I bought an Android Tablet running an ARM11 (v6) processor. Not sure what the GPU is. However, I ran the NeoCore benchmark for the GPU and got an average of about 13.2 FPS. I compared this against Droid Incredible (Qualcomm Snapdragon) and Droid X (TI OMAP 3630). Both of which as you know that are Cortex A8 variance. All are running at 1.0Ghz.
I also ran Softweg's Benchmark on those 3 devices.
For my tablet, I got scores of about 98 for the GPU and 936 for the CPU.
For the Droid Incredible I got NeoCore score of 26 FPS and for the X, 42 FPS. However, they only scored 27 & 30 respectively on Softweg's GPU test.
I would believe the NeoCore score as I am sure the GPU is poor on the tablet. Why would Softweg's Benchmark app show higher scores on my GPU versus those more advanced Android phones?
My CPU score is also higher than the phones which is not possible. Thank you.
interesting , but u know what , maybe its that neocore bench renders the same amount of data all the time , and the other bench resizes it to the size of the screen , so if ur tab has resolution lower than 480*800 thats the reason why
Actually the resolution is 1024x600. I could understand why the NeoCore score is low which is what I would expect. However, I do not know why my CPU & GPU scores under Benchmark would be higher than a better Cortex processor.
Please post some benchmark scores for us future buyers please? Linpack, neocore, and quadrant (don't kill me) etc..
Linpack 39.67 MFLOPS
Quadrant 1879 just now. Usually between 1700-1900
Neocore 58.0 fps.
Pretty solid.
Mine are in the 1700's as well. I'm truly impressed with that for a $99 phone out of the box. Although I did see lge quad posted on engadget yesterday and it dominates all.
Sent from my Inspire 4G
40.4 mflops and 1897 quadrant on first boot out of the box after installing 20 apps from the market lol, im curious to see scores after root and custom roms and overclocking kernels are released.
Linpack
Hi 40.05 MFLOPS Time: 2.09
Low 36.00 MFLOPS Time: 2.21
Quadrant
1st run: 1845
2nd run: 1895
Neocore
With sound 57.8 fps.
Without sound 58.6 fps.
Hi All,
I was looking HTC one S benchmark over the web and there is one bench called Electopia for graphics.
Just wondering if one of you has already run this bench in full resolution on HTC one S to see how GPU performs.
Thanks !
Hey,
I ran the benchmark in fullscreen and got 60.42 fps (vsync limited)
Feel free to ask if you want some more benchmarks
New Snapdragon S4 is great. I'd love to see Nenamark test
Nenamark1: 60 FPS (vsync limited)
Nenamark2: 60 FPS (vsync limited)
Wow! Really great results
Those are some interesting results, I can't wait until mine arrives.
Hi ossy1337,
Just wondering if you ran electopia in Full resolution or WVGA?
Could you run also Basemark Taiji, nenamark2 & GLBenchmark 2.1.1 (offscreen)
This GPU looks pretty powerfull, but electopia seems ot be too much Qualcomm friendly .
Hey,
I ran it in fullscreen mode not 800x480
Basemark Taiji: 47.16 FPS
GLBenchmark Egypt Offscreen (720p): 56 FPS
Nenamark2: 60 FPS (as mentioned before)
greets,
ossy1337
I've used results from Anandtech and Primate Labs
iPhone 5 scores on the left and One X on the right.
