Just noticed after doing benchmark tests that the results are showing all 8 cores clocking in at max of 1.8Ghz (which should be the co processors speed). The main should be 2.3Ghz according to Huawei's spec on the phone. Battery Management setting is set to Performance.
Just wondering if anyone else noticed this or has input?
yeah ive been seeing this on pretty much all apps thus far. my guess is that its an issue with reading the correct clocks on the A72 cores, only the weaker A53 cores seem to be registered and monitored. as long as your scores are fine i wouldnt worry too much about it cpu-wise, the kirin 950 is on par with SD820 and exynos 8890
Sent from my Huawei Mate 8 NXT-AL10 using Tapatalk
jbmc83 said:
yeah ive been seeing this on pretty much all apps thus far. my guess is that its an issue with reading the correct clocks on the A72 cores, only the weaker A53 cores seem to be registered and monitored. as long as your scores are fine i wouldnt worry too much about it cpu-wise, the kirin 950 is on par with SD820 and exynos 8890
Sent from my Huawei Mate 8 NXT-AL10 using Tapatalk
Click to expand...
Click to collapse
Just making sure we're not getting a device that's quite a step down in clock speed from what the manufacturer is telling us.
2.3 & 1.8 is quite a difference.
Sent from my HUAWEI NXT-L29 using XDA-Developers mobile app
mthorn79 said:
Just making sure we're not getting a device that's quite a step down in clock speed from what the manufacturer is telling us.
2.3 & 1.8 is quite a difference.
Sent from my HUAWEI NXT-L29 using XDA-Developers mobile app
Click to expand...
Click to collapse
2.3GHz is the max speed of big cores, and 1.8 is the max speed of the little cores. A lot of apps incorrectly report max speeds, core number...
ufoman said:
2.3GHz is the max speed of big cores, and 1.8 is the max speed of the little cores. A lot of apps incorrectly report max speeds, core number...
Click to expand...
Click to collapse
I can understand that maybe some outdated apps could do this, like say Quadrant that hasn't been updated since 2012, but a top rated app like Antutu also reports 1.8...even CPUZ which has been the most accurate in desktop and mobile data even shows 1.8ghz. Should we just assume these apps can't read the A72 chip?
«Mate8»
So after using these tools:
Antutu
Geekbench 3
Quadrant
Vellamo
Cpuid
Vellamo was the only one to distinguish processors#1-#4 [email protected] and processors#5-#8 at 2.3ghz
«Mate8»
mthorn79 said:
So after using these tools:
Antutu
Geekbench 3
Quadrant
Vellamo
Cpuid
Vellamo was the only one to distinguish processors#1-#4 [email protected] and processors#5-#8 at 2.3ghz
«Mate8»
Click to expand...
Click to collapse
Antutu 6 definitely shows 2,3GHz in info screens.
in Vellamo, if you look at benchmark details, you can see CPU going up to 2.,3GHz.
For most comprehensive information about the CPU, you might want to install AIDA64.
mthorn79 said:
So after using these tools:
Antutu
Geekbench 3
Quadrant
Vellamo
Cpuid
Vellamo was the only one to distinguish processors#1-#4 [email protected] and processors#5-#8 at 2.3ghz
«Mate8»
Click to expand...
Click to collapse
This was taken using Chainfire's CPUMon while running Antutu :
if the A72 cores were really only ramping up to 1.8 ghz we wouldve noticed by now according to the geekbench scores, trust me
Sent from my Huawei Mate 8 NXT-AL10 using Tapatalk
Ok so ran Antutu again and now it's showing max freq @ 2.3...I swear last week it was showing 1.8 as max...oh well, all is good. False alarm.
«Mate8»
maybe theyve updated smth in the meantime
Sent from my Huawei Mate 8 NXT-AL10 using Tapatalk
Related
See attached..... run them just now on my new transformer eeepad
www.iamdarren.com/xda/1.png
www.iamdarren.com/xda/2.png
www.iamdarren.com/xda/3.png
Lets hope this is a starting point, as soon as I can get root I will run again. With overclocking. Waiting on Paul's release from modaco
Sent from my Transformer TF101 using Tapatalk
Okay few questions.
1. Why the hell does the iPad 2 score over a 170 in linpack
2. Why is the stock transformer scoring more than stock xoom in linpack and Quadrant?
seshmaru said:
Okay few questions.
1. Why the hell does the iPad 2 score over a 170 in linpack
2. Why is the stock transformer scoring more than stock xoom in linpack and Quadrant?
Click to expand...
