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

[Zarathustra[H]]

I know you covered it before and I'm not exactly ignoring your posts or disagreeing with you. But seeing as AMD's history with motherboards (note their sub-par boards aren't exactly a 'new' thing either) you'd think that there would be a lot more attention from them trying to smooth things over with the OEM's. Especially now seeing as they have a product people REALLY want.

And I know, this isn't exactly a big incident either, I'll get over myself sometime later after I finish my cheese.

It's difficult to place demands on motherboard OEM's when you have a second rate product.

Now that they are in play with Intel I expect the x570 boards to get every bit as much attention as Intel's equivalent boards. AMD have proven themselves as not just an "also ran" now. With it will certainly come a little bit more respect from motherboard vendors

 

[Dan_D]

It's difficult to place demands on motherboard OEM'a when you have a second rate product.

Now that they are in play with Intel I expect the x570 boards to get every bit as much attention as Intel's equivalent boards. AMD have proven themselves as not just an "also ran" now. With it will certainly come a little bit more respect from motherboard vendors

Exactly. Also, keep in mind that AMD doesn't have near Intel's platform experience. Things will potentially always be a little rougher, but it will get better over time. AMD has shown it can learn. X399 and X470, B350, B450 etc. were all better launches than X370 was.

 

[Uvaman2]

It's difficult to place demands on motherboard OEM'a when you have a second rate product.

Now that they are in play with Intel I expect the x570 boards to get every bit as much attention as Intel's equivalent boards. AMD have proven themselves as not just an "also ran" now. With it will certainly come a little bit more respect from motherboard vendors

Thats true... But also dumb on motherboards manufacturers part... Death of the only other x86 supplier could be horrendous for their business. But yeah, short sighted is the rule for sure.

 

[OFaceSIG]

Exactly. Also, keep in mind that AMD doesn't have near Intel's platform experience. Things will potentially always be a little rougher, but it will get better over time. AMD has shown it can learn. X399 and X470, B350, B450 etc. were all better launches than X370 was.

True

 

[Zarathustra[H]]

Thats true... But also dumb on motherboards manufacturers part... Death of the only other x86 supplier could be horrendous for their business. But yeah, short sighted is the rule for sure.

Well, financially, three things go into a calculation when starting a new project to design anything. How much is it going to cost to develop, and how many can we sell, at how much profit.

When designing boards for budget processors, from a second tier manufacturer, the volume and price isn't going to be there. No one is going to pay for a premium motherboard to use with a low cost alternative CPU.

The motherboard manufacturers need evidence that they are going to have a market to sell into before they commit to bringing out high end products which cost money to develop.

Dan brings up a good point about Intel's launch interfering with the 300 series chipset motherboards from AMD which I was unaware of, but I also think a huge part of it was that they didn't know what to expect from Ryzen. They were burned before just like we all were. I bought a very good 990FX motherboard in 2011, a few months before the Bulldozer launch which I used with a Phenom II 1090T, and was planning to drop in a Bulldozer CPU when it launched, only to be greatly disappointed by that launch. I think many motherboard manufacturers were as well, as the motherboard offerings were pretty decent for that launch, it was the CPU that was historically bad.

So, I'm guessing the motherboard makers didn't want to risk a repeat of 2011 and proceeded very cautiously with the initial Ryzen launch. Now they seem to be fully on board. I mean, there are high end $700 x570 Socket AM4 boards... That's just nuts.

 

[Dan_D]

Well, financially, three things go into a calculation when starting a new project to design anything. How much is it going to cost to develop, and how many can we sell, at how much profit.

When designing boards for budget processors, from a second tier manufacturer, the volume and price isn't going to be there. No one is going to pay for a premium motherboard to use with a low cost alternative CPU.

The motherboard manufacturers need evidence that they are going to have a market to sell into before they commit to bringing out high end products which cost money to develop.

Dan brings up a good point about Intel's launch interfering with the 300 series chipset motherboards from AMD which I was unaware of, but I also think a huge part of it was that they didn't know what to expect from Ryzen. They were burned before just like we all were. I bought a very good 990FX motherboard in 2011, a few months before the Bulldozer launch which I used with a Phenom II 1090T, and was planning to drop in a Bulldozer CPU when it launched, only to be greatly disappointed by that launch. I think many motherboard manufacturers were as well, as the motherboard offerings were pretty decent for that launch, it was the CPU that was historically bad.

So, I'm guessing the motherboard makers didn't want to risk a repeat of 2011 and proceeded very cautiously with the initial Ryzen launch. Now they seem to be fully on board. I mean, there are high end $700 x570 Socket AM4 boards... That's just nuts.