Memory Performance iPhone5 HTC One X
Read Sequential ST- 1.78GB/s 289MB/s
Read Sequential MT- 1.35GB/s 1.56GB/s
Stdlib Allocate ST- 1.92Mallocs/s 2.42Mallocs/s
Stdlib write 6.06Gb/s 1.94GB/s
Stdlib copy 2.26Gb/s 1.94GB/s
Stream Performance
Stream Copy 1.9GB/s 468MB/s
Stream Scale 994MB/s 420MB/s
Stream Add 1.39GB/s 533MB/s
Stream Triad 1019MB/s 537MB/s
Integer Performance
Blowfish ST 23.4MB/s 28.0MB/s
Blowfish MT 45.6MB/s 50.1MB/s
Text Compress ST 2.79MB/s 1.95MB/s
Text Compress MT 5.19MB/s 4.32MB/s
Text Decompress ST 3.82MB/s 2.21MB/s
Text Decompress MT 5.60MB/s 3.55 MB/sec
Image Compress ST 7.31 Mpixels/sec 6.47 Mpixels/sec
Image Compress MT 14.2 Mpixels/sec 16.6 Mpixels/sec
Image Decompress ST 12.4 Mpixels/sec 10.9 Mpixels/sec
Image Decompress MT 23.0 Mpixels/sec 23.7 Mpixels/sec
Lua ST 455 Knodes/sec 471 Knodes/sec
Lua MT 887 Knodes/sec 1.78 Mnodes/sec
Floating Point Performance
Mandelbrot ST 397 MFLOPS 379 MFLOPS
Mandelbrot MT 766 MFLOPS 1.33 GFLOPS
Dot Product ST 322 MFLOPS 465 MFLOPS
Dot Product MT 627 MFLOPS 1.40 GFLOPS
Lu Decomposition ST 387 MFLOPS 168 MFLOPS
Lu Decomposition MT 767 MFLOPS 326 MFLOPS
Primality ST 370 MFLOPS 171 MFLOPS
Primality MT 676 MFLOPS 254 MFLOPS
Sharpen Image ST 4.85 Mpixels/sec 3.99 Mpixels/sec
Sharpen Image MT 9.28 Mpixels/sec 9.59 Mpixels/sec
Blur Image ST 1.96 Mpixels/sec 1.95 Mpixels/sec
Blur Image MT 3.78 Mpixels/sec 5.24 Mpixels/sec
Okay so I'm not a benchmark geek, and its mentioned at Anandtech they don't compare Geekbench scores from platform to platform but these scores were pretty interesting. Our quad core is what makes a difference in a lot of tests, but the A6 is a very powerful processor bundled with really fast memory. I'm not sure what version of android/kernel is used for the One X, I assume the test was taken when One X came with 4.03.
P.S Excuse any typos.
i think these are more clear
hamdir said:
some interesting multicore scores, like predicted a faster newer architecture core but can't beat a quad A9
in real world tough test like image blur and sharpen, we are over 2x
These are the floating point tests, you can see the same equivalent in the integer tests we are nearly 2x
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Processor floating point performance - Multi core: Winner One X
One X
Mandelbrot multi-core scalar score 2345 - 1.53 Gflops
Dot Product multi-core scalar score 4545 - 2.07 Gflops
LU Decomposition multi-core scalar score 907 - 795 Mflops
Primality Test multi-core scalar score 1762 - 327 Mflops
Sharpen Image multi-core scalar score 6514 - 15.0 Mpixels/sec
Blur Image multi-core scalar score 8227 - 6.47 Mpixels/sec
iPhone 5
Mandelbrot multi-core scalar score 1222 - 800 Mflops
Dot Product multi-core scalar score 1413 - 644 Mflops
LU Decomposition multi-core scalar score 877 - 769 Mflops
Primality Test multi-core scalar score 3786 - 703 Mflops
Sharpen Image multi-core scalar score 4177 - 9.63 Mpixels/sec
Blur Image multi-core scalar score 4981 - 3.92 Mpixels/sec
Processor floating point performance - Single core: Winner iPhone 5
One X
Mandelbrot single-core scalar score 514 - 342 Mflops
Dot Product single-core scalar score 1151 - 556 Mflops
LU Decomposition single-core scalar score 233 - 207 Mflops