Click to collapse
The Ipad score seems high to me too, the only difference i can come up with the Ipad 2 is its DDR2 memory is clocked at 1033mhz
As for the Tegra 2 Processor, the DDR2 memory is clocked at 133mhz
Plus we dont really know the bus speed of the Apple A5 or Tegra 2
I know the Tegra 2 has GPUs that come in 300 or 333 Mhz which could affect the quanrant score, also the speed of the nand could affect quadrant
Who knows
What comparisons can we make between A5 and Tegra 2 processors?
One hypothesis was that the A5 may have NEON enabled, while Tegra2 doesn't.
According to Anandtech, Qualcomm's dual core offerings should have NEON support, and it impacts Linpack performance:
The Scorpion’s VFPv3 FPU is fully pipelined. As a result, floating point performance is much improved. Qualcomm also implements support for NEON, but with a wider 128-bit datapath (compared to 64-bit in the A8 and A9). As a result, Qualcomm should have much higher VFP and NEON performance than the Cortex A8 (we see a good example of this in our Linpack performance results).
Click to expand...
Click to collapse
Perhaps benchmarking Linpack on the HTC Sensation & Evo3D will shed some light on this.
Brandon
I think I will run the tests again with the class 2 micro sd taken out.... will post back later after work.
Sent from my HTC Desire using Tapatalk
I now have root, modaco custom rom, clockwork mod, and ive done anotherquadrant...
2986... setcpu just goes up to 1000, so need yo get some customer frequencies and voltages, time to build a setcpu txt file i think.......
Lets see if we can make it any faster.
www.iamdarren.com/xda/4.png
Sent from my Transformer TF101 using Tapatalk
Is it a standard kernel?
Hi all,
I can't find any benchmark about Nexus 10 on GeekBench. I need to compare it with Apple Ipad 4 CPU (already listed on geekbench website).
Someone could help me?
Where can I find a benchmark?
crashlab said:
Hi all,
I can't find any benchmark about Nexus 10 on GeekBench. I need to compare it with Apple Ipad 4 CPU (already listed on geekbench website).
Someone could help me?
Where can I find a benchmark?
Click to expand...
Click to collapse
It will smash the iPad 4. But no-one has had a chance to bench it yet.
Dual Exynos 5250 ARM-Cortex A15 CPU vs @1.7GHz
Dual Apple A6 ARM-Cortex A9 (with some architectural enhancements from A15) @ 1.3GHz
Mali-604 GPU which is 4-5x faster than previous generations Mali-400 GPU
PowerVR SGX543MP4 GPU which is great compared to Tegra and OMAP, but not compared to Mali-604
RAM is double on Nexus 10.
Geekbench on N10 is 2315 as per ZDNET. On N4, it's 1961.
I think they should improve further as Google clearly haven't optimized the drivers properly. A lot of work needs to be done from google to to optimize the devices. Expect to improve further in future.
Just got 2657 on geek bench on my nexus 10. It's rooted with a custom rom but not overclocked
I got only 1590 with geekbench on my CM11+KTmanta Nexus 10..
so how can you get 2700..?
rhyxos said:
I got only 1590 with geekbench on my CM11+KTmanta Nexus 10..
so how can you get 2700..?
Click to expand...
Click to collapse
Hi
Geekbench 3 scores shouldn't be compared with Geekbench 2 scores. AFAIK your score is related to GB3 while theirs (that are older) are related to GB2 .
~Lord
XxLordxX said:
Hi
Geekbench 3 scores shouldn't be compared with Geekbench 2 scores. AFAIK your score is related to GB3 while theirs (that are older) are related to GB2 .
~Lord
Click to expand...
Click to collapse
I thought my device had some power issues with that score (1500)
Post on Gsmarena:
"The motherboard has a Qualcomm Snapdragon S4 APQ8064-1AA chipset, which as it turns out is a downclocked Snapdragon 600 chipset – four Krait 300 cores (rather than Krait 200) at 1.5Ghz and Adreno 320. There are also four Elpida 512MB RAM chips, SK Hynix 16GB eMMC storage, a Wi-Fi a/b/g/b and BT4.0 capable Qualcomm chip and an Analogix SlimPort transmitter."
PS: Sorry if this piece of info was clarified somewhere else in the forums, I did not found it.
Yes it is a downclocked s600 which comes with krait 300 and lpddr3 ram.
Sent from my HTC One X using xda app-developers app
Yeah, this is a fairly good explanation of why these things are so battery efficient and smooth.