You are exactly correct. There was a big investment from motherboard makers to get ready for Bulldozer. A CPU that was delayed prompting the release of 990FX boards before the CPU's were available. This turned out to be a huge mistake as the boards didn't review well using older Phenom II CPU's. Those boards were optimized for Bulldozer and sure enough, later on they were better with the Bulldozer CPU's, but the damage was done. Not knowing what to expect from Ryzen was why motherboard manufacturers concentrated on Intel and practically phoned it in on early X370 boards.

 

[Ready4Dis]

You guys act like Intel never had any issues. I was an early adopter of skylake, b150m ASRock board. Kept getting random memory issues and computer crashes for a few weeks. Ran all sorts of tests and everything kept passing. Tried other ram, new psu, different GPU... Finally updated bios came out, lol and behold system has been Rock solid since. Growing pains of early adoption. AMD has a lot less resources so my assumption is they take a little longer to resolve things, but hey .. my b150m has almost no upgrade path (7000 series isn't really much/any different) even though the only difference to the 8th Feb was what, one ground pin that was proven unecessary? I am one that upgrades CPUs, during athlon, x2, x3 BE, x4, core 2 duo, quad, upgraded the dual xeons server... It's good to get some benefit without having to upgrade an entire system.

 

[sleepybp]

With all the issues pointed out with Ryzen bios in this specific thread, its a miracle all the Ryzen 3000 reviews were able to publish on time, and without any notes of instability or crashing during testing.

Sucks that I had to wait 7 days for a bios update to run Ryzen 3000 on my old recycled X470 motherboard.
I would rather go Intel and buy a new motherboard every CPU release.


/S

 

[Rockenrooster]

Dan_D



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

My brother had a MX100 SSD with his i5 4690k.
That thing was never stable or consistent, his system would just hang saying 100% disk activity with nothing going on, it was showing errors and stuff so he RMA'd it. Got a MX200 back and it had the SAME issue. It would freeze within 5min.
Installing the intel rapid storage driver fixed the issue for a while but it came up again, so RMA'd that one too, so they sent him a MX500 and it exhibited the SAME issue. He said eff it and got a Samsung 860 EVO and its been stable ever since.

I've seen this EXACT SAME issue with a FX 8300 system (inlaws), MX500 didn't work right (freezing within 5 mins) got a Samsung SSD and its been working fine ever since (they have since upgraded to a Ryzen 2600 Cyberpower prebuilt and kept the same SSD with no issues)

That said I'm staying away from Crucial, as in my experience, it's NOT an AMD thing. It suck because the m4 SSD back in the early days was the bomb!

 

[Dan_D]

You guys act like Intel never had any issues. I was an early adopter of skylake, b150m ASRock board. Kept getting random memory issues and computer crashes for a few weeks. Ran all sorts of tests and everything kept passing. Tried other ram, new psu, different GPU... Finally updated bios came out, lol and behold system has been Rock solid since. Growing pains of early adoption. AMD has a lot less resources so my assumption is they take a little longer to resolve things, but hey .. my b150m has almost no upgrade path (7000 series isn't really much/any different) even though the only difference to the 8th Feb was what, one ground pin that was proven unecessary? I am one that upgrades CPUs, during athlon, x2, x3 BE, x4, core 2 duo, quad, upgraded the dual xeons server... It's good to get some benefit without having to upgrade an entire system.

Not at all. I remember the Intel 820 issues with DDR RAM, the Z68 issues, the Pentium 60/66 FDIV issue, the 1.13GHz Pentium III and the 3.8GHz Pentium III's that both got pulled from the market because they ran too hot and would lock up at those clocks. Generally, Intel doesn't have issues as consistently or as often as AMD does. That's not to say they are perfect, far from it, but platform wise Intel's are usually less trouble.

 

[Ready4Dis]

I agree, and them having more resources and more pull with their board vendors helps a lot. I wasn't saying you, but some others act like if they buy Intel they are guaranteed to not have any issues ever, while that's far from the truth. AMD do and did have issues and it seems more often than Intel, however their issues aren't normally hardware related (fdiv bug, lots of security issues from taking shortcuts, recalling CPUs that run to hot). Not saying never, and and do a tend to have more consistently had problems, but it's a crap shoot with new technology no matter which brand you get. If you're worried, don't buy it u til the bugs are worked out. That goes for amd, intel or Nvidia or any company.

 

[Dan_D]

I waited a bit to by an RTX 2080 Ti due to the space invaders issue. Sometimes it pays off to be patient.

 

[Zarathustra[H]]

I waited a bit to by an RTX 2080 Ti due to the space invaders issue. Sometimes it pays off to be patient.

I did the same, and as a result I haven't bought one.

I find that buying these expensive top end boards at launch is the only way I can justify the price, because then it gets spread over a longer period of use. After I got comfort that the Space Invaders issue was addressed I felt it was too late for me.

Next generation.

