Develop Bugs

Is Qualcomm cheating in Vellamo?

Some food for thought,,
Is Qualcomm cheating in Vellamo?
A few days ago we were happy to report about the first set of leaked HTC One X scores on an American device powered by Qualcomm’s new 28nm S4 chip.
The scores look quite good for Qualcomm, but since we got hold of an Asus Transformer Prime we decided to dig a bid deeper. Both Nvidia and Qualcomm do not want to get involved in official fight but we were quite surprised with what we learned. We also talked to some engineers that want to remain unnamed and we came up with a few interesting things.
The fastest CPU in Quadrant is Hummingbird S5PC110 from Nexus S as tested here.
Since we have tested this phone you can easily tell that this benchmark doesn't really use two or four cores properly. In this test S4 scores 4920 where the Tegra 3 based Transformer Prime scores 3954 and Tegra 2 scores 2154 on a custom ROM Optimus 2X.
The second test is the Vellamo benchmark was a bit more disturbing as once you run it you see Qualcomm ads all over it. This didn't give us much confidence and it turns out that Qualcomm has a lot of power over this particular benchmark.
Asus Transformer Prime scores 1408, while Qualcomm in Vellamo scores over 2000, our guess is between 2200 and 2300 as we didn't see the full number. Our good buddy Anand compared the One S powered by the S4 as well as the One X powered by Tegra 3 and you can see that Tegra 3 on this phone usually ends up faster or tied with the S4. It is faster in Sunspider Javascript benchmark 0.9.1, loses by a few points in Browsermark to S4 based HTC One S and dominates GLBenchmark. There is no 2X performance lead that we saw in the leaked S4 benches and frankly we see no point in taking Vellamo seriously until the issue is addressed.
Our engineering friends are telling us that Velamo disables some hardware acceleration in compositing Deap Sea Canvas and See the sun canvas subtest. Honeycomb and ICS support hardware acceleration by default and disabling this probably hurt the general score. One can argue that it hurts S4 scores as well, but it definitely hurts Tegra 3 more. The benchmark isn't flushing commands in the Pixel Blender subtest and there is a suspicion that this might help Qualcomm S4 to gain a better score.
The most important issue is the fact that it is unclear how Vellamo scores sub test scores. In Third party benchmarks such as Sun Spider and Google V8 it turns out that Vellamo penalizes high Google V8 scores and if your score gets too high in V8, the general score gets lower. There are a lot of benchmarks out there and some of the ones that like more cores include Antutu, CF benchmark and Moonbat.
French enthusiasts managed to run Antutu here, and Tegra 3 phone scores 10597, while the S4 based HTC One S scores 6458. This doesn't look so good for the phone that is based on S4 cores that should go after ARM’s upcoming A15 and it looks like that it cannot really beat the A9-based Tegra 3. In the real world, as long as the application is aware of four cores, there is a good chance that Tegra 3 will end up faster than the S4.
Naturally if you are reading this from the US and you really like your LTE from AT&T or Verison, I guess that you won’t have much choice and you will get the HTC One X with Qualcomm S4 as this chip also supports LTE. In the US it’s all about LTE and in Europe despite the fact that countries like Austria have LTE for more than a year now, even at €50 for 40GB there is almost no interest whatsoever. Networks need to put more advertising money and make 4G cool and it might happen. Of course, the lack of 4G devices is also an issue, but technology has a way of catching up.
The story gets even better when you know that there are lot of former ATI employees who take care of Adreno graphics and if natural selection and theory of evolution have taught us anything, it is that Nvidia and ATI are sworn enemies in any universe. Some readers might see a pinch of poetic justice in all of this, as Nvidia was accused of tweaking its GPU drivers to score more in PC benchmarks years ago.
Tudor Brown, the president of ARM that we meet a few years back at GlobalFoundries Dresden fab, once said that ARM does not want to get involved in GHz fight and this is now exactly what is going on, as punters are using benchmarks to prove of A is faster than B. Frankly I would be more concerned about battery life that I can get from a brand new phone as we got from five day battery life on feature phones to a day or so of battery life, and with LTE and heavy use, even this can go down to a few hours at best. If I found myself in Qualcomm’s shoes, this is what I would emphasize, the new 28nm process and energy efficiency, not skewed benchmarks.
Phones should be about the overall user experience, but how can you benchmark experience? It's a very subjective category and can hardly be expressed with cold numbers and statistics. We are not talking about PC components and the race for more performance, smartphones should be viewed as a complete package, battery life, build quality, UI and design come into play. Sheer performance is just one aspect and for many users it is still not the deciding factor and we believe it should not be.
In any case, watching the phone market will definitely be fun in the months to come.
Click to expand...
Click to collapse
Source: Here

