[SOLVED] Does PBO degrade the CPU?

[andy0x2a]

Hi,

I have not been able to conclusivley find anything online, so I thought I would ask the experts here.
I want to enable PBO, but I don't want to degrade my CPU. I have read that PBO voids the AMD warranty; is that because it pushes the CPU to unsafe levels?

I have a 3800X with an MSI x57-a-pro motherboard, and EVGA 850W PSU. Cooling it is an h100 Pro XT.

I have tried setting PBO to motherboard settings, with a +300 Mhz, and have noticed that hwinfo shows a few of the cores going up to 4675. Which is nice. Idles at 28-32 (ambient +7), dev tasks are low 50s, compiling brings it up to low 60s.

So my question is: will PBO degrade the CPU, and does scalar value matter for CPU degredation? Is there an optimal/safe value for the scalar?

Thanks

 

[zx128k]

Or run it at extremely high temperatures. PBO done right leaves the processor's boosting algorithm in charge of both voltage and clock speed, so it can lower either/both as processing load rises to keep temperature in safe range.

But as far as degrading your processor goes...well, just using your processor degrades it and using it harder degrades it faster. If you have enabled PBO in order to use your processor harder then yes, it will degrade faster if for no other reason than you use it harder.

And then there's the question of how much faster? is it enough to matter when most of us live on 5 year upgrade cycles anyway? I seriously doubt it. But if it should degrade just remove the PBO and use it at full-on stock to return to stability, which is what you'd have been doing anyway.

Right voltage and like buildzoids 3700x you will need higher voltage at stock.

 

[drea.drechsler]

Right voltage and like buildzoids 3700x you will need higher voltage at stock.

omg...i don't have the fortitude to handle a buildzoid video right now.

but as i recall, that was as much an issue of that particular motherboard as anything else. My motherboard's pretty bad too, so I rely on the STI2 TFN voltage which the CPU reports and is pretty much independent of motherboard mashups.

 

[andy0x2a]

Thanks everyone for the thoughtful discussion.

I guess I will have to live with no warranty, and the chance of degredation. I am running it with a 1x scalar, to ensure my CPU makes it to atleast zen-3 release, and I am hoping the next version of zen will fit into the AM-4 socket.

I figure that will give me enough lifespan with a 1x scalar on PBO. And I am more than happy seeing 4675Mhz as a max frequency for my CPU.

 

[drea.drechsler]

... And I am more than happy seeing 4675Mhz as a max frequency for my CPU.
...

Plus you're hitting over 4500 or over on every core! That's very good boosting ... a prime example of how well binned 3800's are!

In contrast, I'm getting 4400 on my 3700X on only 5 cores, occasionally a 6th. I can get all 8 to hit 4400 but I have to overvolt a tad. But since it only boosts one core at a time it doesn't really matter so I'm OK with what I got.

I'd suggest that if you wanted to undervolt a bit by dialing in one or two notches of negative offset voltage it might bring the numbers down to 4500 across the board (which is spec for the processor). It shouldn't hurt single thread performance much and might actually help with multi-thread performance. The lowered voltage keeps the temperature lower under heavy loads, so it won't dial clocks back as soon or as far. Something easy to experiment with, and perfectly safe.

 

[andy0x2a]

Thanks. I will try undervolting it to see if it makes a difference.

 

[CountMike]

Turbo, boost, overdrive, PBO XFR, XMP.... can't be considered overclocking because it's just bringing affected parts to best specs manufacturer claims. If that poses a risk, it's manufacturer's risk they have to take and can't be a warranty restriction.
By some analogy, car manufacturer can't void warranty because somebody drive a car 100Mph although they claim it can go that fast.
Both, Intel and AMD were taken to task for misleading product information. in the past.

 

[zx128k]

Turbo, boost, overdrive, PBO XFR, XMP.... can't be considered overclocking because it's just bringing affected parts to best specs manufacturer claims. If that poses a risk, it's manufacturer's risk they have to take and can't be a warranty restriction.
By some analogy, car manufacturer can't void warranty because somebody drive a car 100Mph although they claim it can go that fast.
Both, Intel and AMD were taken to task for misleading product information. in the past.

