Gadgetory

with the launch of the Rison 3000 series processors we've noticed a distinct confusion among viewers when it comes to the phrases precision boost to xfr precision boost overdrive which is different from precision boost and precision boost to and Auto Oh see there's a lot of confusion about what's considered stock what PBO even does or if it works at all or if it's been oversold as a feature by Andy and how thermals impact frequency of rise in CPUs today we're demystifying these names and demonstrating the basic behaviors of each solution as tested on two different motherboards before that this video is brought to you by us and the gamers Nexus toolkit on store documents axis net our brand new toolkit just launched and contains 10 custom made drivers for video card disassembly reap hasting and tear downs the 8 core tools are made of high-quality chromium vanadium alloy steel that's built for long service life and resistance to wear during use the other two tools are carbon steel hex heads that were custom ground down for capacitor clearance on video cards all the tools are easily mounted to a pegboard or stored in the GN made tool bag for easy transport learn more at the link in the description below precision boost overdrive is a new technology to the rise and family of desktop processors having first been introduced on the thread Ripper SKUs previously and technically rise in 3000 already uses something called precision boost to PBO with the overdrive in there is explicitly different from precision boost and precision boost to adding the overdrive is an important change this is where a lot of the confusion comes about so precision boost is not a shorthand for precision boost overdrive just to be very clear there it's a different thing precision boost is kind of like XFR andy has this extended frequency range boosting table for boosting a set number of cores a little bit higher under limited thread workload this is something introduced on Rison 1000 series and act so far has more or less been rolled into precision boost too at this point and its entirety so XFR was introduced a while ago and precision boost takes into account three numbers when deciding how the boosting behavior should perform and it's looking at how many cores should boost when should they boost and under what conditions and then how high can they boost so the numbers that are taken into consideration are PBT TDC and EDC and we'll define each of those there's also temperature and the chips maximum boost clock precision boost is enabled on a stock CPU so precision boost not precision boost overdrive to be really clear here again precision boost is considered stock precision boost overdrive is not considered stock and this is something that AMD states as well so what PBO does not ever do is boost the frequency beyond the advertised CPU maximum frequency so if the advertised frequency boost is four point six which does not mean all core by the way it means limited thread-like one core scenarios then PBO isn't going to get you to four point seven or anything like that so that's another point to demystify here will quote directly from andy's review guide just to firmly define what PPT TDC and EDC are and then we'll go through some additional explanations and a whole lot of testing we've done to try and see if these features even work properly package power tracking is PBT the PBT threshold is the allowed socket power consumption permitted across the voltage rails supplying the socket this is again a quote from the Andy review guide applications with high thread counts and or heavy threads can encounter PBT limits that can be alleviated with a raised PBT limit defaults for socket AM for is at least 142 watts on motherboards rated for 105 watt TDP processors default for socket am for is at least 88 watts on motherboards that rated for 65 watt TDP processors thermal design current or TDC is next the maximum at current in amps that can be delivered by a specific motherboards voltage regulator configuration in thermally constrained scenarios is what AMD defines as thermal design current defaults for socket am for is at least 95 amps on motherboards rated for 105 watt TDP processors default for socket am for is at least 60 amps on motherboards rated for 65 watt TDP processors electrical design current is EDC this is the maximum current in amps that can be delivered by a specific motherboards voltage regulator configuration and peak or spike condition for a short period of time default is 140 amps on motherboards rated for 105 watt TDP processors and 90 amps for motherboards rated for 65 watt TDP processors knowing all of this there's a few important things to cover next first of all this is the explanation of why TDP doesn't equal power consumption so when you see 140 watts down EPS 12 volt rails to a 3900 X and it not matching 105 watts that's why TDP doesn't equal power consumption necessary next up PBO only affects these three power limits that's what it does so the effect on CPU clock speed is in direct auto OC is a different feature by the way it's not bundled into PBO that is a separate thing entirely but it is contained within the menu PBO will never boost again the