Primality Test single-core scalar score 1398 - 209 Mflops
Sharpen Image single-core scalar score 1590 - 3.71 Mpixels/sec
Blur Image single-core scalar score 2369 - 1.88 Mpixels/sec
iPhone 5
Mandelbrot single-core scalar score 603 - 402 Mflops
Dot Product single-core scalar score 669 - 323 Mflops
LU Decomposition single-core scalar score 436 - 389 Mflops
Primality Test single-core scalar score 2540 - 379 Mflops
Sharpen Image single-core scalar score 2090 - 4.88 Mpixels/sec
Blur Image single-core scalar score 2490 - 1.97 Mpixels/sec
Processor integer performance - Multi core: Winner One X
One X
Blowfish multi-core scalar 1687 - 69.1 MB/sec
Text Compress multi-core scalar 1439 - 4.72 MB/sec
Text Decompress multi-core scalar 898 - 3.58 MB/sec
Image Compress multi-core scalar 3059 - 25.7 Mpixels/sec
Image Decompress multi-core scalar 2080 - 33.9 Mpixels/sec
Lua multi-core scalar 4891 - 1.88 Mnodes/sec
iPhone 5
Blowfish multi-core scalar 1141 - 46.8 MB/sec
Text Compress multi-core scalar 1659 - 5.45 MB/sec
Text Decompress multi-core scalar 1466 - 5.84 MB/sec
Image Compress multi-core scalar 1737 - 14.6 Mpixels/sec
Image Decompress multi-core scalar 1480 - 24.2 Mpixels/sec
Lua multi-core scalar 2410 - 927 Knodes/sec
Processor integer performance - Single core: Winner iPhone 5
One X
Blowfish single-core scalar 658 - 28.9 MB/sec
Text Compress single-core scalar 476 - 1.52 MB/sec
Text Decompress single-core scalar 560 - 2.30 MB/sec
Image Compress single-core scalar 789 - 6.52 Mpixels/sec
Image Decompress single-core scalar 650 - 10.9 Mpixels/sec
Lua single-core scalar 970 - 374 Knodes/sec
iPhone 5
Blowfish single-core scalar 535 - 23.5 MB/sec
Text Compress single-core scalar 898 - 2.87 MB/sec
Text Decompress single-core scalar 1011 - 4.16 MB/sec
Image Compress single-core scalar 889 - 7.35 Mpixels/sec
Image Decompress single-core scalar 750 - 12.6 Mpixels/sec
Lua single-core scalar 1208 - 465 Knodes/sec
Memory goes to the i5, not surprising given the dual memory channel and newer architecture
Memory performance: Winner iPhone 5
One X
Read Sequential single-core scalar 230 - 289 MB/sec
Write Sequential single-core scalar 1946 - 1.33 GB/sec
Stdlib Allocate single-core scalar 522 - 1.95 Mallocs/sec
Stdlib Write single-core scalar 960 - 1.99 GB/sec
Stdlib Copy single-core scalar 1906 -1.96 GB/sec
iPhone 5
Read Sequential single-core scalar 1479 - 1.81 GB/sec
Write Sequential single-core scalar 1970 - 1.35 GB/sec
Stdlib Allocate single-core scalar 531 - 1.98 Mallocs/sec
Stdlib Write single-core scalar 3038 - 6.29 GB/sec
Stdlib Copy single-core scalar 2252 - 2.32 GB/sec
Stream Memory bandwidth performance: Winner iPhone 5
One X
Stream Copy single-core scalar 296 - 415 MB/sec
Stream Scale single-core scalar 301 - 401 MB/sec
Stream Add single-core scalar 296 - 458 MB/sec
Stream Triad single-core scalar 318 - 451 MB/sec
iPhone 5
Stream Copy single-core scalar 1381 - 1.89 GB/sec
Stream Scale single-core scalar 745 - 991 MB/sec
Stream Add single-core scalar 919 - 1.39 GB/sec
Stream Triad single-core scalar 721 - 1021 MB/sec
Click to expand...
Click to collapse
My bad, I didn't see it. But tegra3 holds its own in multithread tests, I'm quite happy about that.
hamdir said:
i think these are more clear
Click to expand...