The specs for proper operation are given on AMD's website. These are the stock speeds. PBO is a feature to help you overclock the cpu by removing power and current limits, increasing boost frequencies by maintaining higher vcore for longer with scalar. PBO increases TDP beyond stock and thus your cpu can overheat by hitting 95c. You can remove that throttling limit for temperature on my motherboard as well as part of PBO.

PBO is just a better version of Intel XTU were the AMD overclocking software leverages the normal boost algorithm and tweaks it beyond box specs. Stock operation is the fused values, PBO is a software override for these fused hardware values. Stock operation is Precision Boost 2 but overclocking is Precision Boost Overdrive. AMD spells this out of their website. This only becomes a problem if you tell AMD about how you baked your cpu by overclocking too far. So long as the issue can happen at stock and AMD can't prove you did anything wrong then you should be fine. If you delid, lap the IHS etc then AMD are going to get an ideal about what you were doing with your cpu.

Taking the car manufacture as an example. You have just put a turbo (more average core voltage, higher power and current limit and water cooling on the PC) on the car engine but failed to tune the car correctly using software (BIOS setting PBO etc). To much boost lead to failure. The engine is now dead and the manufacture wont honor the warranty because of improper use and testing.

 

[CountMike]

PBO is a feature of the processor, not some feature MB and BIOS to overclock it. If 3700x for instance is advertised with base speed of 3.6GHz on all cores and up to 4.4GHz on one core, nothing in between can be considered an OC. Those are just specs AMD says they are for that model. Now, if you were to set clock to 4.3 or 4.4GHz on all cores, that could be considered an OC even if you don't exceed max single core frequency.
You can say it's similar if not same with XMP which is set of settings manufacturer recognizes and supports for best stock performance of given RAM. Memory advertised as for instance 3600MHz can still have base frequency of 2133MHz but that doesn't mean 3600 is overclock.

 

[drea.drechsler]

....
By some analogy, car manufacturer can't void warranty because somebody drive a car 100Mph although they claim it can go that fast.
....

Calling PBO 'overclocking' is just sliding the meaning of the term around. So long as there are no official and unvarying definitions, it's pointless to argue about it.

But more to the point: even true 'overclocking' in and of itself is arguably not going to void your warranty! It's not even mentioned in the legalese version of AMD's limited warranty only abuse or misuse is. Even if it was mentioned, voiding their limited warranty can not have any effect on the implied warranty of fitness. The marketing, advertising and literature makes clear these processors are fit for the purpose of 'overclocking' and that makes for an un-voidable implied warranty of fitness.

But about that analogy: if you had to bypass a governor to go over 95MPH they could invalidate the warranty until you bring it back to stock. And then, if the reason you brought it in for repair was because the engine was damaged by misuse/abuse/racing (common terms I've seen in my autos' warranty booklet) they could void their limited warranty for the engine.

In the case of an engine it's much cheaper relative to repair cost to determine that it was abused. But a microprocessor is way more expensive compared to simply replacing it. And besides, it's not going to be damaged by PBO in any reasonable time frame and they know it. As long as voltage and frequency is left in 'auto' it throttles itself, by design, until it's safe.

The motherboard VRM, however, is another story. If you enable PBO and set high enough values, a weak enough VRM may toast itself trying to feed a 3950, for instance, the power it needs.

 

[CountMike]

Calling PBO 'overclocking' is just sliding the meaning of the term around. So long as there are no official and unvarying definitions, it's pointless to argue about it.

But more to the point: even true 'overclocking' in and of itself is arguably not going to void your warranty! It's not even mentioned in the legalese version of AMD's limited warranty only abuse or misuse is. Even if it was mentioned, voiding their limited warranty can not have any effect on the implied warranty of fitness. The marketing, advertising and literature makes clear these processors are fit for the purpose of 'overclocking' and that makes for an un-voidable implied warranty of fitness.

But about that analogy: if you had to bypass a governor to go over 95MPH they could invalidate the warranty until you bring it back to stock. And then, if the reason you brought it in for repair was because the engine was damaged by misuse/abuse/racing (common terms I've seen in my autos' warranty booklet) they could void their limited warranty for the engine.

In the case of an engine it's much cheaper relative to repair cost to determine that it was abused. But a microprocessor is way more expensive compared to simply replacing it. And besides, it's not going to be damaged by PBO in any reasonable time frame and they know it. As long as voltage and frequency is left in 'auto' it throttles itself, by design, until it's safe.