CPU past the advertised clocks at best it will allow the CPU to maintain its boost clocks for potentially a longer period of time and that's particularly going to be noticeable or have the strongest effect in scenarios where the CPU can already boost so limited Core scenarios would be the best way to demonstrate when PBO comes into play if it does depending on the motherboard you're working with because one of the constraints is thermal it gets a little tricky to test and PBO will also have less effect on CPUs that are already well cooled and are not bumping against the thermal limit so if you have a high-end cooling solution on your CPU PBO will in theory have less of an impact and this is something we'll explore today as well remember that in addition to these three power limits precision boost is constrained again by temperature and maximum boost clocks of the CPU as defined by the spec on the box these limits are not affected by PBO those last two sentences might sound contradictory on the surface of it so we need to further clarify and define a couple of additional parts of PBO and the PBO menu PBO will raise the amount of power that is deliverable to the CPU when it's too hot but it will not raise the temperature that the CPU considers to be too hot so those are two different things further platformer thermal throttling it is the the temperature at which you would encounter issues just like with Intel you can boost t.j.maxx and some of the BIOS options so platform thermal throttle emmitt is not part of PB o that has to be clarified as well PB o has three options in it that are part of PB o and then the PB o menu has additional options which are not part of PB o and this is all defined very clearly in Andy's own documentation so that's largely how we know what's part of PB o and what technically isn't but it's still a feature so the thermal throttle limit is still a feature it's not a PB o feature when we did our initial reviews for the 3600 and 3900 X we had some frequency over time plots and we can just let those run now for a moment Andy has been extremely reluctant to commit to numbers from single core boosting and all core boosting because precision Boost uses what is called an opportunistic algorithm to boost frequency until it hits one of these limits and E says the following of this quote at the limit the processor will draw back on boost and dither the frequency until the situation changes this dithering and analysis loop occurs every one millisecond within infinity fabrics command and control faculties and they further say quote the processor is designed to use thermal Headroom to drive higher average frequencies and thus more performance there's two ways to look at this and on the it's it's sort of the same line as GPU boost 4.0 so the optimistic way to look at this is that AMD is out of the box automatically pushing the CPU to more or less where it can maximally stay stable and that's why you see all core overclocking lose some of its efficacy for this generation when it was already not particularly effective in 2000 series it is less effective here and in 1000 series all core overclocking was actually pretty useful so that's one way to look at it this is more efficient than manual overclocking efficient being a key word here you can get higher with manual overclocking still in some instances and not all but more efficient certainly in terms of a power to performance output it's also more sensitive than manual overclocking and then the pessimistic way to look at it is that it's another variable out of control of the user in some ways until you disable the features and that makes it difficult to do testing or benchmarking which is not particularly a valid user environment so we don't really factor that in much and three views but it's something that people need to be aware of when they're doing comparisons of their own products or processor they've installed versus data they've seen online and it is again sort of like gpb's 4.0 on nvidia where the clock is boosting contingent upon a few different parameters primarily power and/or current and voltage as you know creating power volts times amps equals watts and then temperature so those are the key ones for NVIDIA GPU boost 4.0 and some of those factors are starting to apply to processors these days it's not like where it used to be with Intel CPUs for example where until you hit t.j.maxx you're fine you're running at spec and that's all there is to it so very different in that regard we're careful to use the same high quality NZXT X 62 liquid cooler for all of our testing that's part of our standardized test benches and we test with the fans in the pumps at max speed to avoid throttling this is all done on purpose to eliminate a potential variable you don't obviously want a CPU to throttle when you're trying to compare them all head-to-head and a like-for-like fair scenario so that's what we use for our standard testing we've added another one for some of the tests today using a lower spec and the cooler on a high spec a and the processor we'll talk about that later and if we weren't using this high-end liquid cooler then performance would be directly affected on rise in 3000 