Tuesday, September 10th 2019

AMD AGESA 1.0.0.3ABBA Detailed, Fixes Zen2 Boost Issues

AMD is giving final touches to an AGESA microcode update that fixes the issue of underwhelming Precision Boost behavior on its 3rd generation Ryzen processors. Version ComboAM4 1.0.0.3ABBA is being pushed to motherboard manufacturers to integrate with their UEFI firmware, and one such dispatch to MSI got leaked to the web on ChipHell. Tom's Hardware grabbed the BIOS as it was compatible with the MEG X570 Creator motherboard they have, and tested the Ryzen 9 3900X and Ryzen 7 3700X with it.

In its testing, posted in a mini-review article, Tom's Hardware observed that with AGESA 1.0.0.3ABBA, their 3700X sample was correctly hitting 4.40 GHz across the board at stock settings. With the older 1.0.0.3AB, it would touch 4.375 GHz. The Ryzen 9 3900X behaves slightly differently with this microcode. Tom's Hardware was able to raise its peak boost frequency from 4.575 GHz to 4.625 GHz (above the 4.60 GHz specification), but in certain tests such as POV-Ray and Cinebench, its boost frequency decays down to 4.250 GHz. Overall, the reviewer tabulated improved performance on the chips with the new microcode. The new microcode also apparently changes the processor's thermal thresholds.

Update (10/9) AMD posted an elaborate release detailing the AGESA 1.0.0.3ABBA update.

Hello, everyone! We're delighted by your support and the strong momentum of 3rd Gen AMD Ryzen processors in the marketplace, and we continue to watch your feedback closely. Today we have some important updates for you concerning processor boost behavior, desktop idle behavior, and a new monitoring SDK. The first two changes will be arriving in BIOSes based on AGESA 1003ABBA, and we are planning to make the SDK public on developer.amd.com with a target release date of September 30.

Boost Changes
Starting with our commitment to provide you an update on processor boost, our analysis indicates that the processor boost algorithm was affected by an issue that could cause target frequencies to be lower than expected. This has been resolved. We've also been exploring other opportunities to optimize performance, which can further enhance the frequency. These changes are now being implemented in flashable BIOSes from our motherboard partners. Across the stack of 3rd Gen Ryzen Processors, our internal testing shows that these changes can add approximately 25-50 MHz to the current boost frequencies under various workloads.

Our estimation of the benefit is broadly based on workloads like PCMark 10 and Kraken JavaScript Benchmark. The actual improvement may be lower or higher depending on the workload, system configuration, and thermal/cooling solution implemented in the PC. We used the following test system in our analysis:
  • AMD Reference Motherboard (AGESA 1003ABBA beta BIOS)
  • 2x8GB DDR4-3600C16
  • AMD Wraith Prism and Noctua NH-D15S coolers
  • Windows 10 May 2019 Update
  • 22°C ambient test lab
  • Streacom BC1 Open Benchtable
  • AMD Chipset Driver 1.8.19.xxx
  • AMD Ryzen Balanced power plan
  • BIOS defaults (except memory OC)
These improvements will be available in final BIOSes starting in about three weeks' time, depending on the testing and implementation schedule of your motherboard manufacturer. Additional information on boost frequency in the 3rd Gen AMD Ryzen Processors can also be obtained from this separate blog update.

Going forward, it's important to understand how our boost technology operates. Our processors perform intelligent real-time analysis of the CPU temperature, motherboard voltage regulator current (amps), socket power (watts), loaded cores, and workload intensity to maximize performance from millisecond to millisecond. Ensuring your system has adequate thermal paste; reliable system cooling; the latest motherboard BIOS; reliable BIOS settings/configuration; the latest AMD chipset driver; and the latest operating system can enhance your experience.

Following the installation of the latest BIOS update, a consumer running a bursty, single threaded application on a PC with the latest software updates and adequate voltage and thermal headroom should see the maximum boost frequency of their processor. PCMark 10 is a good proxy for a user to test the maximum boost frequency of the processor in their system. It is fully expected that if users run a workload like Cinebench, which runs for an extended period of time, the operating frequencies may be lower than maximum throughout the run.