The motherboard VRM, however, is another story. If you enable PBO and set high enough values, a weak enough VRM may toast itself trying to feed a 3950, for instance, the power it needs.

VRM is part of motherboard and MB manufacturers shouldn't advertise compatibility with processors that need so much power at their stock settings (including boost, pbo..... ) if it can't take it in long run. They do state max TDP of processor and then list processors above that in their compatibility list.

 

[drea.drechsler]

VRM is part of motherboard and MB manufacturers shouldn't advertise compatibility with processors that need so much power at their stock settings (including boost, pbo..... ) if it can't take it in long run. They do state max TDP of processor and then list processors above that in their compatibility list.

I can agree with that... but then I also believe most motherboards have a capable enough VRM to handle it in spite of what we may think of them. FET temperatures may get to 100C doing it, but when FET's are rated to 125C and the VRM force-throttles the CPU at 110-115C is that as bad as we think? But I also don't know how many people have used a 3950 with those weak(er) VRM'd B350 (for instance) motherboards that lack decent heatsinks and PBO'd them with the 'EDC=0' bug exploit to put them to the test.

And lastly: Processor TDP ratings have proved to be useless for what we think they're useful for. The better tech reviewers (Steve at GamersNexus in particular) are very critical of what they've become, whether AMD's or Intel's. It's not surprising MB MFR's would do that.

 

[CountMike]

I can agree with that... but then I also believe most motherboards have a capable enough VRM to handle it in spite of what we may think of them. FET temperatures may get to 100C doing it, but when FET's are rated to 125C and the VRM force-throttles the CPU at 110-115C is that as bad as we think? But I also don't know how many people have used a 3950 with those weak(er) VRM'd B350 (for instance) motherboards that lack decent heatsinks and PBO'd them with the 'EDC=0' bug exploit to put them to the test.

And lastly: Processor TDP ratings have proved to be useless for what we think they're useful for. The better tech reviewers (Steve at GamersNexus in particular) are very critical of what they've become, whether AMD's or Intel's. It's not surprising MB MFR's would do that.

Well yes, TDP is not reliable measure, MB and VRM wise it would be more helpful to know wattage and amperage a certain MB can withstand at recommended max temperatures.
I have seen some test where someone stuck 2900x on a A320 MB but not sure how long that lasted before white smoke.

 

[drea.drechsler]

... it would be more helpful to know wattage and amperage a certain MB can withstand at recommended max temperatures.
...

How so often I've wondered why they don't furnish that kind of data. We have to wade through Buildzoid's very informative but unmerciful ramblings, and hope he happens to cover either our motherboard or one with an identical VRM topology, to have any idea of what it might be.

That's just not right.

 

[zx128k]

PBO is a moot point, I can't find anyone that states that their cpu degraded because they use PBO.
There are still reports that high static vcore does degrade and running higher than Fmax.
View: https://www.reddit.com/r/overclocking/comments/bbxot0/psa_they_werent_kidding_about_how_fast_ryzen/

https://www.youtube.com/post/UgwIDprhUtrJzFuhW0N4AaABCQ

Strictly technical: Matisse (Not really)

07/08/2019 6:33 PM (GMT) - Update on the bios issue on Crosshair VIII Hero motherboard ("the thing"). Earlier today I received a response to my inquiries from ASUS. The response was rather technical and I cannot go into the specifics of what exactly it involved. However, it confirmed my...

www.overclock.net

Precision Boost Override is a motherboard feature. https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/page-72#post-39391302

The "Precision Boost Override" feature available on 400-series motherboards allows increasing the physical limiters mentioned earlier. On SKUs belonging to the 105W TDP infrastructure group, the default limiters are following: PPT 141.75W, TDC 95A, EDC 140A and tJMax of 85°C (absolute, excl. offset).

When "Precision Boost Override" mode is enabled (AGESA default), PPT becomes essentially unrestricted (1000W), TDC is set to 114A and EDC to 168A.

FIT is still enabled with PBO.

Where is the limit?