Anybody excited to kick this baby up to stock speed and see how she flies
Sent from my Nexus 4 using Tapatalk 4 Beta
swagstr said:
Anybody excited to kick this baby up to stock speed and see how she flies
Sent from my Nexus 4 using Tapatalk 4 Beta
Click to expand...
Click to collapse
It's not exactly a 1.9GHz SD600 (APQ8064T, APQ8064AB) downclocked, but a Krait 300 variant (APQ8064-1AA) that is intended to run at 1.5GHz, so overclocking I'd imagine would come with plenty of throttling and battery consumption.
It really is the perfect HW for current generational need, it's not like we can do pro-grade stuff that needs Core i5 power, so it's best using the stock clock which is well balanced between performance and battery life.
Good suff
Sent from Nexus 7 FHD from XDA Premium HD
BoneXDA said:
It's not exactly a 1.9GHz SD600 (APQ8064T, APQ8064AB) downclocked, but a Krait 300 variant (APQ8064-1AA) that is intended to run at 1.5GHz, so overclocking I'd imagine would come with plenty of throttling and battery consumption.
It really is the perfect HW for current generational need, it's not like we can do pro-grade stuff that needs Core i5 power, so it's best using the stock clock which is well balanced between performance and battery life.
Click to expand...
Click to collapse
I read it was a 1.7Ghz model... Must have been misinformed. She's a beast nonetheless.
Sent from my Nexus 4 using Tapatalk 4 Beta
swagstr said:
I read it was a 1.7Ghz model... Must have been misinformed. She's a beast nonetheless.
Sent from my Nexus 4 using Tapatalk 4 Beta
Click to expand...
Click to collapse
I can't imagine overclocking to 1.7 would drain anymore than a percent or 2 of total battery consumption. With the current setup it looks like it can even be done with the same voltage settings to me.
Info from my Nex7
Sent from my Nexus 7 using Xparent Green Tapatalk 2
conan1600 said:
I can't imagine overclocking to 1.7 would drain anymore than a percent or 2 of total battery consumption. With the current setup it looks like it can even be done with the same voltage settings to me.
Click to expand...
Click to collapse
I mean, I'm not the type to OC obsessively. I mean, underclock = better battery... But if we could bump up to "stock" it prob wouldn't be that bug of a difference. BUT apparently this model of the S600 is supposed to run at 1.5. I'll probably keep it there.
Sent from my Nexus 4 using Tapatalk 4 Beta
On Wikipedia it says that the APQ8064-1AA has a way faster memory speed than the S600 in the HTC One (1600MHz DDR3L 12.8 GB/sec vs. 533MHz LPDDR3 8.5 GB/sec).
I don't know if I can trust it, but...
swagstr said:
I mean, I'm not the type to OC obsessively. I mean, underclock = better battery... But if we could bump up to "stock" it prob wouldn't be that bug of a difference. BUT apparently this model of the S600 is supposed to run at 1.5. I'll probably keep it there.
Sent from my Nexus 4 using Tapatalk 4 Beta
Click to expand...
Click to collapse
Underclocking doesn't necessarily lead to a better battery life. It depends on the efficiency at the actual frequency. If you underclock, the chip needs less current but needs more time to finish the work load because it is slower. So what matters is total energy consumption: W = U * I * t (Voltage * Current * time). So to save energy, the current has to decrease more than the time increases for a given workload. Also we can see that decreasing the voltage (undervolting) also saves energy (in this case we don't loos calculation speed because the frequencies are untouched)
antutu is biased because it is favouring 64bit processors. a 32bit processor with 3 gb ram will always work better than 64bit processor with 3gb ram. this is an old debate on the desktop world when intel introduced 64bit processors. 64bit processors require double the ram just to give the same performance as as similar 32bit processors with half the ram. here are the Quadrant benchmark scores which does not bias/cheat results in favour of 64bit just to drive market sales.
23751 oneplus x
24187 oneplus one
26459 oneplus two
New perspective, interesting.
Quadrant? The 2012 app?
Enviado desde mi ONE E1005 mediante Tapatalk
any names for 32 bit benchmark apps ?
ChavitoArg said:
Quadrant?
Click to expand...
Click to collapse
yes back when benchmarks were not biased in favour of 64bit.
Mr.3bood said:
any names for 32 bit benchmark apps ?
Click to expand...
Click to collapse
Quadrant
psychem said:
antutu is biased because it is favouring 64bit processors. a 32bit processor with 3 gb ram will always work better than 64bit processor with 3gb ram. this is an old debate on the desktop world when intel introduced 64bit processors. 64bit processors require double the ram just to give the same performance as as similar 32bit processors with half the ram. here are the Quadrant benchmark scores which does not bias/cheat results in favour of 64bit just to drive market sales.
