Author Topic: GPD Q9 Video Review Emulation Android tests.Benchmarks & Tegra 4 comparison! (Read 69234 times)

nielo360 · « **Reply #40 on:** April 15, 2015, 12:54:00 pm »

Quote from: Melquiades on April 15, 2015, 12:49:58 pm

Wow, thanks for the review.

Unfortunately you are scaring the shit out of me. When I receive it I will use it for some weeks and then decide which one should I sell, this or the S7800B...

Dont worry, as all reviews are really subjective, you might love the device and have fewer issues, I would highly recommend looking at Deenox's gpd Q88/Q89 review as he goes into much greater detail about the device's build (they have the same mold)

Melquiades · « **Reply #41 on:** April 15, 2015, 01:04:20 pm »

Of course I have been waiting to get my hands on that DPAD for a year. What pisses me off is the heating. It totally renders a handheld useless. On the other hand, if to avoid heating I have to limit clockspeed and to set a low governor, I fear I will lose the boost in android performace for things like Asphalt, Real Racing,... but also for games like Banner Saga or Limbo. Lets see, I am starting to think I should have bought a Q89.

Taru · « **Reply #42 on:** April 15, 2015, 02:31:41 pm »

I just thought id mention I used to have a gpd q88+ and I found a way to make the analog nubs glide easier, I noticed that when you press on stick from a certain angle it catches the casing so I cut out 2 circular pieces of thin clear plastic from packaging I had laying around then cut a thin key hole shade out of the middle of the circles I made and fit them underneath the nubs above the casing no need to take anything apart just slip em around the tiny stick underneath the flat part of the nub, the plactic pieces need to cover the holes in the casing under the nubs.

procyon · « **Reply #43 on:** April 15, 2015, 03:09:07 pm »

I used my device for nearly 2 hours straight last night, mostly emulators and some casual Android apps. Battery went from 100% to 60% in that time. Either Nielo360 has a particularly defective unit, or he is far more sensitive to the heat than I am. I did notice that the unit gets warm, particularly behind the right side of the device, but in no way did it get to the point of discomfort or even distraction. I only noticed it was warm when I stopped playing whatever I was playing. I think (fortunately or unfortunately, depending on how you look at it) it's entirely subjective, and will only come down to trying it out and seeing if it's beyond your personal threshold.

So far, I'm extremely happy with the device, if for no other reason than purely as a replacement to my jxd 7800, which I constantly felt was fraught with problems. The battery and power supply on the 7800 is a nightmare, and the buttons are loud and/or creaky. The Q9 has none of that. I thought I would care about the decrease in resolution, and I honestly couldn't even really tell the difference while playing games. For some reason, I was convinced that going from an RK3188 to an RK3288 would be like going from a Core i3 to a Core i5, but it's really not. But I didn't really buy the device to take specific advantage of an improved processor, so I can't complain. All I know is that my 7800 better get used to being stuck in a drawer for a long time, or until I decide to sell it...

Edit: I would conclude that nielo's review is accurate for power android gamers, while my perspective is better for gamers more interested in old-school emulation and casual gaming.

eragon2890 · « **Reply #44 on:** April 15, 2015, 03:24:23 pm »

The review puts a huge smile on my face because of the great game performance. On my rk3188 archos gamepad 2, exiles has an outside framerate of about 5 (Seriously) , he says it runs great on this device. Which is also true with other videos and t764 benchmarks. The gaming performance for android games is literally 6 - 8 times higher, and that does not take the lower resolution into account even! I saw video of icestorm benchmark on youtube, where framerate went from about 10, to more then 45 between the two devices. The difference between mali 400 mp4 and t-764 is seriously insane. ANother fact, to indicate the huge gap: triangle performance went from 55 million per second to 1300 million (!).
Ad in the screen resolution and this thing has insane game performance, can be seen in video too, it sends everything from nvidia shield to galaxy s5 and xperia z2 crying to mommy in terms of framerates. Just for fun, can the reviewer (who apparently has an xperia ultra tablet) run some graphic benchmarks on both devices and compare the fps results? I am almost sure this will actually perform (much) better since the GPU slaughters the adreno 330 used in all of 2014 ultra-high-end devices on paper on same resolution devices, and the screen resolution is only half...