by using a lower quality cooler that can't keep the thermals at a lower temperature so this isn't just under t.j.maxx now it's it needs to be under it's it's done in steps and increments where the frequency will boost depending on how high the temperature is again GPUs 4.0 as an example boosts about every 5 degrees Celsius so a few things here all this first of all further reinforces our opinions on cases and choosing ones that have good thermals so you can go check out our back catalogue of case reviews now because it's definitely relevant here another one is the performance that our reviewer sees on a good cooler might not match what you see in a thermally constrained environment and it's all just part of precision boost to the feature because it's thermally dependent again so precision abused to overdrive emphasis on the overdrive is what AMD is calling when these three thresholds are bumped up the default setting for PBO is Auto which the stock configuration specifies is disabled so Auto should mean off if everyone's following spec further and these own review guide states to disable PBO for reviews again emphasis on the overdrive overdrive is not spec that's what we did for our review it's what most people did for their reviews if not all and it's just how we disable MCE for intel the default numbers can be pulled from the chip with overdrive disabled for the the three limiting numbers or the higher numbers can be pulled from the motherboard with overdrive enabled allowing you to bypass some of those limitations setting overdrive to enabled or setting overdrive to advanced and specifying the motherboard as the source did the same thing on the two boards we tested for this content the godlike in the x5 sony master but the two motherboards set different values on the gigabyte x5 so many master with the 3900 X P Bo limits were as follows PPT was 1200 watts TDC was 540 amps and EDC was 600 amps on the MSI godlike the limits were set to 1000 watts for 90 amps and 630 amps but at the end of the day all it really means is they're functionally disabled at this point they're so high that you're not going to hit those limits with PBO disabled the limits were PPT 142 watts TDC 95 amps and EDC 140 amps on both motherboards which is the correct and the specification for 105 watt TDP processors the limits adjusted downwards to spec when a 65 watt processor was installed with PBO still disabled it sounds like motherboard manufacturers are free to set higher numbers based on what they expect their own boards and vrm to be capable of although some of them other board makers we spoke to are under the impression that AMD is supposed to help provide these numbers so that's another issue entirely and it's a teething issue because the limits change based on the processor U is 105 watt vs. 65 watt for example this means that the advantages of PBO are dependent on things like the motherboard model specifically the processor model the specific process sir not the skew but the individual unit the cooling solution the temperature of the room and the alignment of the planets that brings us to a couple of other things like Auto OC so the 200 megahertz offset that's been mentioned a lot in conjunction with PB o is called Auto OC it's a separate AMD feature it's not technically part of P Bo but it is part of the PB o menu in BIOS so and these put it in the same menu it is not however a part of pbl so it's a bit confusing the way BIOS is laid out but that's how that's how the spec explicitly defines it PB Oh sounds good it is good in a lot of ways this allows users to circumvent some of the tachi GPU ask clock behaviors that are starting to emerge on CPUs these days and that's great especially for users who have the vrm and cooling solutions they need to handle those higher speeds to bypass potential power and and other limits that we've gone over and so on paper it sounds like a great way to get rid of the vastly diminishing returns from all court over clocks over the past generation of AMD processors and give overclockers access to a little performance via a different overclocking option and the intense PBO and auto OC to be used in conjunction with each other with Auto OC more useful in theory for single core boosting and PBO mostly benefiting the heavier core workloads where you're saturating more of them and challenging that power limit but the theta is in our testing none of this helps at least not much and certainly not for us further note that entering the number 200 does not mean it will actually offset by 200 it's more of a you can offset within this limit and in that regard again we look back at GPU clocks where you set an offset it doesn't mean it's going to actually offset by that number because the base will change so the base from which you are offsetting is variable and the actual offset is not necessarily equal it's the 200 number you might enter into auto Oh see we have a table from AMD that helps illustrate the intended use and these reviewer guide says that the offset only apply to the max boost clock but a chart later on clarifies that this is the max boost clock for any number of course the risin nine 3900 X has a max listed boost clock of 4.6 