23751 oneplus x
24187 oneplus one
26459 oneplus two
Click to expand...
Click to collapse
My CPU score is saying something different... CPU score More than One plus Two ... Ooops
rockuppl said:
My CPU score is saying something different... CPU score More than One plus Two ... Ooops
Click to expand...
Click to collapse
23751 is the total score
A thread where all benchmarks are posted. Especially when comparing exynos vs SD 820
http://m.gsmarena.com/samsung_galaxy_s7_and_s7_edge_benchmarked_the_exynos_flavor-news-16794.php
Sent from my SM-G925F
Seems way of if you ask me.. should kill the z5... There is deff something wrong here!
johanbiff said:
Seems way of if you ask me.. should kill the z5... There is deff something wrong here!
Click to expand...
Click to collapse
Z5 is rendering in 1080p so there's no surprise it comes ahead in onscreen benchmarks. The mali-gpu is also not the strongest. Pretty sure the 820 will perform better.
---------- Post added at 12:55 AM ---------- Previous post was at 12:14 AM ----------
http://www.phonearena.com/news/LG-G5-shakes-hands-with-Snapdragon-820-to-shatter-AnTuTu-records-benchmark-test-scores_id78636
LG G5 seems to be scoring almost 20k higher than the exynos 8890-equipped S7. S820 looks to be the better SoC by far at this point.
https://youtu.be/qMJ2x6POZak
128k there.i guess its the exynos
Anyway not enough to surpass the iphone
http://www.antutu.com/en/view.shtml?id=8184
s3ns3lol said:
https://youtu.be/qMJ2x6POZak
128k there.i guess its the exynos
Anyway not enough to surpass the iphone
http://www.antutu.com/en/view.shtml?id=8184
Click to expand...
Click to collapse
Seriously ? Do you even know how ios and android operate ?
Sent from my SM-G925F
To sum things up: the Snapdragon 820 sports a better GPU, the Exynos 8890 sports a better CPU and a better DAC (Qualcomm DACs just haven't got the best of reputations, while the Exynos usually sport a decent Wolfson DAC), Qualcomm SOCs however usually sport a better baseband/radio than the competition.
I would say, in daily usages, the performances should be negligible, the real impact between both should be battery life related, an early preview done on the Exynos 8890 version claims a 12hrs battery life at maximum brightness on the S7 (not the S7 Edge), I guess we will see how it goes when more reviews come in.
mathieulh said:
To sum things up: the Snapdragon 820 sports a better GPU, the Exynos 8890 sports a better CPU and a better DAC (Qualcomm DACs just haven't got the best of reputations, while the Exynos usually sport a decent Wolfson DAC), Qualcomm SOCs however usually sport a better baseband/radio than the competition.
I would say, in daily usages, the performances should be negligible, the real impact between both should be battery life related, an early preview done on the Exynos 8890 version claims a 12hrs battery life at maximum brightness on the S7 (not the S7 Edge), I guess we will see how it goes when more reviews come in.
Click to expand...
Click to collapse
No Wolfson this time. It seems Samsung is using an in-house DAC.
http://www.sammobile.com/2016/02/22...ony-imx260-camera-sensor-in-house-audio-chip/
Sent from my MI 3W using Tapatalk
skivnit said:
Seriously ? Do you even know how ios and android operate ?
Sent from my SM-G925F
Click to expand...
Click to collapse
Yes.And its irrelevant.that test is cross platform
s3ns3lol said:
Yes.And its irrelevant.that test is cross platform
Click to expand...
Click to collapse
No you dont get it. Android uses multi cores to the fullest thats why multi core performance is the thing to look at, i suggest u read a piece on the subject on Anandtech
Sent from my SM-G925F
another benchmark between sd 820 and 8890:
www.anandtech.com/show/10075/early-exynos-8890-impressions
i hope the 4 core difference between the two doesn't mean worse performance on the sd 820 variant. Also if you in the EU you will -apparently- be getting the exynos variant :crying:
i also read somewhere that said that the sd 820 had 2x custom (kyro) a-72 cores and 2x custom (kyro) a-53 cores and not 4x cutom (kyro) a-72 cores, hope its not true.
s3ns3lol said:
Yes.And its irrelevant.that test is cross platform
Click to expand...
