An Antutu benchmark discovered today shows us the type of specifications to expect for the Meizu Pro 6, but are these specs you expected from the Meizu MX6 instead?
Last year Meizu seemed to have drawn a line between their flagship phones. From what many Meizu fans understood the MX range of devices would remain as the flagship Mediatek powered phones, while the Pro line would feature higher specifications and Samsung processors. The first examples of this were the Meizu MX5 with Mediatek Helio X10 and the Pro 5 with Samsung Exynos chip.
However the lines have blurred once again with the revelation that the next generation Meizu Pro 6 all get a Helio X25, and after today’s Antutu leak a spec close to what we were expecting on the MX6.
Gizchina News of the week
The Antutu benchmark shows the Meizu Pro 6 with a 1920 x 1080 display, 3GB RAM, 32GB memory, 5 mega-pixel front camera, 21 mega-pixel rear, Android 6.0 and a Mediatek MT6797T aka Helio X25 (the X20 with a slight speed boost). The Pro 6 managed to get 99,948 points in the test.
For those who know your Meizu phones you will see that the only update here from the MX5 is the chipset, so once again the update from phone to phone this year from Meizu seems to be quite a small one.
Another change we might see is the absence of laser autofocus and a slightly redesigned body.
What do you think of these Pro 6 specs? are they Pro worthy or should they be on the MX6 instead?
Look at single core integer and FP scores, the memory score just distorts the total http://browser.primatelabs.com/geekbench3/compare/5751599?baseline=5751615
GPU vs Mate 8 https://gfxbench.com/compare.jsp?benchmark=gfx40&did1=31491219&os1=Android&api1=gl&hwtype1=GPU&hwname1=ARM+Mali-T880&D2=Huawei+Mate+8+%28NXT-xxx%29
vs the Exynos https://gfxbench.com/compare.jsp?benchmark=gfx40&did1=31491219&os1=Android&api1=gl&hwtype1=GPU&hwname1=ARM+Mali-T880&D2=Samsung+Galaxy+S7+%28Mali-T880%2C+SM-G930x%29
The big question being how much it throttles.
I didn’t know they had charts like that! 🙂 thanks
Wow. Thew memory performance is a shock.
It’s normal not a shock and pretty irrelevant.The Geekbench memory score pretty much scales with the memory bandwidth, unless there is another bottleneck. But the memory perf depends on latency too and actual apps react in very different ways to memory perf.
Then again, you are just trolling.
Not actually true as the GPU is clocked higher than in Kirin 950 and should deliver absolutely playable performance for all current games running 2K res.
Sure, the GPU isn`t as powerful as the Adreno 430 and Mali 760MP8 from last year, but it has a more rich feature set and it won`t have any problems running anything currently on the Android market.
The GPU is clocked lower than Kirin 950. 850mhz for x25 and 900mhz for the Kirin. Go and see the french website frandroid Kirin 950 has insane performance in game better than Samsung S7
It could be, I assumed as much from the tests(showing slightly higher scores on the X25), but no chance that it performs better than the MP12 Version inside the Exynos 8890 or Adreno 530(SD820).
The Exynos 8890 doenst keep fps for a long time so the average is lower than Kirin 950 : http://www.frandroid.com/marques/samsung/347416_performances-galaxy-s7-s7-edge-compares-aux-galaxy-s6-huawei-mate-8
The difference in screen res matters a lot there but , on the other hand, they don’t seem to test it long enough to let the Samsungs cook, they should run the games for 30 mins or more,that’s when you see how low they can go.
Real Racing 3 is launched in FHD for both phone. Only little game are launched in QHD but yes gaming test are more interesting than synthetic
You are right they do say the 2 games are at 1080p and it seems that the Hitman score is not lower for the Samsungs because they don’t use antialiasing for some reason.
So uf Frandroid made no mistakes, the Mate 8 crushes the Exynos, with a 3 times smaller GPU.
In pure theory that depends.
You got 3-3.5W that you can use and how you manage those between CPU, GPU and memory can make a big difference. SD820/Eyynos 8890 throttle up to 40-50% in GFXBench long term perf test but that one is not CPU heavy compared to some games.
Using the small cores more and better managing power in general could push a much smaller GPU ahead of a much bigger GPU.
Ofc we don’t know how much the X25 throttles and chances are it will throttle some but in pure theory it is doable.