In addition, we do want to address recent questions about reliability. We perform extensive engineering analysis to develop reliability models and to model the lifetime of our processors before entering mass production. While AGESA 1003AB contained changes to improve system stability and performance for users, changes were not made for product longevity reasons. We do not expect that the improvements that have been made in boost frequency for AGESA 1003ABBA will have any impact on the lifetime of your Ryzen processor.

Revisiting Calmer Idle
In late July, we implemented a series of software changes that would help the processor ignore requests for voltage/frequency boost from lightweight applications. The goal was to make the processor more relaxed at the desktop, but poised to react for serious workloads. While many of you were happy with the effect of the software changes, some of you were still grappling with cases where the CPU was a bit overzealous with boost. We wanted to smooth those out, too.

Today we're announcing that AGESA 1003ABBA carries firmware-level changes designed to do just that. The changes primarily arrive in the form of an "activity filter" that empowers the CPU boost algorithm itself to disregard intermittent OS and application background noise. Example test cases might include: video playback, game launchers, monitoring utilities, and peripheral utilities. These cases tend to make regular requests for a higher boost state, but their intermittent nature would fall below the threshold of the activity filter.

Net-net, we expect you'll see lower desktop voltages, around 1.2 V, for the core(s) actively handling such tasks. We believe this solution will be even more effective than the July changes for an even wider range of applications.

Please keep in mind, however, that this firmware change is not a cap. The processor must still be free to boost if active workload(s) seriously require it, so you should still expect occasions where the processor will explore its designed and tested voltage range of 0.2 V to 1.5 V.

New Monitoring SDK
Obtaining reliable data about the operating behavior of a processor is important to enthusiasts such as myself. There are many monitoring utilities on the market, and we work with many of them to ensure they're accessing telemetry data in a sensible manner. Regardless of the utility, however, it's common sense that all the tools should roughly correlate when you ask a simple question like "what's my CPU temperature?"

Enabling a consistent experience across monitoring utilities is important to us. That's why we're announcing the September 30 release of the AMD Monitoring SDK that will allow anyone to build a public monitoring utility that can reliably report a range of key processor metrics in a consistent manner. Altogether, there are 30+ API calls within the first SDK release, but we've highlighted a few of the more important or interesting ones below:
  • Current Operating Temperature: Reports the average temperature of the CPU cores over a short sample period. By design, this metric filters transient spikes that can skew temperature reporting.
  • Peak Core(s) Voltage (PCV): Reports the Voltage Identification (VID) requested by the CPU package of the motherboard voltage regulators. This voltage is set to service the needs of the cores under active load, but isn't necessarily the final voltage experienced by all of the CPU cores.
  • Average Core Voltage (ACV): Reports the average voltages experienced by all processor cores over a short sample period, factoring in active power management, sleep states, Vdroop, and idle time.
  • EDC (A), TDC (A), PPT (W): The current and power limits for your motherboard VRMs and processor socket.
  • Peak Speed: The maximum frequency of the fastest core during the sample period.
  • Effective Frequency: The frequency of the processor cores after factoring in time spent in sleep states (e.g. cc6 core sleep or pc6 package sleep). Example: One processor core is running at 4 GHz while awake, but in cc6 core sleep for 50% of the sample period. The effective frequency of this core would be 2 GHz. This value can give you a feel for how often the cores are using aggressive power management capabilities that aren't immediately obvious (e.g. clock or voltage changes).
  • Various voltages and clocks, including: SoC voltage, DRAM voltage, fabric clock, memory clock, etc.
A Preview in Action
This SDK will be available for public download on developer.amd.com on September 30. As a preview of what the new SDK can enable, AMD Ryzen Master (version 2.0.2.1271) has already been updated with the new Average Core Voltage API for 3rd Gen Ryzen Processors. It's ready for download today!

As noted above, Average Core Voltage shows you average voltages that all CPU cores are experiencing over a short sample period after you factor in sleep states, idle states, active power management, and Vdroop. Depending on the load on the processor, this value might be quite different from Peak Core(s) Voltage.