arthasz said:
Some food for thought,,
Source: Here
Click to expand...
Click to collapse
haha fun stuff ill quote some of it in my thread if you don't mind thanks
to be honest this is exactly what i felt the moment i thought the score, its ridiculous
glad we got confirmation from some devs too
---------- Post added at 12:13 PM ---------- Previous post was at 11:56 AM ----------
speaking of which, their Adreno3x series is far from ready
Qualcomm has Imagination
Apparently they might go with PowerVR for the windows s4 devices, i find this hard to believe
Semi accurate is part of this fud game

Now things make sense. That's why people don't blindly trust these benchmark score. Try real work applications you'll better idea of the device like running 12 HD video in T3 etc.

I've always trusted Antutu
& apparently, the tegra 3 version is a beast, scoring over 10,000!
My SGS2 scores just 4,700 on gingerbread & around 5,700 on ICS.
Sent from my GT-I9100 using Tapatalk

My overclocked (1,4 GHz) Desire Z (ICS with Sense 4.0A) got around 1 200 points in Vellamo, in graph it was just above Transformer Prime. Qualcomm must cheating.
Edit: added screenshots (due to my bad memory - it was just above while at 1,4 GHz ), at 1,6 GHz it is will be better than Prime (according to benchmark)

Reremnu said:
My overclocked (1,4 GHz) Desire Z (ICS with Sense 4.0A) got around 1 200 points in Vellamo, in graph it was just above Transformer Prime. Qualcomm must cheating.
Click to expand...
Click to collapse
hahaha now that is crazy, disappointed that qualcomm took this direction they always been more transparent
actually its wrong that they make a test themselves it can never be considered impartial
snapdragon 4 cores are impressive as they are they don't need this kind of fud just because they don't have a quad ready

If this is really true, how about other benchmarks with other devices? can they be trusted?

[Q] GPU comparison Mali400 vs Adreno 225 vs Adreno 320

Hi,
I am planning to swap my Note 2 and GFs S3 with either Xperia T or Nexus 4
Nexus is the best when it comes to gaming with its superb GPU but it has its own problems like performance drops due to heating, cracking class, relatively worse screen, lack of SD card slot etc.
So I am firstly thinking of Sony devices as they have almost zero manufacturing defects unlike cracking, heating LG or burn-in suffering and eMMc frying Samsung
I use the CPU at 1200 Mhz since it is not necessary for todays needings. But GPU is important for future games and I am afraid our mali 400 is outdated.
When we compare 3 GPUs with GLBenchmark Adreno 320 wins all rounds but things are complicated between Adreno 225 and mali 400.
Mali 400 is not exceeding 19FPS (normally 17) no matter how much i OC or what resolution I choose on the Egypt 2.5 test. Adreno 225 scores 22-23 FPS on 720p without OCing.
Things change when we compare older 2.1 Egytp test. With Xperia T on both onscreen or offscreen scores goes pretty low behind Note 2.
It gets more interesting when I run Egytp 2.5 with 4xAA. Note 2 barely gets effected from it and it still gives the 17FPS. Apperiently the part which makes the AA is not where Mali 400 is bottle necked. On the other hand Adreno 225 gets slower on the AA side and gets 16FPS.
I understand that the Adreno is newer so is Egytp 2.5 test. But which one should i take into account for the future gaming?
And more importantly should I make the swap?
Thanks for reading.