The maximum safe voltages for CPUs are an eternal riddle, as neither of the two manufacturers release this information for public consumption. Public or even the NDA documents generally specify a vague limit, which most of the time relates to a point where the catastrophic failures become more common instead of specifying the voltage that is safe to sustain without causing any damage to the silicon. Such limit is admittingly rather hard to specify, as the limit will vary between the different CPU specimens (silicon variance, SIDD) and operating scenarios (peak current in different utilization scenarios, temperature, etc.).

In order to get the most accurate answer for this question I ended up “asking” the CPU itself. As stated previously, the CPU features various different limiters / safe guards (Package Power Tracking: PPT, Thermal Design Current: TDC, Electrical Design Current: EDC, thermal protection and FIT).

“FIT” as the name suggest is a feature to monitor / track the fitness of the silicon and adjust the operating parameters to maintain the specified and expected reliability. Many semiconductor manufacturers utilize such feature to eke out every last bit of performance, in an ERA where most of the semiconductors are process bound in terms of performance. In short: FIT feature allows the manufacturers to push their designs to the very limit out of the box, without jeopardizing the reliability of the silicon. A practical example would be the knock sensors on an engine. The control unit of the engine always tries to advance the ignition timing as much as possible, to produce the best possible power / torque figures. The purpose of the knock sensors is to listen if knocking occurs and tell the ECU to reduce the timing advance when it does, in order to protect the engine.

To see what the actual maximum voltage FIT allows the CPU to run at in various different scenarios is, I disabled all of the other limiters and safe guards. With every other limiter / safe guard disabled, the reliability (FIT) becomes the only restrain. The voltage command which the CPU sends to the VRM regulator via the SVI2 interface and the actual effective voltage were then recorded in various scenarios. In stock configuration the sustained maximum effective voltage during all-core stress allowed by FIT was =< 1.330V. Meanwhile, in single core workloads the sustained maximum was =< 1.425V. When the “FIT” parameters were adjusted by increasing the scalar value from the default 1x to the maximum allowed value of 10x, the maximum all-core voltage became 1.380V, while the maximum single core voltage increased to 1.480V. The recorded figures appear to fall very well in line with the seen and known behavior, frequency, power and thermal scaling wise.

The seen behavior suggests that the full silicon reliability can be maintained up to around 1.330V in all-core workloads (i.e. high current) and up to 1.425V in single core workloads (i.e. low current). Use of higher voltages is definitely possible (as FIT will allow up to 1.380V / 1.480V when scalar is increased by 10x), but it more than likely results in reduced silicon lifetime / reliability. By how much? Only the good folks at AMD who have access to the simulation data will know for sure.

These figures will almost certainly vary between the different CPU specimens (due to SIDD and other silicon specific factors), however the recorded values were almost identical on all of the tested samples (within 20mV, lowest-highest leaking specimen).

Also note that the figures stated here relate to the actual effective voltage, and not to the voltage requested by the CPU. The CPU is aware of the actual effective voltage, so things like load-line adjustments and voltage offsets will modify the CPUs voltage request from the VRM controller accordingly. The most accurate method to measure the effective voltage on AM4 platform is to monitor the “VDDCR_CPU SVI2 TFN” voltage, which is available in HWInfo. This reading is sourced directly from the VRM controller (through SVI2 interface) and generally it is the most accurate reading available to end-users by far. As a side note, while the TFN (“telemetry function”) voltage readings are always generic (and accurate), never blindly trust the reported current and power readings (as every motherboard model needs separate calibration).

From AMD.

Q: WHAT VOLTAGE IS SAFE?

A: The one the CPU picks on its own. The firmware is actively monitoring and managing more chip health data than you can possibly imagine or digest--every second--to ensure performance is being reliably maximized for n cores under load at y temperature for x time. Stepping in to fix a static vcore, because you "feel" something about what the CPU is doing on its own, or you misread what someone said on the internet, won't help you. It will likely subtract performance (reduced Vmax = reduced perf), and it will likely raise the average voltage the chip experiences over a 24H period vs. what it will self-manage to on its own accord (you've disabled idle voltages). If the purpose of setting a manual vcore was to reduce voltage, you missed: you actually just raised the average voltage.

OKAY, BUT WHAT ABOUT OVERCLOCKING?