For example: if the processor is lightly loaded on a few cores, the overall activity level of all the CPU cores will be relatively low and, therefore, the Average Core Voltage will be low as well. But the active cores still need intermittently higher voltages to power boost frequencies, which will be reflected in the Peak Core Voltage. As the CPU comes under full load, these two values will eventually converge representing that all cores are active at approximately the same intensity. The overall goal of these two values is to show you what's happening moment-to-moment the most loaded cores (Peak), and what's happening more generally to the CPU cores over time (Average).

We hope new APIs like Average Core Voltage give you a better understanding of how our processors behave, and we can't wait to see more tools make use of the new monitoring SDK. Visit amd.com on September 30 for the first public release!

What to Expect Next
AGESA 1003ABBA has now been released to our motherboard partners. Now they will perform additional testing, QA, and implementation work on their specific hardware (versus our reference motherboard). Final BIOSes based on AGESA 1003ABBA will begin to arrive in approximately three weeks, depending on the testing time of your vendor and motherboard.

Going forward, we'll continue providing updates in this format as the updates are being prepped for release. Sources: Tom's Hardware, ChipHell
Add your own comment

106 Comments on AMD AGESA 1.0.0.3ABBA Detailed, Fixes Zen2 Boost Issues

#51
sutyi
lexluthermiester, post: 4114250, member: 134537"
To be fair, 200mhz out of 4400+mhz isn't enough to make a big deal about. That's less than 5%.
Doesn't matter... it should do what it says on the box, simple as that. If it says 4.6GHz boost, it should boost to 4.6GHz on at least 2 cores even if it only holds it for a single second.

My 0.25$: This whole boost thing got a bit out of hand... Not to say something was not 100% OK with the boost algorithm, but most people experiencing wierd boost behavior is like 30% AMD, 20% user settings and 50% of MB vendors doing wierd stuff with their own tweaks to EDC, PPT and TDC behavior in their BIOS code ending up with clocks all over the place... Just like ASUS turning on MCE along with XMP, but that might have been already fixed.

W1zzard, post: 4114282, member: 1"
Just waiting for asrock taichi version to come out
Great Tech Merlin may I have a question regarding that mobo? How are PCH temps with a graphics card in place in considering the fan placement on the board just under the first PCIe x16?
Posted on Reply
#52
Chrispy_
TheLostSwede, post: 4114088, member: 3382"
And you base this on running a Xeon rig? Seriously?

So AMD fixes the issue, yet all the Intel Fanboi's are here slinging mud... :banghead:
I'm not slinging mud. I'm moaning that the fix is pointless because there was no issue in the first place. The net result of this AGESA patch is that the cores are pushed harder than before, and as a result they are generating heat faster than before. All other things being equal - that means they'll hit their power/temp/cooling equilibrium sooner so the steady-state, minimum boost will occur earlier.

And yes I have Xeons. I have a whole goddamn server room, 60+ node CPU compute farm, 200+ node GPGPU compute farm and access to around 2000 different machines spanning multiple disciplines, generations, vendors, purposes and price brackets. It's my job to select the best hardware and software for specific purposes, test it, and then buy that hardware in bulk for multiple companies. I have contacts in AMD, Intel, and Nvidia and haven't missed a SIGGRAPH or Computex in almost a decade.

I don't know how I can state this any more clearly; I am vendor agnostic. I want the best solution regardless of who makes it. I do not like marketing spin, because I don't get paid for marketing spin. I get paid for actual real-world results and that means I need performance/$ and no-nonsense approaches to everything. Yes, I am biased. I am biased AGAINST BS and AGAINST shilling, because that makes my job harder. I enjoy tech as a hobby and gaming enthusiast, but first and foremost I need to understand all of the tech because my livelihood is financed by it.
Posted on Reply
#53
Nordic
Xzibit, post: 4114226, member: 105152"
He did try to pass it as such. Roman in his video said he filtered out the "noise". How is that even possible when their wasn't a validation process for the survey. Anyone could fill in the survey with no proof at all.
He did not try to pass it as such. Filtering out the noise in this case would be removing bad data so that the information becomes clearer. Filtering out the noise is a way to try and improve data quality.

He was very clear that he had data quality issues. He was very clear that his methodology was flawed. He did as scientific an analysis as he could given the data he collected.

Even though I think it was clear, it seems it wasn't enough for everyone to understand that.