gigahertz it may or may not ever hit this clock speed depending on the CPU itself and the usage and probably not on more than one core with Auto OC this theoretical limit is raised to 4 point 8 gigahertz but there's still no guarantee the chip will ever hit that from AMD quote this feature will not guarantee the higher boost clock on any number of course the frequency cores booster ation will still depend on the firmware manage the limits even though those limits are higher than OEM when PBO has been enabled by this direction enabling PBO and the offset make it more likely that the 3900 acts will maybe hit 4.8 gigahertz on one core once in a while in Cinebench single threaded benchmarks and sometimes in extremely heavy all core loads the CPU will run a little faster than it would with stock PBT TDC or EDC limits although the all core load is barely affected by the 200 megahertz auto OC offset in our testing and spoiler alert there we'll get to that once we had a handle on what PBO and auto OC are supposed to do we started testing we began with an r5 3600 on the gigabyte horas master used for a verizon 3000 reviews and the msi godlike reasoning that the most power hungry sample chip we had with the lowest stock clocks would benefit the most from removing current limitations this is the CPU that took more than one point 4 volts as a reminder to complete our tests at 4.3 gigahertz all core when the 3900 X could manage the same at 1.35 it's also the model of CPU that Andy showed gaining the most points in single threaded performance with PBO and auto OC we have charts of that from their own reviewers guide as well note that we're talking stock numbers with PBO as the a/b test here so all core o'seas are not part of this just a reference point a messiah has another menu in addition to the generic and the PPO submenu with some basic preset PPO profiles in it we verified that all these profiles do is set PPO to Advanced Mode and max all the settings except Auto C which again is technically a separate feature so what at MSI I was paying attention to Phoebe Oh enough to create these profiles so we can assume that they've tailored the motherboard PBT TDC and EDC limits to fit their boards in the interest of not making this piece 10 hours lon will summarize the tests we ran and discuss all the results together we tested the r5 3600 and r9 3900 X in the gigabyte X 570 or as master and MSI x5 so many god-like motherboards and we ran them with PBO disabled we also tested PPO enabled to auto PPO maxed with a plus 200 megahertz auto OSI offset including max in thermal throttle limits and PPO scalar menu options we also tried them minimized we found that manually maxing the PPO variables and allowing the motherboard to set them accomplished functionally the same thing so for quote max tests we used the motherboard values we ran a mix of Cinebench r15 R xx and 1080p game tests we didn't run all tests on all configurations because it would have been a waste of time the tests and the charts are the tests that we ran we also ran to further passes on the 3900 acts in the godlike board with the CLC fan and pump speeds manually slowed we tested with PP o on and off to see whether overdrive is more helpful when the CPU is warmer by again slowing down the CPU fan speed on the CLC to a point where it was running close to 85 90 degrees somewhere in that range on the X 570 master with the r5 3600 on PBO disabled Cinebench are 20 scored 37:59 multi-threaded 484 single threaded simply setting PB o 2 enabled scored 37 48 and 485 and setting PB o to motherboard limits functionally the same as max maxing the throttle temperature in PB o scalar and applying the auto OC 200 megahertz offset scored 37 65 and 496 that's a 0.2% improvement over stock for the multi-threaded score absolutely within range of variants and a 2.5 percent improvement in single core performance AMD showed a 21-point increase for one threaded performance in this test we showed a 12 points increase in hours and we can't tell you the percent difference for AMD's numbers because they omitted a vertical axis on their marketing charts attempting to render them as useless as possible cinema charge one a single-threaded on the r5 3600 is and these absolute best case scenario for PBO and auto SE and it barely does anything in their own testing if their stock single core score somehow exactly matched ours and it should be very close and these 21 points would be a 4.3 percent increase on the godlike board we saw a 3.5 percent increase from PPO disabled to PBO max with auto OC we also ran Cinebench r15 so let's take a look at the r9 3100 X for that benchmark the lowest single threaded score we saw was with PPO and auto C on the horse board and this is where our next big hurdle started to become visible precision boost is so reactive that the variance between test results from one day to the next can be larger than the difference between OC and stock or stock and PB o for a more logical comparison the godlike improves to 2.5 percent moving from PPO off to PPO maxed with 200 megahertz auto OC but then zero single threaded improvement we didn't bother running more synthetics than sentiments for this content piece because