We need good writers, he's fine where he's at. He can always do consulting like Anand did. This is a great investigative piece and i'm glad Anandtech has linux guru's who can make their own OS through a patchwork of kernal modifications.
Excellent investigation. Samsung designed the M3 core with wider decode, lower frequency, and potentially better integration with its own hardware yet still could not match the performance and efficiency of ARM A75. Why did they not do their internal testing of A75 before deciding to go with M3?
My strong guess is that Samsung has the same idea that Apple has for its future MacBooks - Intel outside. In mobile systems (phones), well executed hardware and software designs of wide and deep cores (Apple: yes, Samsung: not really) can offer great peak performance but have to throttle heavily due to thermal and power constraints. Same core designs in a laptop with good heat management and a much larger battery, those restraints are loosened significantly. Samsung's attempt failed mostly due to their typical report card - hardware: A- or B+, software: D or F. That being said, Apple's drive towards in-house chips plus the ongoing Windows 10 on Snapdragon 835/845 initiative by Qualcomm and MS doesn't augur well for Intel's almost-monopoly in the ultraportable laptop and 2-in-1 market, especially once Samsung & Co. get their act together.
I used to share your optimism on ARM muscling into Intel territory but now I'm not so sure. Snapdragon Win10 PCs have been announced but nobody's buying them - this really does smell like WinRT and Surface RT all over again. Microsoft, Qualcomm and PC OEMs are adopting a wait and see approach when they should be going all-out on ARM.
That's Microsoft's fault. Instead of focusing on a purely mobile architecture and the "enhancing" it for the desktop after it failed, they should have worked on the desktop experience from the start to be closer to mobile in terms of efficiency and hardware acceleration without sacrificing features, information density and responsiveness, relieving it of its dependence on general cpu compute.
Instead of going platform agnostic for a native desktop, they went with a *form factor* agnostic approach that focused on the lowest common denominator (mobile) that didn't benefit either.
If the software is right, and workloads are properly accelerated/deligated, then the CPU of choice wouldn't be nearly as critical to get the optimum experience and performance of the device and the purpose it was is designed to serve.
Indeed, you may be on to something there. Microsoft added tablet and mobile features to Windows 10 in a halfhearted way, like how the touch keyboard on Windows tablets is a laggy and inaccurate mess compared to the iPad keyboard. I bring along a folding Bluetooth keyboard with my Windows tablet because using the virtual keyboard is such a painful experience.
That said, there's probably a ton of Win32 baggage that relies on x86 to run so it could take years for Microsoft to refactor everything.
Responsiveness and accuracy of your on-screen keyboard are largely determined by hardware, not MS. E.g. the keyboard on my wifes low-end Lumia 535 is pretty bad, whereas the same software on my Lumia 950 performs great.
The onscreen keyboard on my cube i7 book works as smoothly and quickly as any ipad or android tab. The Core m3 keeps it ticking along nicely. I use it purely as a tablet, even while working on it occasionally, with the type attachment collecting dust in my drawer.
I think poor quality storage and CPU are the reasons for the bad experience.
Yet to finish reading but for clarity. the memory controller is at half speed only in the Samsung power saving mode and not with your custom configs?
PC Mark clearly does not depend on core perf much and maybe that's what's confusing. It's seen as mostly a CPU benchmark with GPU in photo editing. And you are kidding about a robotic arm but you only need a moving fingertip with a sensor for the most basic testing and that's easily doable in days. I know Sparkfun has a 40$ IR sensor for robotic fingers but you can go with other sensors too. There are dedicated robots but , to pivot a bit to a slightly different topic, you could test app load times with a high speed phone camera and we would be happy. Wish you guys did that, test load times better than the folks on Youtube - so the bar is very low right now.
Is it possible to disable 1-2 big cores and run config 3, maybe with more aggressive settings? And why blame the M3 core for web perf and not everything else that might add latency?
Anyway, this core feels like it was aimed at 7nm and it has potential if they improve on it.
I still say old logic is much more superior (HMP, interactive & core_ctl). Windows load tracking isn't something new it just whose not used much in HMP. EAS is just a big miss. Also seems that 2GHz remains sane limit for sustainable leaking.
Makes me wonder if this Exynos chip would be better as a midranger like the SD650, with the A55s doing most of the work and the M3 cores running in short bursts. A cut-down M3 with a lower frequency limit could work better; at 2 GHz+ the M3 seems to guzzle power for not much of a performance increase.
What kind of magic sauce did Qualcomm use in the SD845 S9 to get 11 hours of battery life vs 7 on the Exynos S9? I'm getting 11 hours on a Mi Max with a much larger battery so any further tweaks would be welcome.
You are getting 11h SOT in real life usage on an old phone with worn out or at least started to get worn out battery. S9 US version ain't getting 11h of SOT, it's getting 7, 8 at best. Equnos is getting 4 to 6. Check out on XDA. The S710 will be cut down version of S845 & 2GHz is limit for 10 nm second Samsung FinFET, check the graph S845 is even more leaking than Equnos when it's crossed.