Chrispy_, post: 4114298, member: 185623"
... the fix is pointless because there was no issue in the first place.
AMD said there was a bug. Are you saying AMD was wrong?
Posted on Reply
#54
lexluthermiester
xorbe, post: 4114269, member: 102945"
Am I the only one that installs the cpu, slaps in 3200 CL14, and just uses the PC as is?
Nope, most people do that.
Posted on Reply
#55
Chrispy_
Nordic, post: 4114309, member: 96457"
AMD said there was a bug. Are you saying AMD was wrong?
AMD said there was a bug in order to shut people up and the fix appears to be to just jam more voltage through the chips so that more of them hit the "up to" boost clocks than currently.

I suspect the AGESA version has two fixes rolled into one:
  1. Workarounds for poor vendor implementations in popular motherboard BIOSes because those vendors haven't got it right yet. This applies to those chips that were hundreds of MHz short of the boost clock and locked at specific frequencies. The onus is on the motherboard BIOS vendor to fix their broken BIOS but AMD can patch around a bad BIOS just as AMD/Nvidia GPU drivers can patch around bad game engine coding. Yes, this is a bugfix. It shouldn't be necessary but AMD are fixing bugs that Asus/Gigabyte/MSI/Asrock/EVGA haven't because this issue has enough press coverage that it has become more their problem than the motherboard vendors' problem.
  2. Increased voltage. Der8auer's survey showed that the number of people achieving the max boost clocks wasn't great - with the vast majority of chips falling 50MHz or 25MHz short. This extra voltage allows all of those 'almost there' chips to hit that exact, round-number 4.4/4.5/4.6GHz. Even if they were already doing that to the nearest significant figure. My only regret with this approach is that in fixing the peak clockspeed they seem to have hurt the overall performance. This image shows that despite the peak being momentarily higher at the very start of the test, the steady-state single-thread clockspeed is 25MHz lower for the vast bulk of the graph because the extra voltage hurts performance when the chips are temperature limited. This image shows that the voltage boost causes the chips to hit that temperature limit faster. So yes, the people whining that they were 50MHz short are now happy but the end result is that everyone's chips are now over-volted that little bit more, and hotter and slower overall because of it.
Posted on Reply
#56
Vya Domus
Nordic, post: 4114309, member: 96457"
He did not try to pass it as such. Filtering out the noise in this case would be removing bad data so that the information becomes clearer. Filtering out the noise is a way to try and improve data quality.

He was very clear that he had data quality issues. He was very clear that his methodology was flawed. He did as scientific an analysis as he could given the data he collected.
You can filter out as much as you want the data and make sure you use the right methodology, it all amounts to nothing. The problem is not with the analysis, it's with the data. Everyone must eventually understand that asking people over the internet what a sensor reads it's the worst way you can go about to do this sort of thing.
Posted on Reply
#57
Dave65
TheLostSwede, post: 4114088, member: 3382"
For the billionth time, NO. It's within AMD spec. Can people PLEASE stop commenting about the 1.5V already?

And it seems the cores are hitting even higher boosts given a bit of time.




And you base this on running a Xeon rig? Seriously?

So AMD fixes the issue, yet all the Intel Fanboi's are here slinging mud... :banghead:
That is what Intel fan babies do.
Posted on Reply
#58
lexluthermiester
Chrispy_, post: 4114298, member: 185623"
I don't know how I can state this any more clearly; I am vendor agnostic. I want the best solution regardless of who makes it. I do not like marketing spin, because I don't get paid for marketing spin. I get paid for actual real-world results and that means I need performance/$ and no-nonsense approaches to everything. Yes, I am biased. I am biased AGAINST BS and AGAINST shilling, because that makes my job harder. I enjoy tech as a hobby and gaming enthusiast, but first and foremost I need to understand all of the tech because my livelihood is financed by it.
Well said. Right there with you, though in a different part of the industry.

Dave65, post: 4114329, member: 82235"
That is what Intel fan babies do.
No it's a real problem. 1.5v will do bad things to 7nm circuit pathways at room temperature. Electron migration and migation become very real problems at that voltage.
Posted on Reply
#59
Xzibit
Nordic, post: 4114309, member: 96457"
He did not try to pass it as such. Filtering out the noise in this case would be removing bad data so that the information becomes clearer. Filtering out the noise is a way to try and improve data quality.