 

[Red Falcon]

And this is, sadly, why Intel remains the only viable upgrade option for me. Time is the most valuable commodity for me and I don't want to waste it fixing AMD's screw-ups.

Because Intel totally didn't screw the pooch on Meltdown, Spectre, Foreshadow, disabling HyperThreading, and the other 10+ hardware exploits...
How many months did we have to wait for vendor patches for Intel's multiple and repeated fuckups?

 

[mda]

For AGESA being such a critical component of the AMD firmware, it seems to be poorly documented what exactly each version fixes/changes.

 

[iDShaDoW]

When it is time to upgrade, I replace the mobo/cpu/heatsink/ram all together. Why the hell would I want to build a system, and replace the cpu every year? Sounds too expensive, it's not like cpu's are $49.99..

Lol damn. I wasn't aware that heatsinks and RAM continue to evolve and improve at that rate.

I upgrade my motherboard/CPU every couple years when there's a decent improvement (like with AMD Ryzen series).

RAM is RAM until DDR5 lands. AIO heatsinks just get some fancy new packaging but aren't giving you noticeable temp improvements. The CPUs are where you get the efficiency/TDP and, as a result, better thermals.

And it's easy enough to pull your motherboard and CPU and sell it on Craigslist or eBay and recoup a decent amount of what you spent on it originally.

 

[NKD]

And this is, sadly, why Intel remains the only viable upgrade option for me. Time is the most valuable commodity for me and I don't want to waste it fixing AMD's screw-ups.

wait you make your decisions on a bios that was buggy and they refused to release it? ROFL! My system is rocking100% stable. I am not going to bitch about them doing the right thing and taking their time working on it.

 

[mda]

New platform problems are nothing new...

AMD and Intel will have them. Haswell MB Sleep Bug? P67 SATA port issues?

If 100% problem free and stability is a requirement, one shouldn't be the first to move to new platforms, regardless of manufacturer. Full stop.

Bring out the pitchforks IF the manufacturer can't reasonably rectify those issues within a reasonable amount of time.

 

[Dan_D]

For AGESA being such a critical component of the AMD firmware, it seems to be poorly documented what exactly each version fixes/changes.

The motherboard vendors are horrendous about actually showing anything in the patch notes of value.

 

[GoodBoy]

Lol damn. I wasn't aware that heatsinks and RAM continue to evolve and improve at that rate.

I upgrade my motherboard/CPU every couple years...

Well, because my last 2 upgrades in total are over 10 years old. The i7-920 was DDR3 pretty sure and lasted 7.5 years in my main box, and it's still good. I need the ram with it, and the new system was ddr4.

As far as upgrading CPU's, there's almost no point anymore. It's not like 18 years ago when every 6 months, a new cpu that was twice as fast came out... now we get a few percent per tock, about 5% per year. Not worth it for most to spend $300 to $600 every year for minimal gains. So don't give a flip if the socket changes. And it apparently helps avoid some of this crap AMD mobo's are having, regardless of who is at fault.

 

[mda]

The motherboard vendors are horrendous about actually showing anything in the patch notes of value.

Reading "improved stability" from the asus website gets boring, but having a BIOS update with a blank note is something else.

 

[IdiotInCharge]

The motherboard vendors are horrendous about actually showing anything in the patch notes of value.

I assume that you've been reading their manuals for at least the 25 years that I have... I can't say that they've changed much. Their English is perhaps a tad better

 

[Red Falcon]

As far as upgrading CPU's, there's almost no point anymore. It's not like 18 years ago when every 6 months, a new cpu that was twice as fast came out... now we get a few percent per tock, about 5% per year. Not worth it for most to spend $300 to $600 every year for minimal gains. So don't give a flip if the socket changes. And it apparently helps avoid some of this crap AMD mobo's are having, regardless of who is at fault.

That's not true any more - if it were Q2 of 2017 or earlier, I would have fully agreed.
With AMD greatly upping the core count, there is a massive difference now.

Going from an Intel 6700K (before and after exploit patches) to an AMD 2700X was night and day, especially in light and heavy SMP applications.
After the Intel exploit patches, the 2700X beat the shit out of the 6700K in single-threaded applications as well.

So, going from 4 cores to 8-16 cores with similar IPC is a massive boost.
Also, Intel's tick-tock strategy died with Kaby Lake back in 2016, hence all of the 14nm++++++ high-priced garbage they are releasing now.

 

[Mchart]

I assume that you've been reading their manuals for at least the 25 years that I have... I can't say that they've changed much. Their English is perhaps a tad better

-SECURITY IMPROVEMENTS
-SUPPORT NEW FEATURE

-¹ª©¥»×¥¿¤U¦C°ÝÃD

 

[ccityinstaller]

Reading "improved stability" from the asus website gets boring, but having a BIOS update with a blank note is something else.