Click to collapse
Quoted from the comments of the article at AnandTech linked above:
It seems perfectly competitive in the graphics benchmarks, and comparing JavaScript benchmarks across different hardware, OS, and browser configurations is useless. To say Apple's Safari team "aggressively optimizes" for Octane and Kraken would be an understatement. Plus we're talking about simple benchmarks that can barely make any use of a second processor core, so of course they make the A9's dual-core CPU design look good next to more parallel competitors. Run something like Geekbench MT or the 3DMark physics test and watch A9 lose out to even Exynos 7420 or SD 810.
Click to expand...
Click to collapse
overlordofdoom1 said:
another benchmark between sd 820 and 8890:
www.anandtech.com/show/10075/early-exynos-8890-impressions
i hope the 4 core difference between the two doesn't mean worse performance on the sd 820 variant. Also if you in the EU you will -apparently- be getting the exynos variant :crying:
i also read somewhere that said that the sd 820 had 2x custom (kyro) a-72 cores and 2x custom (kyro) a-53 cores and not 4x cutom (kyro) a-72 cores, hope its not true.
Click to expand...
Click to collapse
The s820 has two custom cores and two lower-clocked A53s. It really won't matter that it has two fewer larger cores, as more cores leads to more heat, and more heat to more throttling. Only in benchmarks will it be noticeable.
Toss3 said:
The s820 has two custom cores and two lower-clocked A53s.
Click to expand...
Click to collapse
Where did you find that the 2 lower clocked are indeed A53s? What I have read, SD820 has 2 high clocked 2.15GHz and 2 low clocked 1.59GHz "Kryo cores"? So those 2 downclocked Kryo cores should be A72 like power not A53?
SAVVAS. said:
Where did you find that the 2 lower clocked are indeed A53s? What I have read, SD820 has 2 high clocked 2.15GHz and 2 low clocked 1.59GHz "Kryo cores"? So those 2 downclocked Kryo cores should be A72 like power not A53?
Click to expand...
Click to collapse
Hmm not sure, but pretty sure they aren't the same cores as the faster ones, as that way they could just have clocked them higher, and have them downclock instead of having them at 1.59Ghz all the time.
GFX Bench battery and throttling test of exynos variant. From 2800 frames to 1400 in 10 minutes of load. http://4pda.ru/forum/index.php?s=&showtopic=707315&view=findpost&p=47363269
If some1 find snapdragon s7 results, please post it here.
TANKRED_ENDURES said:
GFX Bench battery and throttling test of exynos variant. From 2800 frames to 1400 in 10 minutes of load. http://4pda.ru/forum/index.php?s=&showtopic=707315&view=findpost&p=47363269
If some1 find snapdragon s7 results, please post it here.
Click to expand...
Click to collapse
Ouch i think for sd820 we have to wait a bit since its us and China only
Sent from my SM-G925F
TANKRED_ENDURES said:
GFX Bench battery and throttling test of exynos variant. From 2800 frames to 1400 in 10 minutes of load. http://4pda.ru/forum/index.php?s=&showtopic=707315&view=findpost&p=47363269
If some1 find snapdragon s7 results, please post it here.
Click to expand...
Click to collapse
Says the test ran for 220 minutes. EDIT: Okey that was the results of the battery-test. But still where did you get 10 minutes from? If you look at the graphs you can clearly see that it dips only once to 1400 and that was at about the 1200 second mark (20 minutes).
http://www.talkandroid.com/286767-vivo-xplay-5-gets-benchmarked-on-antutu/#more-286767
Vivo Xplay 5 scored around 160k on Antutu and that is with the Snapdragon 820. Think Samsung should have stuck with Qualcomm for all regions this time around. Wish we could get the sd-version here in Europe as well.
Toss3 said:
http://www.talkandroid.com/286767-vivo-xplay-5-gets-benchmarked-on-antutu/#more-286767
Vivo Xplay 5 scored around 160k on Antutu and that is with the Snapdragon 820. Think Samsung should have stuck with Qualcomm for all regions this time around. Wish we could get the sd-version here in Europe as well.
Click to expand...
Click to collapse
Its probably fake as most of the score comes from GPU which is impossible and then theres 1080p vs qhd screen question
Sent from my SM-G925F
Throttling looks much better compared to 7420.
They have made GPU wider and lower frequency, also better manufacturin process. Bound to get better compared to 7420.
---------- Post added at 12:02 AM ---------- Previous post was at 12:01 AM ----------
skivnit said:
Its probably fake as most of the score comes from GPU which is impossible and then theres 1080p vs qhd screen question
Sent from my SM-G925F
Click to expand...
Click to collapse
Yes, in latest version of Antutu it gives more priority to onscreen numbers and single core performance. That's why iPhones are at top of chain in Antutu