He was very clear that he had data quality issues. He was very clear that his methodology was flawed. He did as scientific an analysis as he could given the data he collected.

Even though I think it was clear, it seems it wasn't enough for everyone to understand that.
Of course his data was flawed. Anyone could fill out the survey with what ever he/she intended and fill it out multiple times. Roman never addressed that issue. All one had to do is click a few options (4), name a MB and that was it. Nothing was checked ever!!! no proof that the submitees had the hardware or complied with his instructions at all.

He used it as a soap-box. Good or Bad.

He has CaseKing as a resource. He could of done actual controlled testing in the 10 days he had the survey open.
Posted on Reply
#60
Redwoodz
lexluthermiester, post: 4114335, member: 134537"
Well said. Right there with you, though in a different part of the industry.


No it's a real problem. 1.5v will do bad things to 7nm circuit pathways at room temperature. Electron migration and migation become very real problems at that voltage.
PROOF. That's the most ubsurd thing I have ever heard. Room temp is like 26c. Secondly- the amount of voltage is not nearly as important as the power draw. Single thread w\ a millisecond of a .2 increase in volts is going to do ZERO to electron migration. You people are NOT even seeing the real time figures yet. Until AMD releases the new monitoring SDK.

Stop the madness.






xkm1948, post: 4113987, member: 50521"
Wonder what all those “This is fine” AMD fanboiz gonna say now.
Told you so. ;)
Just because we knew it was no big deal we also knew it would be addressed by AMD. Fine wine baby....fine wine. ;)
Posted on Reply
#61
Nordic
Chrispy_, post: 4114324, member: 185623"
This extra voltage allows all of those 'almost there' chips to hit that exact, round-number 4.4/4.5/4.6GHz.
I already had you on my ignore list, so it is totally my fault that I am in this stupid argument with someone who doesn't understand what is going on. I know you don't understand because of this one little statement. It is a hard wall that these chips hit that is preventing them from getting to those advertising boost clocks. More voltage doesn't fix that the chip hits a wall at some arbitrary point.
Posted on Reply
#62
TheLostSwede
hzy4, post: 4114228, member: 186895"
I though the Agesa was released just for the board vendors, and the implementation into BIOS updates could take up to the 30th September. Is the Agesa available through AMD? Can you clarify where you got it from?
See my first post in this thread.

xorbe, post: 4114269, member: 102945"
Am I the only one that installs the cpu, slaps in 3200 CL14, and just uses the PC as is?
You don't even set the correct memory speed in the UEFI? :roll:
Posted on Reply
#63
voltage
ZoneDymo, post: 4113949, member: 66089"
I doubt anyone would care if someone mentions their cpu died after 6+ years of use tbh.
one sided. If this was an Intel issue, you wouldn't be saying that. you would care if your cpu died in 6 years. I personally am not falling for amd's hype right now, but i would like a new system, as I am using a workstation that just turned 8 years old, but works like new. An old Intel i7, with a second gen plextor SSD, back when they were one of the best. Next year when amd had newer, and intel finally comes out with Tiger Lake, ill evaluate both, and buy a new laptop with either, and a new work station, but personally, I do care if my cpu lasts, and I know plenty of people who care also.
Posted on Reply
#64
TheinsanegamerN
Arc1t3ct, post: 4114122, member: 190440"
My AMD K6-2 @450Mhz still runs fine. What’s your point exactly?

Are you implying that AMD cpus are not as well made?
No, the exact opposite, that AMD knows what they are doing and they are not going to push an update that reduces their CPU lifespan, and that ZoneDymo's insinuation that nobody cares if a 6 year old CPU cooks itself due to vendor BIOS is hilariously wrong.