For this part, Spock should‘ve used a tunnel bore instead of a mere chainsaw, because a chainsaw doesn‘t seem to accomodate the right amount of „POWER!“ that one would require in order to deal with Samsung‘s crapload of shady implementation. SCNR
Software can still play a part in the web workloads, as the Javascript and browser engines are probably better tuned for the ARM cores. Of course tt could also be the M3 core struggling to extract enough ILP out of the branchy and cache unfriendly JS code, ARM has certainly tuned their cores for this over the years.
Outstanding work, and much appreciated. Thank you, Andrei!
I fully recognize that one only has so much time to commit to writing pieces like this, and that it requires a fair amount of personal interest in the subject matter at hand to do it right. However, it would be great to see Anandtech do a deep dive into the iPhone battery issue. Obviously Apple is intentionally opaque regarding a fair amount of the iPhone and iOS internals, which would make it somewhat more challenging. But I have yet to see anyone publish anything that even quantifies the basics of the reduced performance modes introduced by the various software updates. I can't imagine it would be too hard to figure out what the maximum clocks are with a new battery vs. what appear to be four distinct lower performance modes, or to determine what metrics are used to trigger those modes. Maybe the folks at the kiosk that does aftermarket battery replacements at the local mall would be willing to set aside a box of pulled batteries for testing? Such an article would arrive late as far as the media cycle is concerned, but this is an issue that continues to affect hundreds of millions of people worldwide.
A few years back, Brian Klug used to catch a lot of flack for being biased, before he (and Anand) left to join Apple. And maybe you'll end up at Samsung in the future, like Kristian. But while I think your tone was quite even handed and appropriate in this article, you seemed much more inclined to denounce Apple as deserving of a class-action lawsuit before doing any in-depth testing in regards to their recent power / performance issues. In your opinion, what is the appropriate response from Samsung in this situation? I mean, obviously pushing out a software update that improves performance would be a good start, but should owners of Exynos 9810 variants of the Galaxy S9 sue for damages and extract their pound of flesh?
@repoman27: "But while I think your tone was quite even handed and appropriate in this article, you seemed much more inclined to denounce Apple as deserving of a class-action lawsuit before doing any in-depth testing in regards to their recent power / performance issues."
I think the difference in tone reflects the difference in situation. In Samsung's situation, the product came to market with a set level of performance and battery life. Very early in the lifecycle of the product, It was discovered that better software would improve their situation to a more or less extent on one or both fronts. In Apple's situation, the product came to market with a set level of performance and battery life. After a certain time period and as a direct result of a software update, performance suddenly and inexplicably (until it was investigated) dropped for older phones (more cycles on the battery).
Samsung's situation left their solution looking immature at best and their engineers looking incompetent at worst. In either case, their misstep appears to be unintentional and only serves to hurt sales of a new product. Apple's situation left them looking misguided at best and abusive(?) at worst. In either case, their actions appear to be intentional (well meaning or not). The actions only affect phones later in their life cycle and, whether as an unintentional side effect of their actions or as the primary goal of their actions, Apple stands to gain by virtue of prompting upgrades. Some people tend to believe the upgrades were the primary goal due to Apple's locked in ecosystem design and marketing strategy, but it could just as easily have been an (un)fortunate side effect of them trying to mitigate another potentially serious blemish to their user experience. Perhaps a case of the cure being worse than the disease. Like you, I wouldn't mind a more in depth look at the situation that could help clarify this.
In any case, I think the difference in tone comes down to the clear lack of intention on Samsung's part (they don't benefit from this) and potentially well meaning but certainly intentional actions on Apple's part.
Excellent article Andrei. One of the best on here in a long time.
Glad i got a import Snapdragon S9+, because it seems no software update will ever properly fix the poor battery life and *actual daily usage* performance of the Exynos version. Absolutely pathetic how poorly Samsung handled all this though. So many poor decisions. Honestly the people responsible need to be fired. People shouldn't have to go out of their way to buy imports.
This article will probably be used as a chainsaw by Samsung top management to get rid of everyone who screwed up the Exynos S9. It's sad how seemingly chasing synthetic benchmarks led to bad decisions from chip design onwards.
AFAIK they are comparing the smaller S9 Exynos (which has 3000mAh battery) vs. S9+ S845 US version (3500mAh). So the battery life obviously will be better with the bigger bro. But, anyway, this doesn't make the exynos version any better.
Very interesting article indeed. I agree the M3 core looks a bit oversized for a smartphone - but a tweaked version of it on 7nm might fare a lot better. For this generation, it definitely wasn't worth the effort over standard A75. And, by the looks of it, recognizing the power issues, samsung made things worse with an inadequate scheduler. I wonder if actually the SD845 could get better battery life at very little performance cost by disabling the highest frequency state - that sure looks inefficient as hell. (Albeit that could only widen the battery life gap between the two S9 versions...). The discrepancy between web tests and microbenchmarks (and spec) in efficiency is also quite interesting - while I'd agree it might be more difficult to extract good IPC of these web tests (putting wide cores at a disadvantage) apple seems to have successfully done so, though I'm ignoring if through software or hardware means.
