 connected in place of the SATA DVD-ROM drive it came with, running Windows Server 2012 R2 as a Plex and storage server.

Edit: Just read your linked article regarding ECC. The CPU ECC was working perfectly fine. The single motherboard (Asrock Taichi x370) that they tested did not issue a halt on error. And this was tested only shortly after the Ryzen 1000 series launch; it is known that all of the launch motherboard BIOS's were all pretty buggy and it took several revisions to settle down. Since there are AM4 motherboards that explicitly support ECC and the function definitely works correctly on them, the issue was the mainboard and then possibly just the BIOS. If it were the chips, the ECC function COULD NOT function at all. Someone should revisit this on those boards so we can see if it was a hardware issue with early boards or just their firmware.
Click to expand...

That's exactly my point on the ECC support. It's up to the board vendors and has nothing to do with AMD.
 
I've had good luck with every AsRock board I've ever purchased....built 10 gaming rigs with no issues so far.

Oth, I've had various issues with ASUS, Gigabyte, and ab older MSI (Although my z77Mpower was rock solid running a 5.2Ghz 3770k.


So I think we all have had issues with SOMETHING.
 
ccityinstaller said:
I've had good luck with every AsRock board I've ever purchased....built 10 gaming rigs with no issues so far.

Oth, I've had various issues with ASUS, Gigabyte, and ab older MSI (Although my z77Mpower was rock solid running a 5.2Ghz 3770k.


So I think we all have had issues with SOMETHING.
Click to expand...

I don't think any manufacturer has ever made a product stack without something in there having issues. I'd like to find the person who designs the layouts on some MSI boards and ask them what the hell they were smoking.
 
ccityinstaller said:
I've had good luck with every AsRock board I've ever purchased....built 10 gaming rigs with no issues so far.

Oth, I've had various issues with ASUS, Gigabyte, and ab older MSI (Although my z77Mpower was rock solid running a 5.2Ghz 3770k.


So I think we all have had issues with SOMETHING.
Click to expand...

I've had good and bad experiences with every brand I've ever dealt with. Including but not limited to: Intel, MSI, ASUS, ASRock, ABIT, Foxconn, DFI, Soltek, Soyo, EpoX, Chaintech, M-Tech, QDI, FIC, GIGABYTE, ECS, Biostar, IWill, Tyan, Supermicro, BFG, EVGA, and probably a few others I've forgotten.
 
Rockenrooster said:
my X370 ASrock only has 16MB EEPROM and it supports ryzen 3000 lol!
I will say ASrock has been killing it with Ryzen.
They will get my money.
Click to expand...

supports it w/o cutting corners? If so, I wonder how or why MSi had to do a lot of elaborate things to get it to fit into theirs? https://www.techpowerup.com/257201/...ail-amds-zen2-backwards-compatibility-promise
From a comment after the article, someone mentioned MSI is going to add back raid support eventually, so I guess thats not too bad as I originally thought
 
Keljian said:
Dan_D

Vive issues- they actually refunded a whole lot of people who bought vives because of this issue: https://community.viveport.com/t5/T...ncompatible-with-RYZEN-cpus/td-p/23285/page/7

Linux issues:

Sata issues:
(There are more, just not digging now)

ECC: http://www.hardwarecanucks.com/foru...ws/75030-ecc-memory-amds-ryzen-deep-dive.html
Click to expand...

Just read through the HTC Vive issues that you linked: A few things...

First, this is specifically with the add-on WiG Wireless Adapter. The Vive stuff works perfectly fine when wired up.

Second, HTC said the problem happened with specific motherboards (and thus OEMs and not AMD specifically) and then proceeded to give no criteria on which specific hardware they were having issues with. This was posed in November of 2018 after there was something of a blow up about it and a LOT of silence about it from HTC (both before and after they acknowledged the issue).

HTC Tech Support Rep said:
I would generally not recommend purchasing a wireless adapter if you have a Ryzen based motherboard. This is a pretty motherboard specific type of error so your success will depend mostly on which specific MB you have - some work just fine, others see the watchdog error.
Click to expand...

Thirdly, the issues were eventually resolved in April of 2019 with driver updates and a firmware update for the WiG card. It took them 6 months or so, but this configuration now works properly.

So in summary: we have an issue that happened because of the way some specific motherboard makers implemented their PCIe on some of their motherboards coupled with a bugged firmware on the WiG card. The issue triggered only under these specific circumstances, and was resolved with a firmware update and a driver update for the card. HTC customer service was pretty shit throughout the process, and their excuse for that silence was that they were working with other vendors. This is an issue that came about because HTC did not properly test their shit before pushing it out there - Ryzen systems existed for a good while BEFORE the Vive wireless adapter kit did. This COULD have been avoided with some proper pre-release QA testing. This is just a case of HTC (and the vendor they licensed the WiG card from) doing a shit job by cutting corners in their testing config. It is very possible considering the scope of effected users that they ONLY tested on Intel stuff. Even if they weren't going to fix it, the issue should have been caught and a warning on the product page issued before they started selling it. Even if they missed it for whatever reason in testing (though due again to the scope of effected users, I doubt they actually tested properly), they should have been a bunch more communicative about it when the issue was discovered. We all call Microsoft to the floor on this skimpy testing bullshit, we should hold everyone else to the same standards. All-in-all, this was a hands-down HTC fail.

Note that this is a different scenario than the Destiny 2 thing. Destiny 2 was out BEFORE Ryzen 3000 series. Bungie didn't really have a way to avoid this. Since it works just fine on Ryzen 2000 and 1000 series chips, and the 3000 series theoretically makes few architectural changes to instruction implementation from the earlier series, there is a good chance that this is an erratum. Hell, AMD has essentially admitted that it is an erratum just by trying to rush out an AGESA fix for it. At the end of the day though, Destiny 2 is still software and a workaround fix will still very likely come from Bungie. Due to the scope of effected users, you KNOW they will be on top of this. I would fully expect an erratum to be addressed in a future AGESA release, but Bungie can't count on people knowing anything about updating their mainboard BIOS's to get it. They want their Ryzen players "Playin' and Payin'" just as much as their Intel ones.

The Linux issues, I will have to concede to you. That shit makes no sense. :)
 
You must log in or register to reply here.
Back
Top