PB o is supposed to be primarily beneficial to gaming and because runs run variants with precision boost alone was starting to drive us insane also this is the one AMD used anyway to demonstrate it we demonstrated why precision boost was driving us crazy in our live stream it's it's not a bad thing ultimately and we'll talk about that at the end but from a testing standpoint it's difficult I will work on another content piece separately to really drive at home to basically every couple degrees change in CPU temperature will affect frequency and a measurable way for results and so the order of the tests completed it will impact results the pause between the tests will impact it the room temperature and more we control all these things as best we can especially room temperature but it's still a tough fight these things are sensitive this is partly a good thing from the standpoint of improving the CPUs performance but again difficult for tiny testing differences and he doesn't guarantee auto OC will actually increase the clocks and PPOs effect is inconsistent at best stick around for the end of this piece where we dig into precision boost behavior at sub-zero temperatures for some explanation of why for game testing we felt things out first by testing with civilization 6 a consistent benchmark that should benefit from frequency increases across a limited number of threads on the 3900 acts turned time completions were almost exactly the same regardless of the cooler or if PBO or otto OC was enabled all the 3900 acts results were gathered on the gigabyte board the test with the CLC and PB o disabled may have been marginally better with the air cooler but a 1.2 percent reduction in turn time isn't concrete results for the 3600 on the MSI board with PB o and Auto AC on and off were identical and the results on the gigabyte board were so close to each other that they may as well have been we have a more limited set of results for total war Warhammer twos battle benchmark just the 3900 acts on that messy board with an air cooler and a liquid cooler for once the results somehow lined up in a sane and reasonable fashion with the to PBO and auto OSI enabled tests at the top and the two stock tests at the bottom with the best results from the CLC and the worst result from an air cooler there's a 2.6 percent improvement between the worst results and the best results Patrick who wrote most of the script - noted that he hopes everyone can appreciate the amount of time and effort he put into testing a feature that nobody understands and that barely does anything and the nobody understand part is definitely true based on our live stream where most people think PB o is the same as precision boost and this is not the fault of the viewers it's the fault of AMD for choosing these names the campaign benchmark also showed some scaling on each of the board and cooler combinations the improvement from PB o off to PB o Otto OC on with the 3900 X and a CLC on the MSI board was a whopping 2.4 percent and 1.5 percent on air despite being a lightly threaded workload it doesn't seem that the total war benefits are massive from the 200 megahertz offset but then again nothing else has been either it's pretty small if measurable in some of the instances it's it's not the 3600 scaled 1.6% with PB o and Otto of C enabled on the gigabyte board we would normally avoid such a patchy mix of hardware and tests but we've been trying to cast as wide a net as possible to find any situation in which PB o and Ottawa C provide a genuine advantage Shadow the Tomb Raider showed no significant change based on PB o or Auto OC on them a cyborg with a 3900 X air cooler or not even we're on the gigabyte board with the 3600 it's hard to get results of this unchanging when we actually want them again the GTA results for the 3900 acts were almost the same all of them regardless of contact result with an air cooler in PBO and Ottawa sea enabled was very slightly worse than the other three presumably because of minor temperature fluctuations from repeated loads in rapid succession or from a reasonable plus or minus one degree Celsius change in room temperature as the testing environment fluctuated but it's still a difference of barely more than one FPS from the top result which was achieved with PBO and auto OC disabled pr5 3600 on the gigabyte board also scored two results within margin of error the worst result for the 3900 X and F 1 2018 was with the air cooler and PBO and auto OSI enabled the best result was with a CLC and PPO turned off our initial theory that PBO would help get around thermal constraints has not panned out Illustrated here the difference is still a minuscule 1.1 percent gain between the worst and best results however so the it's kind of a wash it's it's hard to tell when the difference is real versus not with a 3600 on the gigabyte board there's also less than a 1% increase between the two scores our final gaming benchmark was hitman 2 the X 12 the best results achieved through the 3900 acts was with PPO and auto OC under a CLC close with liquid cooler and the worst result was with an air cooler and PPO disabled the gap is still so incredibly small it's even smaller with the 