I'd be REALLY pissed if I had an Exynos S9+. Seems to me like that would feel like the S6 in terms of battery life, that phone was terrible. Incidentally, I had the S6, and I hated the battery life even as a light user. And IIRC that phone had an Exynos in it even in the US version. Hmmmmmmm.
As an owner I wouldn't say I'm pissed as the battery life overall is pretty decent (getting around 6h+ of SOT which is similar to what most SD845 users are getting). You shouldn't base battery life on just web browsing as people tend to do a lot of other stuff on their phones besides that (check other Youtube comparisons and you'll see that they are on par pretty much). Definitely sucks that Samsung hasn't optimized the performance, and they can't really change the clockspeed now after they've released it.
WOW. Excellent work, great results, and awesome writeup! Bravo Andrei. I love all the detail about what works and what doesn't.
I'd be interested in seeing even rough numbers for what performance and battery looked like when using WALT and when you tried using 8ms half-life with PELT. Like: "Using WALT only gave ~5% performance improvement over Config 2 at the cost of cutting battery life down to ~3 hours..." of course using your figures instead of the ones I made up.
If I remember correctly 6h with WALT and 6.5h with 8ms PELT at 1794MHz, performance was great but just murder on the battery. Obviously something wasn't right with the WALT config so that's why I didn't post the result as it wasn't representative.
I love these Anandtech articles that read like a school textbook. They're challenging, and I genuinely feel like I better understand technology by the end of them. Well done.
Love that you replaced the confusing double barchart thing with a scatterplot of power curves. Much much clearer. I would suggest showing the same data as a energy/performance plot - this way one can see from the same plot if the performane to power trade-off is worth it.
A single tech writer with some smarts is able to do what a tech conglomerate with multiple billions of dollars can't. That is absolutely fucking pathetic on Samsung's part; I used to think they were just incompetent, but to be able to design and fab their own CPUs, yet not provide appropriate working drivers for that CPU? Words quite honestly fail me.
An alternate view is that they have great teams for chip design (though perhaps not as much this one), semi-conductor fabrication, and phone hardware integration., but not particularly good (being kind) teams for their software/firmware development.
And the verdict is ... if you want the best Android experience, either pick up a Pixel 2 XL or wait a few months for the P3. If you want the highest performing smartphone, tolerate Apple and the never-quite-works-properly Siri and their gimmicky fashion-first business model. I expected better out of Samsung ... The Note9 is supposed to be quite an evolution ... hopefully enough that I can be convinced to trade in my Note5. As for everyday use, my Pixel 1 XL is still running brilliantly and there simply isn't enough of a performance delta to risk a third-party Android build that may or may not receive OS updates 24 months from now.
Magnificent article, thanks for wrapping everything up in such a detailed manner.
The more I‘m looking at it, the more I‘m asking myself: „WTF were they smoking when they came up with the ingenieus idea of putting all Meerkat cores into a single clock- & voltage-domain?“. And tbh, even today I can‘t come up with a proper explanation other than „Time to market & saving xtors for the sake of not blowing up the chips budget“.
Another majestic one. As expected from the initial experiment, though I was under the remark that the custom tuning would at least fix up the issue a bit, it did but the performance/efficiency is just bad. Very bad for such a high profile flagship device.
A pity that this level of SoC should need 4000Mah capacity. It would have been much better for tuning and custom software enthusiasts and considering the constant high performance the voltage/power scaling is fine so vs the SD845 this SoC won't kill the total endurance of the battery fast. That's the only good thing of 9810 custom tuned/stock vs 845. But the XDA developer community and the top devs will get much more resources to work with with ease on the other phones with 845 platform like Pixel 3 or OP6 (Unfortunate Notch B$) don't have hopes on HTC as their 10 was a fail even with optimization the efficiency was off small battery and WiFi issues. OP3(T) and 5T seem the best choice for now.
Much thanks Andrei for this, superb analysis and thanks for letting us know about the EAS too, that was total gold. I needed that. I guess I will pass this. I don't want another 6hr SOT, OP3 barely has 3-5 and 6 is super bad with custom tuning that too after a gap of 1Yr. Will wait for the next Exynos if it has a headphone jack / OP7.
And please keep this work coming going forth, don't leave us in the dark, Ofc it would be great if you get a good opportunity but we need you.
On my OnePlus 5, I go from 11 hours SOT (with the standard setup) to about 7 (using EAS+schedutil) and performance in most benchmarks regresses - it clearly needs a lot of work.
If you see part 1, EX developer Flar, has an Elemental X custom kernel with the traditional hotplugging (which Andrei removed for EAS) from Samsung but tweaked.