How much you get out of Smaller A510 with lower frequencies? & performance is on pair with T880 MP4 @850 MHz.
https://gfxbench.com/compare.jsp?benchmark=gfx40&did1=27447856&os1=Android&api1=gl&hwtype1=GPU&hwname1=Qualcomm+Adreno+%28TM%29+510&D2=Meizu+Pro+6
we’ll see how much the x20/25 throttles
Gfx benchmark is rather light on CPU usage it’s also a little on down side when it comes to Adrenos. The comparation you chose for the throttling is totally inadequate as their are no data at all considering battery life & consistency for Xiaomi Note 6 Pro. Something also isn’t right about that Xiaomi Note 6 Pro score as it scores more than HiSilicon Kirin 950 that is 50 MHz higher clocked so that score is not valid.
Now let’s try to make this a bit more fair.
https://gfxbench.com/compare.jsp?benchmark=gfx40&did1=27447856&os1=Android&api1=gl&hwtype1=GPU&hwname1=Qualcomm+Adreno+%28TM%29+510&D2=Huawei+Mate+8+%28NXT-xxx%29
Now keep in mind that X20 is around 30% less power efficient, lower clocked & cetera compared to HiSilicon Kirin 950.
As the number’s show estimated battery life of Galaxy A9 in T-rex is about 47% more wile delivering 16% less performance. Keep in mind that quality of rendering is higher With Mali driver but Overheads are much better with Adreno driver, personally I think overheads are more important for gaming experience.
So the clear winer is A510 even when compared to T880 higher clocked & based on more advanced lithography.
If you have time try to find a 3D mark complete comparation for as similar as possible phones based on S650 to X20 & S652, crucial is a similar battery capacity & Display technology/size, I don’t have time for it & it would give as clear as possible picture.
lol you started the insanity again, couldn’t you just stay banned?
Just stop putting words in my mouth, claiming i said things i never did and more importantly just stop conversing with me.
I whosent banned, just didn’t want to argue with someone prone & ticked as you… Anyhow I don’t have anything from it.
Wouldn’t like to ban you even if I could. Who would I laugh at then?
However what I did do is I had a good laugh at your comments at EE Times but you didn’t know it’s me.
You always try to do that stunt how horrible terrible I am always when something totally argumented with a facts by me doesn’t fit in your litle fantasy & bold claims.
Well fuck you, it’s a free forum! You don’t like it? Well do a something else in your life. Hopefully something you are better at.
Ah and yes the Kirin has LPDDR4 memory whereas X25 has LPDDR3 which is far slower…
not really.
memory perf depends on bandwidth and latency not just bandwidth. DDR4 has higher bandwidth but loses in latency. it also depends on the app, some apps like bandwidth, some like low latency, some don’t care much about memory perf. As DDR4 gets faster, the benefits will increase.
in this particular case, the Kirin has rather low bandwidth for the memory they are using (DDR4 @ 1333MHz), might be that the old interconnect they are using, just can’t do better.
Scroll down to the memory score http://browser.primatelabs.com/geekbench3/compare/5754368?baseline=5753920
GPU’s care just about bandwidth when it comes to it.
GPUs do need BW but there are other factors to consider. T880 does have some neat memory bandwidth saving features, the size of the GPU cache used matters and can help and then, how much memory bandwidth is enough depends a lot on screen res. Hard to say if the GPU here (or in the Kirin) would be held back by the memory BW at 1080p, doesn’t seem so and at 1440p we don’t have any data.
GPU’s simply love a memory bandwidth. You can use Desktop comparation as simple as it is Nv GTX950 have a 128 bit mem bus & smaller bandwidth compared to GTX760 that is 192 bit, even the GTX9xx series have a lot enchanted compression along with improved testilation engine it still in benchmarks for older games that use less testilations isn’t enough for the GTX950 to beat the older GTX760. Think that comparation is transparent enough.
http://www.techspot.com/review/1049-nvidia-geforce-gtx-950/page2.html
lol those 2 have nothing to do with each other, 2 different architectures, different sizes. You compare apples to elephants.
The 950 is Maxwell with 768 shading units, 48 texture mapping units and 32 ROPs.
The 760 is Kepler with 1152 shading units, 96 texture mapping units and 32 ROPs.
If anything you are showing that going with less BW is doable.
Not remotely true, as a high clocked DDR3 has much better latency than a normal clocked DDR4 chip. Memory Latency matters more than clock speeds, but the DDR4 will eventually beat that, when it starts going over 2000Mhz for mobile, but so far that’s not the case.
“For those who know your Meizu phones you will see that the only update here from the MX5 is the chipset” If the operating system is the same then it really is the same phone just a different variant.
This roadmap is incorrect just looking at the Pro 6 specs. Meizu already confirmed they weren’t using the Exnyos.