Honestly speaking I doubt it will make much noticeable difference with games for this year. It has managed to run smoothly at every game I've thrown at my note 2. I think its best if you stick with your note and your gf S3 for the time being. Its funny tho I got my wife an S3 and myself the Note 2. She got envious until she used my phone with one hand and dropped it twice, since then she's quite happy sticking with her S3.-_-
Sent from the Rabbit Hole

You are thinking about Nexus 4 - where can you actually get one? Last time I checked it's been sold out everywhere. If it's not available to buy, then it's an easy choice

I dont think you will
Sent from my GT-N7100 using xda premium

Hey dude..i will tell.you my own experience:
Last year i boughy my first Android Device which was Sony Xperia S...it was the..no 1 at high end devices...it beated every phone.
It has a snapdragon.S4 1.5 ghz dual core.
Adreno.320 gpu.
Honestly at stock ...and even after rooting and overclocking....nothing run smoother then it was already.
But i put it in his box because these high end specs..didnt give the phone enough juice the whole day just using it for chatting.
Then i saw the new note 2..woow..the biggest point that i bought my note 2 was..the battery at first...and the big screen..and the secs.
I bought every high end game on the market..Asphalt 7..finished..moderm combat 4..at 2nd level playing...nova 3 finished..and my loved first fps game shadowgun...i also installed back stap hd...and gt racing..max payne ..gta 3 and vice city....believe me not for a second the games hanged anywhere...they where soo smooth and without lags.
Your note 2 will last you at least 2 more years for high end games.
It is the best phone on the market.
Best regards.
Sent from my GT-N7100 using xda premium

zdravke said:
You are thinking about Nexus 4 - where can you actually get one? Last time I checked it's been sold out everywhere. If it's not available to buy, then it's an easy choice
Click to expand...
Click to collapse
This made the choice very very easy for me!

_delice_doluca_ said:
Hi,
I am planning to swap my Note 2 and GFs S3 with either Xperia T or Nexus 4
Nexus is the best when it comes to gaming with its superb GPU but it has its own problems like performance drops due to heating, cracking class, relatively worse screen, lack of SD card slot etc.
So I am firstly thinking of Sony devices as they have almost zero manufacturing defects unlike cracking, heating LG or burn-in suffering and eMMc frying Samsung
I use the CPU at 1200 Mhz since it is not necessary for todays needings. But GPU is important for future games and I am afraid our mali 400 is outdated.
When we compare 3 GPUs with GLBenchmark Adreno 320 wins all rounds but things are complicated between Adreno 225 and mali 400.
Mali 400 is not exceeding 19FPS (normally 17) no matter how much i OC or what resolution I choose on the Egypt 2.5 test. Adreno 225 scores 22-23 FPS on 720p without OCing.
Things change when we compare older 2.1 Egytp test. With Xperia T on both onscreen or offscreen scores goes pretty low behind Note 2.
It gets more interesting when I run Egytp 2.5 with 4xAA. Note 2 barely gets effected from it and it still gives the 17FPS. Apperiently the part which makes the AA is not where Mali 400 is bottle necked. On the other hand Adreno 225 gets slower on the AA side and gets 16FPS.
I understand that the Adreno is newer so is Egytp 2.5 test. But which one should i take into account for the future gaming?
And more importantly should I make the swap?
Thanks for reading.
Click to expand...
Click to collapse
aloush5592 said:
Hey dude..i will tell.you my own experience:
Last year i boughy my first Android Device which was Sony Xperia S...it was the..no 1 at high end devices...it beated every phone.
It has a snapdragon.S4 1.5 ghz dual core.
Adreno.320 gpu.
Honestly at stock ...and even after rooting and overclocking....nothing run smoother then it was already.
But i put it in his box because these high end specs..didnt give the phone enough juice the whole day just using it for chatting.
Then i saw the new note 2..woow..the biggest point that i bought my note 2 was..the battery at first...and the big screen..and the secs.
I bought every high end game on the market..Asphalt 7..finished..moderm combat 4..at 2nd level playing...nova 3 finished..and my loved first fps game shadowgun...i also installed back stap hd...and gt racing..max payne ..gta 3 and vice city....believe me not for a second the games hanged anywhere...they where soo smooth and without lags.
Your note 2 will last you at least 2 more years for high end games.
It is the best phone on the market.
Best regards.
Sent from my GT-N7100 using xda premium
Click to expand...
Click to collapse
xperia s has s3 and adreno 220....

[Q] Is exynos worth buying?