I used to rely on xda dev for any performance thinkering and trigger the knox (on samsung) but now anand had them too. Wow. Hopefully you will cook some custom rom there andrei since i saw your username too =)
Spelllling errrrors: Therefor my subjective experience of the phone being so fast... Aught to be: Therefore my subjective experience of the phone being so fast...
If you know an individual task’s load, then you can make batter scheduling decisions on which CPU cores to place it. Aught to be: If you know an individual task’s load, then you can make better scheduling decisions on which CPU cores to place it.
Hey, i'm going to switch from iphone 6s to samsung s9, but i live in europe, and i heard very bad about the european s9.Is there any possibility that samsung would fix this thing?
Andrei, thanks a lot for your great review. Would you mind revisiting it with the new update of the B-Train (BRE5 and upwards)? It seems snappier afterwards.
I feel cheated... I bought the Galaxy S9+ 3 weeks ago thinking I would get the Snapdragon CPU. That's what it was advertised as everywhere. I didn't even know there was a different version. And now it looks like I basically got an intentionally defective version for the full price.
Why all posts dont talk about radio performance. Testing both chipset under good and poor coverage to see how 3G and LTE speed is really important test to consider
Thank you very very much for sourcing this.. I merged most of your tweaks in my kernel, along with a different approach to overclocking. The results of my NOTE 9 are even better than those posted in the article, maybe its the improved cooling... https://photos.app.goo.gl/TJjCFY5BVKkYKMi89
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
76 Comments
Back to Article
zepi - Friday, April 20, 2018 - link
Andrei, how come you are still with Anandtech and not working for one of the big smartphone manufacturers, Qualcomm or maybe ARM?zepi - Friday, April 20, 2018 - link
Excellent piece, like always btw.tipoo - Friday, April 20, 2018 - link
(smartphone manufacturers, please don't take him too! :P )Manch - Friday, April 20, 2018 - link
AndreiTechMorawka - Friday, April 20, 2018 - link
We need good writers, he's fine where he's at. He can always do consulting like Anand did. This is a great investigative piece and i'm glad Anandtech has linux guru's who can make their own OS through a patchwork of kernal modifications.RaduR - Tuesday, April 24, 2018 - link
Andrei was working for ImgTech if i'm not mistaking. Unfortunately since Apple move ImgTech without MIPS I dont't think will ever come out with a SOC.Real one not just on paper.
So unfortunate that MIPS+PowerVR was never to become a successful competition to ARM.
juicytuna - Friday, April 20, 2018 - link
Great stuff. Reads like a job application to the S.LSI BSP team.fishjunk - Friday, April 20, 2018 - link
Excellent investigation. Samsung designed the M3 core with wider decode, lower frequency, and potentially better integration with its own hardware yet still could not match the performance and efficiency of ARM A75. Why did they not do their internal testing of A75 before deciding to go with M3?eastcoast_pete - Friday, April 20, 2018 - link
My strong guess is that Samsung has the same idea that Apple has for its future MacBooks - Intel outside. In mobile systems (phones), well executed hardware and software designs of wide and deep cores (Apple: yes, Samsung: not really) can offer great peak performance but have to throttle heavily due to thermal and power constraints. Same core designs in a laptop with good heat management and a much larger battery, those restraints are loosened significantly. Samsung's attempt failed mostly due to their typical report card - hardware: A- or B+, software: D or F.That being said, Apple's drive towards in-house chips plus the ongoing Windows 10 on Snapdragon 835/845 initiative by Qualcomm and MS doesn't augur well for Intel's almost-monopoly in the ultraportable laptop and 2-in-1 market, especially once Samsung & Co. get their act together.
serendip - Friday, April 20, 2018 - link
I used to share your optimism on ARM muscling into Intel territory but now I'm not so sure. Snapdragon Win10 PCs have been announced but nobody's buying them - this really does smell like WinRT and Surface RT all over again. Microsoft, Qualcomm and PC OEMs are adopting a wait and see approach when they should be going all-out on ARM.lilmoe - Friday, April 20, 2018 - link
That's Microsoft's fault. Instead of focusing on a purely mobile architecture and the "enhancing" it for the desktop after it failed, they should have worked on the desktop experience from the start to be closer to mobile in terms of efficiency and hardware acceleration without sacrificing features, information density and responsiveness, relieving it of its dependence on general cpu compute.Instead of going platform agnostic for a native desktop, they went with a *form factor* agnostic approach that focused on the lowest common denominator (mobile) that didn't benefit either.
If the software is right, and workloads are properly accelerated/deligated, then the CPU of choice wouldn't be nearly as critical to get the optimum experience and performance of the device and the purpose it was is designed to serve.
serendip - Saturday, April 21, 2018 - link
Indeed, you may be on to something there. Microsoft added tablet and mobile features to Windows 10 in a halfhearted way, like how the touch keyboard on Windows tablets is a laggy and inaccurate mess compared to the iPad keyboard. I bring along a folding Bluetooth keyboard with my Windows tablet because using the virtual keyboard is such a painful experience.That said, there's probably a ton of Win32 baggage that relies on x86 to run so it could take years for Microsoft to refactor everything.