Click to collapse
omg why do i get 1200?? O.O what disapponts me is that i get just 1800 in floating point operations it's strange, it's just sbout the speed in calculating with floating point numbers, even if benchmarks aren't so reliable it's really really strange!!
cuzzo94 said:
omg why do i get 1200?? O.O what disapponts me is that i get just 1800 in floating point operations it's strange, it's just sbout the speed in calculating with floating point numbers, even if benchmarks aren't so reliable it's really really strange!!
Click to expand...
Click to collapse
well yea its not really reliable test, restart and test on a fresh start
---------- Post added at 03:18 PM ---------- Previous post was at 03:17 PM ----------
death__machine said:
My bad, I didn't see it. But tegra3 holds its own in multithread tests, I'm quite happy about that.
Click to expand...
Click to collapse
its just a dual krait my friend, nothing more we all know Tegra 3 beats the 2x krait
check how similar
One XL vs iPhone 5
although the iPhone 5 beats it @ 1.0 ghz which is very impressive, but if those tests are indeed unreliable we can't draw any conclusion
what its very clear
its 2 custom cores like krait or a15 vs a quad A9 = sorry you can't beat that which just two cores
One X and SGS3 remains faster in terms of multitasking and general performance
hamdir said:
well yea its not really reliable test, restart and test on a fresh start
---------- Post added at 03:18 PM ---------- Previous post was at 03:17 PM ----------
xd i tryed another time without restarting and now it's 1400 XD quadrant it's more stable with results
Click to expand...
Click to collapse
cuzzo94 said:
hamdir said:
well yea its not really reliable test, restart and test on a fresh start
---------- Post added at 03:18 PM ---------- Previous post was at 03:17 PM ----------
xd i tryed another time without restarting and now it's 1400 XD quadrant it's more stable with results
Click to expand...
Click to collapse
Still would be interesting to know which rom you are on anyway? 1.28,1.29? 2.17? custom rom?
IMO i don't really care much about competition. New phones will always get outperformed by newer phones. Do i like to see how a phone performs benchmark wise? Sure thing! Will that make me regret my phone decision etc? Not anytime soon, no!
I just hope htc does well in future. This is my first HTC phone, and i've even been considering a windows phone if they improve in future. It's either going to be a nokia or a htc phone anyway
Click to expand...
Click to collapse
@shadehh
Those are my exact same thoughts. I wouldn't have regretted if it didn't beat the scores, I would have told myself oh but this came out MONTHS after the hox xD. But it is holding its own! That's what makes me happy.
shadehh said:
cuzzo94 said:
Still would be interesting to know which rom you are on anyway? 1.28,1.29? 2.17? custom rom?
IMO i don't really care much about competition. New phones will always get outperformed by newer phones. Do i like to see how a phone performs benchmark wise? Sure thing! Will that make me regret my phone decision etc? Not anytime soon, no!
I just hope htc does well in future. This is my first HTC phone, and i've even been considering a windows phone if they improve in future. It's either going to be a nokia or a htc phone anyway
Click to expand...
Click to collapse
i'm on stock rom updated at 4.0.4. however i always get around 6000 on quadrant
Click to expand...
Click to collapse
the memory performance tests are kinda strange we match in Write Sequential and Stdlib Allocate
i dont think this test is very reliable to be honest especially in memory tests
---------- Post added at 03:58 PM ---------- Previous post was at 03:50 PM ----------
death__machine said:
@shadehh
Those are my exact same thoughts. I wouldn't have regretted if it didn't beat the scores, I would have told myself oh but this came out MONTHS after the hox xD. But it is holding its own! That's what makes me happy.
Click to expand...
Click to collapse
half a year to be exact
also Tegra 3 is now almost a year old
hamdir said:
the memory performance tests are kinda strange we match in Write Sequential and Stdlib Allocate
i dont think this test is very reliable to be honest especially in memory tests
---------- Post added at 03:58 PM ---------- Previous post was at 03:50 PM ----------
half a year to be exact
also Tegra 3 is now almost a year old
Click to expand...