Unless the Ninja'd it, it says it fixes issues with mouse cursor lag.

At least you have had a bios in the last month. My board has not had one since mid June.

My 3700x works, but I really wanted them to release an update by now. I guess with AMD pulling the AGESA release it makes sense there is a delay since Asus didn't do their QA stuff.


I really hate Asus as a company. I only purchased bheir board since I needed 3 16x slots since I was using 3x VEGAs with you blocks.

Initially I got the x370 prime pro and had an old stock one that had the random bios corruption issue. Thankfully I purchased a replacement plan with MC.


I got the x470 Prime Pro (even though I didn't want to give Asus another dime) because it was either that, or one more expensive Gigabyte board (iirc)....

Figured I would stick to the devil I knew since the board layout was the same.

 

[ccityinstaller]

That's not true any more - if it were Q2 of 2017 or earlier, I would have fully agreed.
With AMD greatly upping the core count, there is a massive difference now.

Going from an Intel 6700K (before and after exploit patches) to an AMD 2700X was night and day, especially in light and heavy SMP applications.
After the Intel exploit patches, the 2700X beat the shit out of the 6700K in single-threaded applications as well.

So, going from 4 cores to 8-16 cores with similar IPC is a massive boost.
Also, Intel's tick-tock strategy died with Kaby Lake back in 2016, hence all of the 14nm++++++ high-priced garbage they are releasing now.

Funny, my brother felt the same way going to a 3700x from a 6700k yet I was basically called a liar in another thread about how a 6600k plays every modern game with zero issues and settings.

 

[mda]

Unless the Ninja'd it, it says it fixes issues with mouse cursor lag.

Prior to the latest one, there were a few releases that were blank which was the one I downloaded when I picked up my 3700X on launch date. Didn't realize there was a new one haha

Funny, my brother felt the same way going to a 3700x from a 6700k yet I was basically called a liar in another thread about how a 6600k plays every modern game with zero issues and settings.

Moving from a 4770K haswell at 4.1-4.2 ghz to a stock 3700x, I notice the difference playing the division 2 and starcraft 2 (large 8 player games). Not a mindblowing change but noticeable for me. Mostly the higher low 0.1%s are the most perceptible. If you don't notice those things, you also won't see much of a difference.

 

[Mchart]

Funny, my brother felt the same way going to a 3700x from a 6700k yet I was basically called a liar in another thread about how a 6600k plays every modern game with zero issues and settings.

I saw an improvement going from a 5820k to a 2700x. It wasn't huge from a max FPS perspective, but the drops were much less and you could tell.

 

[Derangel]

I saw an improvement going from a 5820k to a 2700x. It wasn't huge from a max FPS perspective, but the drops were much less and you could tell.

What was your 5820K set to and what res were you playing at? Looking at reviews the numbers I've seen haven't been quite enough to make me jump at Zen 3 yet, but if it makes a big difference on a mins I might have reason to stop fence-sitting.

 

[Mchart]

What was your 5820K set to and what res were you playing at? Looking at reviews the numbers I've seen haven't been quite enough to make me jump at Zen 3 yet, but if it makes a big difference on a mins I might have reason to stop fence-sitting.

I never really benchmarked between the two. I mostly did it because I had an early x99 board that didn't have NVME support, among other things. Certain games felt better on the 2700x, there just wasn't as much frame drops as I was seeing on the 5820k. I had my 5820k clocked at 4.3ghz. Not to mention the 5820k system ran a lot hotter.

Now i'm on the 3800x and the 5820k system just isn't even close. The 2700x was about on par than the 5820k for single threaded stuff, while the 3000 is notably better. The 5820k, even with the average low/mid 4ghz overclock is just ancient at this point. It's not bad by any means, but it's medium-end at best at this point.

But the games that can use the two extra cores it really made a difference.

 

[Dan_D]

I assume that you've been reading their manuals for at least the 25 years that I have... I can't say that they've changed much. Their English is perhaps a tad better

Yes, I have. The English is better. We don't have the "No any drive found" message for drive detection on a controller with nothing connected to it anymore.

 

[Red Falcon]

Funny, my brother felt the same way going to a 3700x from a 6700k yet I was basically called a liar in another thread about how a 6600k plays every modern game with zero issues and settings.

Depends on what you are doing.
I can still use a vanilla 6600 at stock and play most modern games without issue, but quite a few games are starting to struggle on it and a 6700K (with HT disabled due to Foreshadow) due to lack of threads.

For gaming, I think it is less of an issue than it is with other applications, such as rendering, transcoding, VMs, databases, SMP/HPC workloads, etc.
Quad-core CPUs get destroyed in those areas, and Intel CPUs especially so due to the exploit patches killing off between 5%-80% performance, depending on the workload and task.