3600 on the gigabyte board we failed to find a single game we're enabling PPO and Auto OC made a significant difference and PB o plus Auto C is intended to improve gaming performance and these marketing video says that will cut to the clip itself which allows potentially the CPU to run at a higher clock speed for a longer period of time and more clocks be for longer means more performance especially in games they just love extra clock speed so the clip literally says more clock speed for longer means more performance especially in games they just love that extra clock speed we could test more combinations of boards and chips but it just isn't happening without completely changing the definition of PB o the power delivery capabilities of the board's we tested is high the rise in master software reports that the 3900 acts in 3600 both fall well below the motherboard defined PBO limit and motherboard manufacturers could triple PPO limits tomorrow if they're even the ones who control those numbers they seem some of them are confused on if that's true and it still wouldn't make a difference based on what we've seen for the final charts we'd like to show some of the differences in power and frequency behavior with PPO on and off we did a lot of logging here but we picked two runs as representative from the 3900 X and the godlike board one with P Bo disabled completely I don't know with PB o and Abel then set to the motherboard limits with a 200 megahertz Auto Oh see we started hardware info logging to put a current clamp around the CPU EPS 12 volt connectors and ran a Cinebench R 20 Ron first multi-threaded then single threaded in the same log file as you can see on the screen for these results we scored a 514 and 70 78 for one thread and multi thread and PPO off we had 520 and 70 to 35 with PBO and auto OC on so you can see the clocks but we've scored better with this chip and board with PBO disabled and there's still only a 1.2 percent improvement one thread and 2.2 percent improvement multi-threaded but if any pass of any test it's going to show a difference in frequency this is it we're showing the peek core clock here because that's the boost clock which PBO and auto se are supposed to explicitly benefit and because the average core clocks are even closer to identical during the multi-threaded load at the beginning PBO and auto SE maintains a 50 megahertz average but during the single threaded test any difference is imperceptible although the higher result with PBO and auto asean indicates there must be some delta averaging the peak frequencies during the single threaded runs comes out to roughly 45 14.4 megahertz with p vo off and 45 19 point 5 mega hertz with PB o and HUD host Heon v core as reported by hardware info had a more obvious delta during the multi-threaded test holding a 1.2 for volt average for PB o off and 1.3 volt average for PB o and auto OC the single threaded test again shows barely any difference and in fact taking an average has PBO disabled ahead of the other results at one point were seven versus one point four six but that difference is negligible and is noise our preferred method for logging powers with the current clamp with PB o disabled we log the eleven point six amps multi-threaded three point three single threaded or 130 9.2 watts and 39.6 watts and with PBO and auto c-max we log between 13.9 amps and 14.2 apps I'm holding threaded and 3.4 M single threaded or 170 point 4 and 40 point 8 watts the multi threaded amperage actually spikes to 14 point 8 at the beginning but dropped as low as thirteen point nine during the fairly brief run which was corroborated by hardware info power numbers by this point in testing we had seen so much variance in clocks and scores that we knew precision boost was affecting our results we have liquid nitrogen on hand so it was easy for us to just dump some of it onto the 3900 X and see what it would do under absolutely ideal conditions stock or with PBO will expand on this more in a separate piece but for now we can say that Cinebench r15 running with TDI at about 62 degrees Celsius scored 31 63 points and that score scaled all the way up to 34 27 points followed by a blue screen at minus 56 degrees Celsius for the score and then blue screen at about minus 70 or so depending and similar more stable scores were achieved that minus 30 C for the record these were just some rough tests to see if PB would continue scaling clocks with temperature down to sub-zero and it did that's just precision boost not overdrive nothing else was changed no overclocking was done overdrive was not enabled we just made it colder and ran it stock we were maintaining up to nearly 4.4 gigahertz all core and Cinebench completely stock PPO disabled higher than the highest all core overclock we could achieve with a 280 CLC and all without voiding the warranty precision boost overdrive voids your CPU warranty not that they can check that but dumping Ellen 2 onto it and leaving it stock technically doesn't void the warranty in fact enabling PBO and 200 Hertz auto se you locked us wait maximum 4.2 gigahertz all core and 4.25 with auto OC disabled this could be down to gigabyte at this point but we did test time