Click to collapse
I just can't, i can't wait to see tegra 4 come out and mop the contest. Just imagine a tegra 4 + 1 and htc.
I'm just speculating, but since Nvidia is boss at developing one of the best GPUs for PCs, i see no reason why they wouldn't be able to bring out a bad-ass GPU for phones as well. Really looking forward to that day.
i wouldn't worry about the GPU on the next Tegra
Their mamoth Kepler GPU was designed from scratch with low power and mobile in mind
they very well intend to take out PowerVR in the next round
their next GPU is expected to have at the very least 24 unified cores
unlike our current situation where 4 vertex cores are often underused
and they were the first to embrace test and trial, ARM's big.LITTLE concept with the LP core on Tegra3, LP core is integral for the a15 adoption (frequency and power hungry)
Tegra 4 = 4x A15 + LP core @ 28nm + 24 unified cores GPU
hamdir said:
i wouldn't worry about the GPU on the next Tegra
Their mamoth Kepler GPU was designed from scratch with low power and mobile in mind
they very well intend to take out PowerVR in the next round
their next GPU is expected to have at the very least 24 unified cores
unlike our current situation where 4 vertex cores are often underused
and they were the first to embrace test and trial, ARM's big.LITTLE concept with the LP core on Tegra3, LP core is integral for the a15 adoption (frequency and power hungry)
Tegra 4 = 4x A15 + LP core @ 28nm + 24 unified cores GPU
Click to expand...
Click to collapse
Too bad I have a 2 year contract hopefull tegra 5 will be some next level shiz like 2ghz 8x A15, is it needed no, but who cares
Sent from my HTC One X using xda premium
Nitin985 said:
Too bad I have a 2 year contract hopefull tegra 5 will be some next level shiz like 2ghz 8x A15, is it needed no, but who cares
Sent from my HTC One X using xda premium
Click to expand...
Click to collapse
nothing will appear in terms of smartphones before next MWC and release almost up to june
plenty of time
look since they release after 6 months, i was worried the i5 to be a One X/SGS3 killer in terms of speed and to be honest it's not
Tegra will keep compensating its GPU using the quad, like it is already does with many THD games
equally Nvidia will not drop T3 from its support like T2, they intend to release a lower priced T3 shrink along T4, so the games support will keep up
One year latter the Tegra 3 still hold on quite well versus brand news SoC.
I love How many people bash Nvidia for how badly made the Tegra 3 is but the fact is it destroyed everything on year ago and set the golden standard for games on Android.
I'm super eager to see what Nvidia have in store with the Tegra 4. From what we ve heard so far it seems it will be a major step up. All very exciting !
kokusho said:
One year latter the Tegra 3 still hold on quite well versus brand news SoC.
I love How many people bash Nvidia for how badly made the Tegra 3 is but the fact is it destroyed everything on year ago and set the golden standard for games on Android.
I'm super eager to see what Nvidia have in store with the Tegra 4. From what we ve heard so far it seems it will be a major step up. All very exciting !
Click to expand...
Click to collapse
well said, i actually see it as a solution for the gaming problems on android
and honestly i don't see how badly made it is?
all issues were software and the chip is the most stable i ever messed with despite all the thermal red lines
hamdir said:
nothing will appear in terms of smartphones before next MWC and release almost up to june
plenty of time
look since they release after 6 months, i was worried the i5 to be a One X/SGS3 killer in terms of speed and to be honest it's not
Tegra will keep compensating its GPU using the quad, like it is already does with many THD games
equally Nvidia will not drop T3 from its support like T2, they intend to release a lower priced T3 shrink along T4, so the games support will keep up
Click to expand...
Click to collapse
important hoping they won't drop it. the support for tegra makes it the best i think. but my contract finishes in 2014 so long wait for me. my sister is getting her iphone 5 on Friday so im going to steal it of her asap and compare. Keeping up with tech is sometimes more stressful than school work
Sent from my GT-I9100 using xda app-developers app
kokusho said:
One year latter the Tegra 3 still hold on quite well versus brand news SoC.