MrSpadge - Saturday, April 21, 2018 - link
Responsiveness and accuracy of your on-screen keyboard are largely determined by hardware, not MS. E.g. the keyboard on my wifes low-end Lumia 535 is pretty bad, whereas the same software on my Lumia 950 performs great.LiverpoolFC5903 - Monday, April 23, 2018 - link
The onscreen keyboard on my cube i7 book works as smoothly and quickly as any ipad or android tab. The Core m3 keeps it ticking along nicely. I use it purely as a tablet, even while working on it occasionally, with the type attachment collecting dust in my drawer.I think poor quality storage and CPU are the reasons for the bad experience.
Alexvrb - Saturday, April 21, 2018 - link
"relieving of its dependence on general CPU compute"Don't smoke crack and post.
tuxRoller - Saturday, April 21, 2018 - link
Otoh, arm servers are looking better all the time.Hurr Durr - Saturday, April 21, 2018 - link
It`s a pipedream, like loonix on desktop. Pretty sure it`s the same people anyway.tuxRoller - Saturday, April 21, 2018 - link
Chromeos is a thing.ZolaIII - Sunday, April 22, 2018 - link
Reed you unlettered kite pipe:https://blog.cloudflare.com/neon-is-the-new-black/...
tuxRoller - Monday, April 23, 2018 - link
That was a great post, and it makes SVE look all the more interesting in this space.jjj - Friday, April 20, 2018 - link
Yet to finish reading but for clarity. the memory controller is at half speed only in the Samsung power saving mode and not with your custom configs?PC Mark clearly does not depend on core perf much and maybe that's what's confusing. It's seen as mostly a CPU benchmark with GPU in photo editing.
And you are kidding about a robotic arm but you only need a moving fingertip with a sensor for the most basic testing and that's easily doable in days. I know Sparkfun has a 40$ IR sensor for robotic fingers but you can go with other sensors too. There are dedicated robots but , to pivot a bit to a slightly different topic, you could test app load times with a high speed phone camera and we would be happy. Wish you guys did that, test load times better than the folks on Youtube - so the bar is very low right now.
Andrei Frumusanu - Friday, April 20, 2018 - link
The memory controller isn't limited in the custom config.jjj - Friday, April 20, 2018 - link
Is it possible to disable 1-2 big cores and run config 3, maybe with more aggressive settings?And why blame the M3 core for web perf and not everything else that might add latency?
Anyway, this core feels like it was aimed at 7nm and it has potential if they improve on it.
ZolaIII - Friday, April 20, 2018 - link
I still say old logic is much more superior (HMP, interactive & core_ctl). Windows load tracking isn't something new it just whose not used much in HMP. EAS is just a big miss. Also seems that 2GHz remains sane limit for sustainable leaking.serendip - Friday, April 20, 2018 - link
Makes me wonder if this Exynos chip would be better as a midranger like the SD650, with the A55s doing most of the work and the M3 cores running in short bursts. A cut-down M3 with a lower frequency limit could work better; at 2 GHz+ the M3 seems to guzzle power for not much of a performance increase.What kind of magic sauce did Qualcomm use in the SD845 S9 to get 11 hours of battery life vs 7 on the Exynos S9? I'm getting 11 hours on a Mi Max with a much larger battery so any further tweaks would be welcome.
ZolaIII - Saturday, April 21, 2018 - link
You are getting 11h SOT in real life usage on an old phone with worn out or at least started to get worn out battery. S9 US version ain't getting 11h of SOT, it's getting 7, 8 at best. Equnos is getting 4 to 6. Check out on XDA. The S710 will be cut down version of S845 & 2GHz is limit for 10 nm second Samsung FinFET, check the graph S845 is even more leaking than Equnos when it's crossed.dave_the_nerd - Friday, April 20, 2018 - link
What's with the splash image of Spock messing with a chainsaw?Ryan Smith - Friday, April 20, 2018 - link
Part 1 was Spock with a screwdriver. This round of testing was much more extreme, so we had to use a more powerful tool.SirCanealot - Friday, April 20, 2018 - link
I just noticed the tool! Oh my, I just cracked up! Thanks for the article and the laugh!! :DPs, Andrei, I love you. (ie, thanks)
Hifihedgehog - Friday, April 20, 2018 - link
Quite the upgrade from stone knives and bear skins!N Zaljov - Sunday, April 22, 2018 - link
For this part, Spock should‘ve used a tunnel bore instead of a mere chainsaw, because a chainsaw doesn‘t seem to accomodate the right amount of „POWER!“ that one would require in order to deal with Samsung‘s crapload of shady implementation. SCNRdjayjp - Friday, April 20, 2018 - link
Wow, amazing work. The definition of above and beyond. I would submit this data to Samsung.StormyParis - Friday, April 20, 2018 - link
Fascinating. Thank you !mad_one - Friday, April 20, 2018 - link
Lovely article!Software can still play a part in the web workloads, as the Javascript and browser engines are probably better tuned for the ARM cores. Of course tt could also be the M3 core struggling to extract enough ILP out of the branchy and cache unfriendly JS code, ARM has certainly tuned their cores for this over the years.