aside for the other benchmarks he saw earlier so we still need to replicate this behavior on the other boards it's bizarre either way we're never seeing the 200 megahertz boost from auto se especially in light of the LNT results we've gathered so far it seems like PBO is largely pointless the limits that it's raising aren't the limits that we're encountering the limit that's far far more important is temperature to the extent that for the CPU for once buying a closed-off case which is an actually good airflow case will be the difference in clock speeds and not to give away the whole future content piece but as we showed in the stream if you run at say 78 degrees Celsius which is not unreasonable on a 3900 X you're looking at maybe 40 50 megahertz average and 72 you're looking at forty seventy-five megahertz average you come all the way down to something like 55 degrees Celsius and now you're adding 200 megahertz to the average core clock with PBO disabled just running precision boost to which are different things again so if PBO doesn't seem to do a whole lot here if anything and yeah this even at the Wraiths spire that isn't rated for 3900 X temperatures that we forcefully ran we use the worst cooler than and then it comes with just to force a bad environment PBO didn't make a significant difference so we'll work on nailing down which settings are restricting the clocks to 4.2 on that gigabyte board but even with the msi board we're just we're not really seeing anything and we can't make wide sweeping statement saying that pv is broken or anything like that because maybe it's working as intended or maybe some of the board's it works on but not on others we haven't looked at as rock for example haven't looked at Asus yet but for these two boards which do represent a large part of the market it appears to do nothing and our understanding is that this behavior should largely stem to the other boards too because AMD provides the bulk of this code so regardless of the verdict we still haven't discovered a single scenario in which PBO makes a significant impact to performance we also have to include the caveat that we don't think rising 3,000 processors are bad unfortunately some of the people who watch this will not have seen our other videos or because people have really short memory when it comes to being enraged on the internet will have forgotten them and will think that because we're saying PBO isn't really worth playing around with and doesn't seem to do much that that somehow translates into us saying the Rison processors are bad so the mark there was some marketing here that was not good and the feature doesn't seem to do what we expect or what we've been led to believe it does but the processor is on the whole are still good we're still recommending buying the 3600 but unfortunate we have to include these notes because of how people just kind of read comments these days and not actually watch the content but anyway precision boost is is from a reviewer standpoint obviously it's a finicky pain that's P P and not P vo and that's because it's harder to control for just like GPUs are hard to control for so for a reviewer standpoint everyone needs to be very careful about things like the environment temperature and we control it hard I mean we were our AC bill is gonna be insane this month because we were running at the same rough 21 22 degrees Celsius setting for all of the tests but even that with that control you're still looking at potential for differences depending on what order you ran the test and then how long the CP has been under load so that's something reviewers need to pay attention to from a consumer standpoint it's probably a good thing because you're effectively maxing out with a CPU can deal without any effort and so from a consumer standpoint always unfortunately except for enthusiasts removing the ability to overclock by way of providing all the performance out of box it's kind of hard to argue against that unless you're an enthusiast so anyway what we can say definitively is that precision boost and precision boost overdrive and auto or C are horribly named features they are very poorly explained even by a in these official videos and they barely do anything in terms of naming other than served to confuse the discussion so we in the stream that was the best example because every 15 20 minutes I had to correct people who were saying he's using P Bo right now no we were just using precision boost they're different it's not the same so that's again that's not the consumers fault it's Andy's fault for naming them confusingly so anyway that's our deep dive on precision boost Patrick spent several days on this and there's a lot more we could do yet but but we got to call it here so there's more we can do with different boards going to older boards things like that seeing where it works that'll wrap it for now so now you know what precision boost is supposed to do and what it does or doesn't do subscribe for more go to store I Kara's access dotnet to help us out directly like by picking up one of our toolkits or you can go to patreon.com/scishow and access to helps out there as well thanks for watching we'll see you all next time