The snapdragon version isn't available in my country, so I will have to buy the exynos (Pretty cheap right now $500 equivalent). The thing is reviews say the snapdragon doesn't lag a bit while exynos is made for a large device.
Is the performance really this bad? I'm not into eons right now by the way.

No its not worth buying the snapdragon version. My s4 is faster than my note...
Sent from my GT-I9505 using xda app-developers app

With HMP enabled there is no comparison between the two, exynos is up to 50% faster and potentially more efficient. With HMP disabled (as things currently are) then qualcomm is the slightly better chip, but I'm not convinced that the difference is enough to prefer one soc over the other...
In short Exynos 5420 is artificially neutered to seem worse than qualcomm, yet -even so- going either way won't make much of a difference...

Do you have any benchmarks to prove your claim of a speed bump of 50 %?
to OP
There are a lot of threads about Exynos vs snapdragon, long story short
Exynos , tad better cpu
Snapdragon tad better gpu
I've had both, ended with exynos , because I didn't need 4g, but needed 32 GB ( in scandinavia 4 G seems to be 16 gb only)
Lag was more or less the same
I felt the battery time on the exynos was a tad better
They felt equally as snappy when they needed to
BUT!!!
App support was a tad better on Snapdragon, ie more apps in the plastore worked with the snapdragon version, a few more games etc... no big deal for me, but still get me ticked of when I noticed a few apps I bought weren't compatible ( yet?!) with the new exynos chip ( but worked with my sammy S3 also exynos chip, older )

Exynos is fine. I've played with both and from a UI and app use perspective you can't tell the difference. Adreno's a bit faster than Mali but no so much as to drastically alter performance. Some games are better optimized for Adreno so depending on your choice of games it could make a difference. As for app compatibility it's more likely the 2,560x1,600 display that's causing the issue not the specific SoC. If there were huge differences between Exynos and S-800 or drastic app performance differences and app compatibility issues it would be all over the N3 forum and it's not.

DeBoX said:
Do you have any benchmarks to prove your claim of a speed bump of 50 %?
to
Click to expand...
Click to collapse
HMP for 8 cores have not yet released but look at Note 3 Neo, it uses 2 less large cores and it posts the same antutu score as our note, so by adding two more large cores you can expect the score to be about 50% more. As I said that is only true were all 8 cores would be used at the same time and they are not throttled (that is why I said "up to").

Stevethegreat said:
Look at Note 3 Neo, it uses 2 less large cores and it posts the same antutu score as our note
Click to expand...
Click to collapse
Not really. It also has a 267 PPI display which is benefitting its graphics scores in AnTuTu compared to the SGS4 at 441 PPI and N3 at 386 PPI.
http://www.nairaland.com/1597298/samsung-budget-galaxy-note-neo
S-800 vs. Exynos on the N3...

BarryH_GEG said:
Not really. It also has a 267 PPI display which is benefitting its graphics scores in AnTuTu compared to the SGS4
Click to expand...
Click to collapse
I was more properly referring to CPU scores which are the only ones benefitted from HMP.
I ran a quick AnTuTu (cpu) test to my Exynos 5420 equipped note and here are the results: http://i.imgur.com/zD32DZQ.png
Notice how remarkably similar they are to note neo's cpu score:
http://www.gsmarena.com/showpic.php3?sImg=newsimg/14/01/sgn3n-leak/gsmarena_006.jpg&idNews=7538
Note that note neo has only two large cores which are clocked lower by 10% compared to exynos 5420 and it still posts almost the same score merely by employing the help of the small cores. Now add two large cores more and you'd get 50% more performance, it's simple math really...
Now I'm not saying that it would be a performance that we would actually see in most occasions , it would either be throttled or -even- not supported by most apps but still it's potentially there (which was my point by saying "up to").
What will *definitely* be there if HMP is to be enabled is better battery -though- as it would make more efficient use of the small cores. Since exynos 5422 is also on 28nm yet has HMP enabled leads me to believe that we lack HMP for strategic reasons (so that samsung will sell more exynos 5422 / qualcomm equipped machines)