repoman27 - Friday, April 20, 2018 - link
Outstanding work, and much appreciated. Thank you, Andrei!I fully recognize that one only has so much time to commit to writing pieces like this, and that it requires a fair amount of personal interest in the subject matter at hand to do it right. However, it would be great to see Anandtech do a deep dive into the iPhone battery issue. Obviously Apple is intentionally opaque regarding a fair amount of the iPhone and iOS internals, which would make it somewhat more challenging. But I have yet to see anyone publish anything that even quantifies the basics of the reduced performance modes introduced by the various software updates. I can't imagine it would be too hard to figure out what the maximum clocks are with a new battery vs. what appear to be four distinct lower performance modes, or to determine what metrics are used to trigger those modes. Maybe the folks at the kiosk that does aftermarket battery replacements at the local mall would be willing to set aside a box of pulled batteries for testing? Such an article would arrive late as far as the media cycle is concerned, but this is an issue that continues to affect hundreds of millions of people worldwide.
A few years back, Brian Klug used to catch a lot of flack for being biased, before he (and Anand) left to join Apple. And maybe you'll end up at Samsung in the future, like Kristian. But while I think your tone was quite even handed and appropriate in this article, you seemed much more inclined to denounce Apple as deserving of a class-action lawsuit before doing any in-depth testing in regards to their recent power / performance issues. In your opinion, what is the appropriate response from Samsung in this situation? I mean, obviously pushing out a software update that improves performance would be a good start, but should owners of Exynos 9810 variants of the Galaxy S9 sue for damages and extract their pound of flesh?
BurntMyBacon - Monday, April 23, 2018 - link
@repoman27: "But while I think your tone was quite even handed and appropriate in this article, you seemed much more inclined to denounce Apple as deserving of a class-action lawsuit before doing any in-depth testing in regards to their recent power / performance issues."I think the difference in tone reflects the difference in situation. In Samsung's situation, the product came to market with a set level of performance and battery life. Very early in the lifecycle of the product, It was discovered that better software would improve their situation to a more or less extent on one or both fronts. In Apple's situation, the product came to market with a set level of performance and battery life. After a certain time period and as a direct result of a software update, performance suddenly and inexplicably (until it was investigated) dropped for older phones (more cycles on the battery).
Samsung's situation left their solution looking immature at best and their engineers looking incompetent at worst. In either case, their misstep appears to be unintentional and only serves to hurt sales of a new product. Apple's situation left them looking misguided at best and abusive(?) at worst. In either case, their actions appear to be intentional (well meaning or not). The actions only affect phones later in their life cycle and, whether as an unintentional side effect of their actions or as the primary goal of their actions, Apple stands to gain by virtue of prompting upgrades. Some people tend to believe the upgrades were the primary goal due to Apple's locked in ecosystem design and marketing strategy, but it could just as easily have been an (un)fortunate side effect of them trying to mitigate another potentially serious blemish to their user experience. Perhaps a case of the cure being worse than the disease. Like you, I wouldn't mind a more in depth look at the situation that could help clarify this.
In any case, I think the difference in tone comes down to the clear lack of intention on Samsung's part (they don't benefit from this) and potentially well meaning but certainly intentional actions on Apple's part.
B3an - Friday, April 20, 2018 - link
Excellent article Andrei. One of the best on here in a long time.Glad i got a import Snapdragon S9+, because it seems no software update will ever properly fix the poor battery life and *actual daily usage* performance of the Exynos version. Absolutely pathetic how poorly Samsung handled all this though. So many poor decisions. Honestly the people responsible need to be fired. People shouldn't have to go out of their way to buy imports.
serendip - Friday, April 20, 2018 - link
This article will probably be used as a chainsaw by Samsung top management to get rid of everyone who screwed up the Exynos S9. It's sad how seemingly chasing synthetic benchmarks led to bad decisions from chip design onwards.madurko - Saturday, April 21, 2018 - link
AFAIK they are comparing the smaller S9 Exynos (which has 3000mAh battery) vs. S9+ S845 US version (3500mAh). So the battery life obviously will be better with the bigger bro. But, anyway, this doesn't make the exynos version any better.mczak - Friday, April 20, 2018 - link
Very interesting article indeed.I agree the M3 core looks a bit oversized for a smartphone - but a tweaked version of it on 7nm might fare a lot better. For this generation, it definitely wasn't worth the effort over standard A75.
And, by the looks of it, recognizing the power issues, samsung made things worse with an inadequate scheduler.
