Samsung Exynos 9810

Do you load webpages in endless loops with your device or do software video encoding? Peak CPU performance is important because CPU workloads are transactional and bursty. I mean I'm not even going into this argument as I specifically changed the way the GPU performance is represented precisely because of these concerns. You're utterly overreacting here and using strawman arguments.

I would love nothing more if you benchmarked use cases that typical cell phone (Galaxy S9) users would encounter. I agree software video encoding is not a typical use case :p, but I do browse web pages, play games, and use apps in sessions longer than 15 seconds! :D Can you tell me how various android phones compare in situations that involve *real applications*? All I could find in your article are a bunch of SPEC, PCMark, and micro-javascript benchmarks.

The rest of the review has the 9810 with the lowest bars all over the place. You're telling me because I dared to test SPEC at peak that this invalidates the whole rest of the article. Again there's reasonable arguments and there's unreasonable ones.

You dedicated a whole page on benchmarks that are completely useless to me as a potential S9 end user. A whole page!!! What does SPEC tell me about how my phone will preform in my day-to-day activities?

And why the hell shouldn't we? Getting both the peak and sustained numbers tells us more about the GPU and the way it's run.

Because we don't live in a fantasy world. Call me crazy but I just want to know how a GPU will preform when I play X, Y, and Z games in a normal situation (i.e. sans dry ice).
 
Last edited:
I would love nothing more if you benchmarked use cases that typical cell phone (Galaxy S9) users would encounter.
Boring.

All I could find in your article are a bunch of SPEC, PCMark, and micro-javascript benchmarks.
This is why Anandtech is interesting in a sea of poor tech journalism.

You dedicated a whole page on benchmarks that are completely useless to me as a potential S9 end user. A whole page!!!
Skip this page. Don't read it. The people who are interested in tech will enjoy reading it.
 
The SD845 being not quite stable (with this particular S9 phone at least) with sustained gpu loads is quite a surprise (since the power usage looks quite reasonable).
Albeit I'm wondering if it could have local temperature hotspots - the mali g72-mp18 may have terrible area efficiency, but of course on the upside this at least means power density is only half as large... (And if the Adreno 630 is actually smaller than the Adreno 540 power density would be larger too - not sure about how it compares in size due to it seemingly being only a 2-core design vs. the previous 4-core one, for that matter are you sure it really doubled alu units and not just alu units per core, since the results do not indicate anywhere close to double the performance in the alu tests I've seen.)

FWIW I think the SPEC comparisons are very interesting.
Although I'd have liked to see voltage / frequency curves for the SD 845 too, I'm kinda wondering if it also sacrifices energy efficiency completely at the highest frequencies.
Surely for the M3 cores the frequencies above ~2.3Ghz are inefficient enough that they should not have been used in the first place (well, in a smartphone). But of course if it would only reach 2.3Ghz then the core would not really be all that much faster than the A75 at its max frequency, making all the spending of transistors and area quite pointless.
If DVFS needs half a second to reach the max frequency that indeed could explain a lot of the trouble with the SoC.
Kind of interesting that everybody was like "the US gets a gimped version of the S9" whereas it looks like in fact you get a much better deal there :). Unless Samsung can improve the DVFS issues - I tend to believe though battery life would still be significantly worse.
 
Last edited:
Back
Top