Glad to see the entire community took my comment the wrong direction. Made perfect sense to me.....
Posted on Reply
#66
lexluthermiester
Redwoodz, post: 4114362, member: 148684"
PROOF
Physics. Look into it.
Redwoodz, post: 4114362, member: 148684"
Secondly- the amount of voltage is not nearly as important as the power draw.
Glad you mentioned that because even at the relatively low power draw of 65w, at 1.5v that's alot of amps to run through such small pathways. More wattage, more power drawn.
Redwoodz, post: 4114362, member: 148684"
Single thread w\ a millisecond of a .2 increase in volts is going to do ZERO to electron migration.
Your understanding of electronics clearly needs improvement.
Posted on Reply
#67
Nkd
sutyi, post: 4114292, member: 112688"
Doesn't matter... it should do what it says on the box, simple as that. If it says 4.6GHz boost, it should boost to 4.6GHz on at least 2 cores even if it only holds it for a single second.

My 0.25$: This whole boost thing got a bit out of hand... Not to say something was not 100% OK with the boost algorithm, but most people experiencing wierd boost behavior is like 30% AMD, 20% user settings and 50% of MB vendors doing wierd stuff with their own tweaks to EDC, PPT and TDC behavior in their BIOS code ending up with clocks all over the place... Just like ASUS turning on MCE along with XMP, but that might have been already fixed.



Great Tech Merlin may I have a question regarding that mobo? How are PCH temps with a graphics card in place in considering the fan placement on the board just under the first PCIe x16?
lol. So even though amd boost says max boost and max boost has always been single core. You want it to boost to two cores because that what you think it should do even though box doesn't say it? hmm lol.
Posted on Reply
#69
TheLostSwede
mahoney, post: 4114400, member: 187374"
Whats up with your chipset fan temps?
That's the internal temp sensor, the chipset is fine at 56C.
Posted on Reply
#70
Poul-erik
I simply don't understand that it can excite people like that. When you buy a new car you are told that it can run 25km on a liter, but in reality it can only drive 23km. Should we all have new engine in or, every man accepts the 23 km on a liter.
Posted on Reply
#71
king of swag187
Chrispy_, post: 4114084, member: 185623"
Clockspeeds are up slightly but boost degrades faster because temperature limits are hit faster.

The net result is probably that overall performance is about the same, but people who wanted their round numbers in synthetic benchmarks will stop whining.

For some (probably Intel shill-invoked reason) 4475 MHz peak and 4350MHz averaged over 300 seconds caused uproar because 4475MHz isn't exactly 4.5GHz.
Now we're going to get 4500MHz or even 4525MHz peak and 4250MHz averaged over 300 seconds and everyone is happy.

Same chip, just running a little hotter and a little less efficiently.

Maybe I'm weird but I'd gladly sacrifice some transient peak value for a higher long-term average clockspeed and lower temperatures.
3.5 Isn't exactly 4GB, No?
Posted on Reply
#72
TheLostSwede
Poul-erik, post: 4114405, member: 98364"
I simply don't understand that it can excite people like that. When you buy a new car you are told that it can run 25km on a liter, but in reality it can only drive 23km. Should we all have new engine in or, every man accepts the 23 km on a liter.
Again, this analogy is flawed, as there are too many variables involved when it comes to cars.
Also, no-one asked for a new engine, as this was a simple fix from AMD's side, as clearly proven.
I've not seen too much feedback from others, but in my case, it more than resolved the issue.
Posted on Reply
#73
thesmokingman
king of swag187, post: 4114421, member: 180102"
3.5 Isn't exactly 4GB, No?
You trolling? Or is that like some Nvidia reference?
Posted on Reply
#74
Chomiq
thesmokingman, post: 4114423, member: 91203"
You trolling? Or is that like some Nvidia reference?
I believe that's the reference to 970. Except I doubt that anyone sane would defend NV in that case.
Posted on Reply
#75
sutyi
W1zzard, post: 4114282, member: 1"
Just waiting for asrock taichi version to come out
Nkd, post: 4114380, member: 42675"
lol. So even though amd boost says max boost and max boost has always been single core. You want it to boost to two cores because that what you think it should do even though box doesn't say it? hmm lol.
These days when everything is multi-threaded apart from some benchmarks, yes I do think it should do that boost clock on 2 cores and it actually did and does.

1T and 2T boost behavior is almost the same and then it drops off from 3T and above. Interestingly TPUs sample of the 3600 non-X held its 4200 boost clock flat.





Posted on Reply
Add your own comment