I wonder if actually the SD845 could get better battery life at very little performance cost by disabling the highest frequency state - that sure looks inefficient as hell. (Albeit that could only widen the battery life gap between the two S9 versions...).
The discrepancy between web tests and microbenchmarks (and spec) in efficiency is also quite interesting - while I'd agree it might be more difficult to extract good IPC of these web tests (putting wide cores at a disadvantage) apple seems to have successfully done so, though I'm ignoring if through software or hardware means.
hansmuff - Friday, April 20, 2018 - link
Fantastic article, thank you so much!I'd be REALLY pissed if I had an Exynos S9+. Seems to me like that would feel like the S6 in terms of battery life, that phone was terrible. Incidentally, I had the S6, and I hated the battery life even as a light user. And IIRC that phone had an Exynos in it even in the US version. Hmmmmmmm.
Toss3 - Saturday, April 21, 2018 - link
As an owner I wouldn't say I'm pissed as the battery life overall is pretty decent (getting around 6h+ of SOT which is similar to what most SD845 users are getting). You shouldn't base battery life on just web browsing as people tend to do a lot of other stuff on their phones besides that (check other Youtube comparisons and you'll see that they are on par pretty much). Definitely sucks that Samsung hasn't optimized the performance, and they can't really change the clockspeed now after they've released it.lucam - Friday, April 20, 2018 - link
When iPhone X Review?MrCommunistGen - Friday, April 20, 2018 - link
WOW. Excellent work, great results, and awesome writeup! Bravo Andrei. I love all the detail about what works and what doesn't.I'd be interested in seeing even rough numbers for what performance and battery looked like when using WALT and when you tried using 8ms half-life with PELT. Like: "Using WALT only gave ~5% performance improvement over Config 2 at the cost of cutting battery life down to ~3 hours..." of course using your figures instead of the ones I made up.
Andrei Frumusanu - Friday, April 20, 2018 - link
If I remember correctly 6h with WALT and 6.5h with 8ms PELT at 1794MHz, performance was great but just murder on the battery. Obviously something wasn't right with the WALT config so that's why I didn't post the result as it wasn't representative.Dizoja86 - Friday, April 20, 2018 - link
I love these Anandtech articles that read like a school textbook. They're challenging, and I genuinely feel like I better understand technology by the end of them. Well done.jospoortvliet - Friday, April 20, 2018 - link
I can only agree. I was looking forward to this article and out is beyond expectation - fantastic work. This is why I come to this site...Lau_Tech - Friday, April 20, 2018 - link
Good job Andrei! Very interesting and unique articlelilmoe - Friday, April 20, 2018 - link
Just looking at the power curves (finally!), this is totally a laptop chip, or a hybrid are least.Dear Microsoft, use this chip for Windows on ARM.
stepz - Saturday, April 21, 2018 - link
Love that you replaced the confusing double barchart thing with a scatterplot of power curves. Much much clearer. I would suggest showing the same data as a energy/performance plot - this way one can see from the same plot if the performane to power trade-off is worth it.Andrei Frumusanu - Saturday, April 21, 2018 - link
It's not a replacement; they serve different purposes.The_Assimilator - Saturday, April 21, 2018 - link
A single tech writer with some smarts is able to do what a tech conglomerate with multiple billions of dollars can't. That is absolutely fucking pathetic on Samsung's part; I used to think they were just incompetent, but to be able to design and fab their own CPUs, yet not provide appropriate working drivers for that CPU? Words quite honestly fail me.BurntMyBacon - Monday, April 23, 2018 - link
An alternate view is that they have great teams for chip design (though perhaps not as much this one), semi-conductor fabrication, and phone hardware integration., but not particularly good (being kind) teams for their software/firmware development.johnnycanadian - Saturday, April 21, 2018 - link
And the verdict is ... if you want the best Android experience, either pick up a Pixel 2 XL or wait a few months for the P3. If you want the highest performing smartphone, tolerate Apple and the never-quite-works-properly Siri and their gimmicky fashion-first business model. I expected better out of Samsung ... The Note9 is supposed to be quite an evolution ... hopefully enough that I can be convinced to trade in my Note5. As for everyday use, my Pixel 1 XL is still running brilliantly and there simply isn't enough of a performance delta to risk a third-party Android build that may or may not receive OS updates 24 months from now.santz - Saturday, April 21, 2018 - link
thank you for the excellent writeup. I just hope the Note 9 will not have exynos for their international version.Seattletech - Sunday, April 22, 2018 - link
Sign me up for2 M3
2 A75
4 A55
N Zaljov - Sunday, April 22, 2018 - link
Magnificent article, thanks for wrapping everything up in such a detailed manner.The more I‘m looking at it, the more I‘m asking myself: „WTF were they smoking when they came up with the ingenieus idea of putting all Meerkat cores into a single clock- & voltage-domain?“. And tbh, even today I can‘t come up with a proper explanation other than „Time to market & saving xtors for the sake of not blowing up the chips budget“.