I love How many people bash Nvidia for how badly made the Tegra 3 is but the fact is it destroyed everything on year ago and set the golden standard for games on Android.
I'm super eager to see what Nvidia have in store with the Tegra 4. From what we ve heard so far it seems it will be a major step up. All very exciting !
Click to expand...
Click to collapse
It only set the gold standard for gaming on Android because of Nvidia's financial muscle. That says nothing of the abilities of Tegra 3's GPU.
@Op,
Compair devices, wrong XDA section and against forum rules.
See first sticky! :cyclops:
Wow, bench marks! Impressive. The HOX is indeed perfect and the best. What a useless POS the iPhone5 is! Can't even make phone calls. HTC is the best, better than Apple in everything !
MIUI 10.2.3.0 Global stable. Not rooted:
# Antutu:
Overall score = 299961
CPU: 94829
GPU: 127706
UX: 66181
MEM: 11245
# 3DMark Sling Shot Extreme:
OpenGL ES 3.1:
Overall score: 3473
Graphics score: 4245
Physics score: 2122
Vulkan:
Overall score: 3168
Graphics score: 3724
Physics score: 2080
MIUI 10.3.4.0 Global stable. Not rooted. (with Game Turbo on):
# Antutu:
Overall score = 299037
CPU: 96859
GPU: 127699
UX: 64577
MEM: 9902
# 3DMark Sling Shot Extreme:
OpenGL ES 3.1:
Overall score: 3648
Graphics score: 4043
Physics score: 2719
Vulkan:
Overall score: 3451
Graphics score: 3970
Physics score: 2367
Performance Summary of 10.3.4.0:
Performance downgrade on:
- OpenGL Graphics (Most used API for android games)
- UX (User experience: scroll delay test, WebView rendering, QR code recognition test)
- MEM (RAM and ROM performance)
Performance upgrade on:
- Vulkan Graphics (Less used for android games)
- CPU (Processor)
- Physics (simulation of physics in games)
Verdict on MIUI 10.3.4.0 :
- Games may run slower on 10.3.4.0 compared to 10.2.3.0
- Downgrade of UX and MEM may affect users in their daily use or may not.
- CPU dependent tasks may perform better.
- Vulkan based rendering may run faster.
Pfffffffff....
For daily use it makes no difference....
As we say in France: you're searching for hair on an egg...
chrissey said:
Pfffffffff....
For daily use it makes no difference....
As we say in France: you're searching for hair on an egg...
Click to expand...
Click to collapse
Can affect some. For example, I'm seeing longer game loading time on 10.3.4 because MEM (RAM & ROM speed) decreased.
Never benched 10.2.3.0 as phone updated the day I received it. Attached are my 3D Mark results.
Stransky said:
Never benched 10.2.3.0 as phone updated the day I received it. Attached are my 3D Mark results.
Click to expand...
Click to collapse
Nice scores. Which MIUI version?
MIui 10.3.4.0 Stock, locked.
Sent from my POCOPHONE F1 using Tapatalk
Stransky said:
MIui 10.3.4.0 Stock, locked.
Sent from my POCOPHONE F1 using Tapatalk
Click to expand...
Click to collapse
Hmm... Seems high for stock miui. Did you use GLTools or any tweaking?
What about antutu scores?
Not tweaks, just pure stock.
Antutu attached.
Sent from my POCOPHONE F1 using Tapatalk
Geekbench 4 MIUI Global 10.3.4.0 pure stock no tweaks or mods. Phone locked.
Flawed comparison.
Do each test 10 times then calculate the average of all 10 tests.
Also be sure to reboot several times and let the device sit idle on charge for an hour or two before doing any tests at all so ART can finish background optimization.
only if benchmark scores were a true reflection of performance. Once people understand that it is not, most rom threads will be clean and actually helpful.