Stevethegreat said:
I was more properly referring to CPU scores which are the only ones benefitted from HMP.
I ran a quick AnTuTu (cpu) test to my Exynos 5420 equipped note and here are the results: http://i.imgur.com/zD32DZQ.png
Notice how remarkably similar they are to note neo's cpu score:
http://www.gsmarena.com/showpic.php3?sImg=newsimg/14/01/sgn3n-leak/gsmarena_006.jpg&idNews=7538
Note that note neo has only two large cores which are clocked lower by 10% compared to exynos 5420 and it still posts almost the same score merely by employing the help of the small cores. Now add two large cores more and you'd get 50% more performance, it's simple math really...
Now I'm not saying that it would be a performance that we would actually see in most occasions , it would either be throttled or -even- not supported by most apps but still it's potentially there (which was my point by saying "up to").
What will *definitely* be there if HMP is to be enabled is better battery -though- as it would make more efficient use of the small cores. Since exynos 5422 is also on 28nm yet has HMP enabled leads me to believe that we lack HMP for strategic reasons (so that samsung will sell more exynos 5422 / qualcomm equipped machines)
Click to expand...
Click to collapse
You can't divorce the impact of display area size and PPI from CPU performance. The GPU doesn't absolve the CPU's role in graphics output. An i3 PC with a killer graphics card will perform worse graphically than an i7 PC with a lesser card because most computational (not rendering, texture mapping, vectoring, and decoding) work is still done on the CPU. So I have no idea what AnTuTu's testing to come up with a CPU rating in isolation but if it's a real-time performance test the CPU's role in graphics output is impacting it. So comparing the Neo with a 5.5" display and 267 PPI against the N10.1-14 with a 10.1" display and 299 PPI isn't going to get you a relevant CPU comparison. That's why I used the N3 and SGS4 as comparisons because only the PPI is off. And the Neo would be well behind the SGS4 in the cumulative AnTuTu test if it had the same PPI because the lower workload of the lower PPI is artificially enhancing its score. At the end of the day an isolated CPU number is pretty meaningless. It's like bench horsepower in a car vs. horsepower to the wheels. A higher bench rating means nothing because none of us drive an engine, we drive a car. The total AnTuTu number (AKA: drive train loss) is more relevant even though it doesn't support the point you're trying to make about HMP.
http://en.wikipedia.org/wiki/Graphics_processing_unit#Computational_functions

BarryH_GEG said:
You can't divorce the impact of display area size and PPI from CPU performance. The GPU doesn't absolve the CPU's role in graphics output. An i3 PC with a killer graphics card will perform worse graphically than an i7 PC with a lesser card because most computational (not rendering, texture mapping, vectoring, and decoding) work is still done on the CPU. So I have no idea what AnTuTu's testing to come up with a CPU rating in isolation but if it's a real-time performance test the CPU's role in graphics output is impacting it. So comparing the Neo with a 5.5" display and 267 PPI against the N10.1-14 with a 10.1" display and 299 PPI isn't going to get you a relevant CPU comparison. That's why I used the N3 and SGS4 as comparisons because only the PPI is off. And the Neo would be well behind the SGS4 in the cumulative AnTuTu test if it had the same PPI because the lower workload of the lower PPI is artificially enhancing its score. At the end of the day an isolated CPU number is pretty meaningless. It's like bench horsepower in a car vs. horsepower to the wheels. A higher bench rating means nothing because none of us drive an engine, we drive a car. The total AnTuTu number (AKA: drive train loss) is more relevant even though it doesn't support the point you're trying to make about HMP.
http://en.wikipedia.org/wiki/Graphics_processing_unit#Computational_functions
Click to expand...
Click to collapse
Maybe so, but the benchmark in question runs off screen. So while in real life resolution matter in Antutu Cpu score, or super pi , or, or, it doesn't. HMP will make the Cpu 50% faster in multi threaded operations, I never claimed it makes the total machine faster by the same amount. For example an HMP equipped note 2014 will score around 40000 in Antutu , NOT 49500. I don't see where we disagree, I merely think you misunderstood my initial claim

If you live for real world use, the Exynos Note is a wonderful tablet. If you live in the world of needing the highest quadrant and antutu scores you should pass.
Sent via Tapatalk and my thumbs.