Quantumz0d - Sunday, April 22, 2018 - link
Another majestic one. As expected from the initial experiment, though I was under the remark that the custom tuning would at least fix up the issue a bit, it did but the performance/efficiency is just bad. Very bad for such a high profile flagship device.A pity that this level of SoC should need 4000Mah capacity. It would have been much better for tuning and custom software enthusiasts and considering the constant high performance the voltage/power scaling is fine so vs the SD845 this SoC won't kill the total endurance of the battery fast. That's the only good thing of 9810 custom tuned/stock vs 845. But the XDA developer community and the top devs will get much more resources to work with with ease on the other phones with 845 platform like Pixel 3 or OP6 (Unfortunate Notch B$) don't have hopes on HTC as their 10 was a fail even with optimization the efficiency was off small battery and WiFi issues. OP3(T) and 5T seem the best choice for now.
Much thanks Andrei for this, superb analysis and thanks for letting us know about the EAS too, that was total gold. I needed that. I guess I will pass this. I don't want another 6hr SOT, OP3 barely has 3-5 and 6 is super bad with custom tuning that too after a gap of 1Yr. Will wait for the next Exynos if it has a headphone jack / OP7.
And please keep this work coming going forth, don't leave us in the dark, Ofc it would be great if you get a good opportunity but we need you.
Thank you again Andrei.
Azurael - Sunday, April 22, 2018 - link
On my OnePlus 5, I go from 11 hours SOT (with the standard setup) to about 7 (using EAS+schedutil) and performance in most benchmarks regresses - it clearly needs a lot of work.Spoelie - Sunday, April 22, 2018 - link
Doesn't seem like someone on XDA is picking up this config yet - would love to try it on my S9.Quantumz0d - Sunday, April 22, 2018 - link
If you see part 1, EX developer Flar, has an Elemental X custom kernel with the traditional hotplugging (which Andrei removed for EAS) from Samsung but tweaked.flar2 - Sunday, April 22, 2018 - link
I'm testing Andrei's changesbairlangga - Monday, April 23, 2018 - link
I used to rely on xda dev for any performance thinkering and trigger the knox (on samsung) but now anand had them too. Wow. Hopefully you will cook some custom rom there andrei since i saw your username too =)krumme - Monday, April 23, 2018 - link
A historical milestone and future benchmark in tech journalism.aryonoco - Tuesday, April 24, 2018 - link
I'm really worried if Andrei continues this, he'll go the same way Brian and Anand went.ballsystemlord - Wednesday, April 25, 2018 - link
Spelllling errrrors:Therefor my subjective experience of the phone being so fast...
Aught to be:
Therefore my subjective experience of the phone being so fast...
If you know an individual task’s load, then you can make batter scheduling decisions on which CPU cores to place it.
Aught to be:
If you know an individual task’s load, then you can make better scheduling decisions on which CPU cores to place it.
Rοb - Thursday, April 26, 2018 - link
Your Samsung Galaxy S9+ (SD845) scores are too low across the board, except for "Video Editing" which scored slightly higher than my phone.Did you charge your phone, put in 'Peformance Mode' (HD+), AND remove the case before running the benchmarks?
Cattepie - Wednesday, May 2, 2018 - link
Hey, i'm going to switch from iphone 6s to samsung s9, but i live in europe, and i heard very bad about the european s9.Is there any possibility that samsung would fix this thing?RagnarAntonisen - Sunday, May 20, 2018 - link
Seems like Exynos is quite a way behind Snapdragon which is quite a way behind Apple.corwin_amber - Thursday, May 24, 2018 - link
Andrei, thanks a lot for your great review.Would you mind revisiting it with the new update of the B-Train (BRE5 and upwards)? It seems snappier afterwards.
krzaku - Tuesday, July 31, 2018 - link
I feel cheated... I bought the Galaxy S9+ 3 weeks ago thinking I would get the Snapdragon CPU. That's what it was advertised as everywhere. I didn't even know there was a different version. And now it looks like I basically got an intentionally defective version for the full price.zhongguoyu - Saturday, August 4, 2018 - link
真希望作者可以放出内核 而不是源代码 （源代码这东西对于普通人而言是没有价值的(ー_ー)!!）Mdad - Tuesday, September 25, 2018 - link
Why all posts dont talk about radio performance. Testing both chipset under good and poor coverage to see how 3G and LTE speed is really important test to considerhanhan - Wednesday, November 14, 2018 - link
Please unleash the gpu power, what if you overclocking the mali g72 to 1.100Mhz per core?xxmustafacooTR - Sunday, August 25, 2019 - link
I know exynos9810 maks supports 600mhz gpu but I am not sure :)gmotux - Friday, January 11, 2019 - link
Thank you very very much for sourcing this.. I merged most of your tweaks in my kernel, along with a different approach to overclocking.The results of my NOTE 9 are even better than those posted in the article, maybe its the improved cooling...
https://photos.app.goo.gl/TJjCFY5BVKkYKMi89