Also, a correction on the video: the CPU is not just a speed bump! The 3188 used cortex a9 cores. The 3288 uses a12 (not a17 what they said) cores. A12 cores *At the same clock speed* are about 40% faster than a9 cores. So factoring in the slight bump in clock speed, CPU performance is about 50% higher, not just a little bit, because they also put in a newer architecture!

I am definitely one of the people who can live with the sticks, and the archos became quite hot also, I found it actually felt pretty good once I stopped worrying about wether or not the device could handle it after a few weeks ^_^

In other words: I can't wait till this thing arrives, I bought exiles and modern combat 5 on the play store but couldn't play them at all due to crap graphics performance, whereas here they run at 30 or probably even 60 fps flat on high(est) settings, so hallllllleeeejuuuuhah! I can't wait to actually finally be able to play exiles!

Oh other nice thing: if oyu have 4k screen, the HDMI port can actually output 4k video signal to it! SO you can game in 4k then

The rk3288 also comes with a 4k video decoder for all common formats (I believe 60 fps even). It's a monster in all possible ways. It's *insane*. That's also why it gets a bit hot... the only way for them to prevent that was to include a desktop style fan like the shield portable, bu tthat would kill portability...

I want it. I want it. Goddamnit

I hope my package arrives soon!

Glad to hear Procyon is also happy with it

nielo360 · « **Reply #45 on:** April 15, 2015, 03:28:30 pm »

Quote from: procyon on April 15, 2015, 03:09:07 pm

I used my device for nearly 2 hours straight last night, mostly emulators and some casual Android apps. Battery went from 100% to 60% in that time. Either Nielo360 has a particularly defective unit, or he is far more sensitive to the heat than I am. I did notice that the unit gets warm, particularly behind the right side of the device, but in no way did it get to the point of discomfort or even distraction. I only noticed it was warm when I stopped playing whatever I was playing. I think (fortunately or unfortunately, depending on how you look at it) it's entirely subjective, and will only come down to trying it out and seeing if it's beyond your personal threshold.

So far, I'm extremely happy with the device, if for no other reason than purely as a replacement to my jxd 7800, which I constantly felt was fraught with problems. The battery and power supply on the 7800 is a nightmare, and the buttons are loud and/or creaky. The Q9 has none of that. I thought I would care about the decrease in resolution, and I honestly couldn't even really tell the difference while playing games. For some reason, I was convinced that going from an RK3188 to an RK3288 would be like going from a Core i3 to a Core i5, but it's really not. But I didn't really buy the device to take specific advantage of an improved processor, so I can't complain. All I know is that my 7800 better get used to being stuck in a drawer for a long time, or until I decide to sell it...

Edit: I would conclude that nielo's review is accurate for power android gamers, while my perspective is better for gamers more interested in old-school emulation and casual gaming.

I agree with your conclusion on the battery, Im not sure why its dropping that way, and GPD has done a great job with faces buttons and dpad as well. The heat is purely subjective, my comparison would come solely from the shield portable/gpd g7 where there was no heat or warmth to be felt. The much i5s did get pretty hot though.

eragon2890 · « **Reply #46 on:** April 15, 2015, 03:33:14 pm »

The analogs are accurate as in if you push forward, you go forward, not left right? Because in that case, I luke nubs, and I want the thing to arrive. NOW

(Won't spam anymore. Just excited to see those games I bougth but couldn't play are gonna run at 60 fps

)

nielo360 · « **Reply #47 on:** April 15, 2015, 03:47:15 pm »

Quote from: eragon2890 on April 15, 2015, 03:33:14 pm

The analogs are accurate as in if you push forward, you go forward, not left right? Because in that case, I luke nubs, and I want the thing to arrive. NOW

(Won't spam anymore. Just excited to see those games I bougth but couldn't play are gonna run at 60 fps )