A: Experiment. Just like every other processor. There are some very smart people doing this work out there. 😉

View: https://old.reddit.com/r/Amd/comments/ck6bji/all_this_confusion_about_voltages_and/evlt3gh/?context=3

 

[drea.drechsler]

PBO is a moot point, I can't find anyone that states that their cpu degraded because they use PBO.
There are still reports that high static vcore does degrade and running higher than Fmax.
View: https://www.reddit.com/r/overclocking/comments/bbxot0/psa_they_werent_kidding_about_how_fast_ryzen/

....

It was hard to tell exactly what he was doing...but it appeared to me he was running that high, fixed voltage while bench-marking. For some sort of competition perhaps? He went on to say temp was sitting at 94-95C, which he was probably aiming for to be just below throttling, and on air.

The temperature, at that voltage (1.475) and therefore current, is what did it in. At that kind of temp, voltage better be down around 1.25 or less or it WILL degrade rapidly! That's why NEVER used fixed voltages! only AUTO or offset, which the way you use PBO when you do it right.

But that's an extreme overclocker for you, it's what they do to get the big benches. He didn't say what clock he was running...maybe later on, these threads are hard to follow with all the commentary.

Even after 'degrading', though, it apparently operated at stock settings (which we're all trying to do anyway)...he just couldn't get as good of PBO results as before. I'm curious what he meant by "PBO Level 3", which was how he determined his processor degraded. Do those top-end overclocking boards have more settings for PBO too? Or might that be a level of performance while using PBO? He was on a Crosshair VI Extreme overclocking board.

 

[zx128k]

It was hard to tell exactly what he was doing...but it appeared to me he was running that high, fixed voltage while bench-marking. For some sort of competition perhaps? He went on to say temp was sitting at 94-95C, which he was probably aiming for to be just below throttling, and on air.

The temperature, at that voltage (1.475) and therefore current, is what did it in. At that kind of temp, voltage better be down around 1.25 or less or it WILL degrade rapidly! That's why NEVER used fixed voltages! only AUTO or offset, which the way you use PBO when you do it right.

But that's an extreme overclocker for you, it's what they do to get the big benches. He didn't say what clock he was running...maybe later on, these threads are hard to follow with all the commentary.

Even after 'degrading', though, it apparently operated at stock settings (which we're all trying to do anyway)...he just couldn't get as good of PBO results as before. I'm curious what he meant by "PBO Level 3", which was how he determined his processor degraded. Do those top-end overclocking boards have more settings for PBO too? Or might that be a level of performance while using PBO? He was on a Crosshair VI Extreme overclocking board.

Most likely scalar x3. That is a big deal, the real performance increase is at scalar 10x. I am hoping that with PBO even using the bug, FIT still keeps the CPU in the save zone. With the edc bug and scalar 10x you get really good performance.

 

[drea.drechsler]

Most likely scalar x3. That is a big deal, the real performance increase is at scalar 10x.....

Even that is confusing... and alone is not a good indicator of degradation. Ryzen's boosting algorithm is sensitive enough to temperature that you have to control for ambient when comparing performance even with identical settings. Just positioning of components around an open air bench, thereby influencing air flow over the cooler or radiator, can have influence. That's why comparing CB20 multi-thread scores ranging +/-100 points can be considered effectively the same.

I can't imagine very many people are careful enough to control their setups and work environment to effectively measure a degradation so fine as one, or even 5, scalar levels in difference.

The usual way to assess degradation is watching a processor drop threads at the same voltage, clock speed and settings where it was previously stable on a Prime95 run of a definite duration.

 

[zx128k]

Even that is confusing... and alone is not a good indicator of degradation. Ryzen's boosting algorithm is sensitive enough to temperature that you have to control for ambient when comparing performance even with identical settings. Just positioning of components around an open air bench, thereby influencing air flow over the cooler or radiator, can have influence. That's why comparing CB20 multi-thread scores ranging +/-100 points can be considered effectively the same.

I can't imagine very many people are careful enough to control their setups and work environment to effectively measure a degradation so fine as one, or even 5, scalar levels in difference.

The usual way to assess degradation is watching a processor drop threads at the same voltage, clock speed and settings where it was previously stable on a Prime95 run of a definite duration.

Its that you start getting errors, BSOD and restart when a processor degrades. So it's likely he's getting one of the three at scalar values >3. Then there is clock stretching when vcore is too low, so performance dies.