Stevethegreat said:
With HMP enabled there is no comparison between the two, exynos is up to 50% faster and potentially more efficient. With HMP disabled (as things currently are) then qualcomm is the slightly better chip, but I'm not convinced that the difference is enough to prefer one soc over the other...
In short Exynos 5420 is artificially neutered to seem worse than qualcomm, yet -even so- going either way won't make much of a difference...
Click to expand...
Click to collapse
How did you enable HMP? My note 3 snap dragon is so much faster than my note.
Sent from my SM-N900T using XDA Premium 4 mobile app

Stevethegreat said:
HMP for 8 cores have not yet released but look at Note 3 Neo, it uses 2 less large cores and it posts the same antutu score as our note, so by adding two more large cores you can expect the score to be about 50% more. As I said that is only true were all 8 cores would be used at the same time and they are not throttled (that is why I said "up to").
Click to expand...
Click to collapse
It will never be released for Exynos 5420 either, unless Samsung want alot of complains about fried Exynos 5420 chipsets. Also they already said it wont release HMP for Exynos 5420 cause of the heat.

dt33 said:
It will never be released for Exynos 5420 either, unless Samsung want alot of complains about fried Exynos 5420 chipsets. Also they already said it wont release HMP for Exynos 5420 cause of the heat.
Click to expand...
Click to collapse
Once again, that's not the reason that they won't release it, if anything the chip would be cooler because more use of A7 cores would be possible and if all 8 cores are needed Samsung could choose to throttle the thing. The reason that they don't release it is the Exynos 5422 which is the same chip but with all 8 cores enabled (also 28nm)...
So no fried socs, lesser profits more like

[Q] same phone different specs....

Hi all,
so me and a friend got our t-mobile Samsung Note 3 within a week of each-other at the same T-mobile store about 8 months ago, so just the other day we were siting around and ran Antunto to see how our phones would ring up, we both have stock phones, he ran 36??? on Antutu and I got 42742, so we started to look at our specs and he has an Qualcomm Snapdragon MSM8974 and I had a Qualcomm Snapdragon MSM8974AB even our dpi are different he has I think 340 and I have 480 dpi...
Any info on this would be appreciated.
Thanks

Did you guys have the same version of Antutu?

Note 3
Techniques said:
Did you guys have the same version of Antutu?
Click to expand...
Click to collapse
Yes Antutu 5, but even other benchmarks there always was at least a 10 to 12% difference...

The Note 3 uses and Snapdragon 800 - 8974AA SOC
I suspect something is off in the build.prop or Antutu had a brain fart.

I agree
rjohnstone said:
The Note 3 uses and Snapdragon 800 - 8974AA SOC
I suspect something is off in the build.prop or Antutu had a brain fart.
Click to expand...
Click to collapse
Don't think im complaining, Sh*&t , I love my phone its fast and vibrant, was just wondering if any one had this happen to them...I feel bad is for my friend who thinks his phone sucks...
but even the screen ppi of 480 is above norm....

480 is stock DPI. I use app settings to maximize real estate on my screen, usually end up using 400/420 DPI myself so it doesn't look like something a 3yr old would use. As for the antutu, installed apps, and apps running in the background, and time since last reboot all contribute to your overall score. Even having a case that restricts heat dissipation will cause you to hit the thermal throttle easier as well.
BTW, same model number with the ab and DPI on my device, 44053 over here.
Sent from my SM-P600 using XDA Premium HD app

Rn8pro performance benchmark

I saw articles of helio g90t proccesor giving great results like 2410 for single core and 7193 for multi core. When testin my rn8pro the results are 505 and 1641 respectivly on geekbench.i will attach photos below.
Anybody knows why so much difference? Is it normal and can we do anything to get higher performance?

You're comparing Geekbench 4 vs 5. Scores on Geekbench 5 are not comparable with scores from Geekbench 4.

reflecx said:
You're comparing Geekbench 4 vs 5. Scores on Geekbench 5 are not comparable with scores from Geekbench 4.
Click to expand...
Click to collapse
i checked it and it is true

ariber14 said:
i checked it and it is true
Click to expand...
Click to collapse
just check the other socs results the g90t is pretty close to the snapdragon 845 multi core score (around 200-300 difference)
and ye those results are from the forth version.