No problem at all, the analogs will work fine for all basic movement.

eragon2890 · « **Reply #48 on:** April 15, 2015, 04:01:52 pm »

Quote from: nielo360 on April 15, 2015, 03:47:15 pm

Quote from: eragon2890 on April 15, 2015, 03:33:14 pm
The analogs are accurate as in if you push forward, you go forward, not left right? Because in that case, I luke nubs, and I want the thing to arrive. NOW

(Won't spam anymore. Just excited to see those games I bougth but couldn't play are gonna run at 60 fps )

No problem at all, the analogs will work fine for all basic movement.

Great! Oh and a tip if you love the dpad so much more: the n64 had no real analog. It could only see 8 directions. So in mupen64, you can just map the dpad's 8 directions to analog stick for games like ocarina of time and super Mario 64. You can then use the awesome dpad instead of analog and lose absolutely nothing compared to a real n64 controller ;-)

acid · « **Reply #49 on:** April 15, 2015, 04:05:30 pm »

Quote from: eragon2890 on April 15, 2015, 04:01:52 pm

Quote from: nielo360 on April 15, 2015, 03:47:15 pm
Quote from: eragon2890 on April 15, 2015, 03:33:14 pm
The analogs are accurate as in if you push forward, you go forward, not left right? Because in that case, I luke nubs, and I want the thing to arrive. NOW

(Won't spam anymore. Just excited to see those games I bougth but couldn't play are gonna run at 60 fps )

No problem at all, the analogs will work fine for all basic movement.

Great! Oh and a tip if you love the dpad so much more: the n64 had no real analog. It could only see 8 directions. So in mupen64, you can just map the dpad's 8 directions to analog stick for games like ocarina of time and super Mario 64. You can then use the awesome dpad instead of analog and lose absolutely nothing compared to a real n64 controller ;-)

Uh I'm sure you're confused because the N64 did have real analog control. Meaning you could mechanically control walking speed and such. That could never translate to a d-pad.

procyon · « **Reply #50 on:** April 15, 2015, 04:19:00 pm »

Quote from: acid on April 15, 2015, 04:05:30 pm

Quote from: eragon2890 on April 15, 2015, 04:01:52 pm
Great! Oh and a tip if you love the dpad so much more: the n64 had no real analog. It could only see 8 directions. So in mupen64, you can just map the dpad's 8 directions to analog stick for games like ocarina of time and super Mario 64. You can then use the awesome dpad instead of analog and lose absolutely nothing compared to a real n64 controller ;-)

Uh I'm sure you're confused because the N64 did have real analog control. Meaning you could mechanically control walking speed and such. That could never translate to a d-pad.

Yeah, where did you come up with that notion? The N64 analog stick is truly analog. The outer edges were simply notched so as to catch the stick in 8 different positions. Presumably Nintendo did that to make the transition from digital to analog easier for players who were unaccustomed to a freer range of motion. But the design that Nintendo used for the internals of the analog stick is pretty much unchanged even today in the DualShock 4 and Xbone controllers.

eragon2890 · « **Reply #51 on:** April 15, 2015, 04:32:55 pm »

Sorry, confused 'same number of directions' with 'same range of freedom'. Stupid me $:-\$

I will just be silent now for a bit and wait for my 7 inch of happiness to arrive XD

So I can finally play exiles and modern combat 5 and run asphalt 8 and all the other cool games at 60 fps....

procyon · « **Reply #52 on:** April 15, 2015, 04:33:57 pm »

I would recommend you also try DomiNations... Not the same type of game, and nothing that will take advantage of the tablet specifically, but a really fun and addictive game none the less.

stpat · « **Reply #53 on:** April 15, 2015, 11:17:42 pm »

To clarify, nielo360 and procyon, where did you guys order from? One of your units came with international firmware installed, the other did not, correct?

midknight · « **Reply #54 on:** April 16, 2015, 07:07:39 am »

for the most part this sounds great i wonder how hard it would be to mod out the problem with this device like maybe glue the tops of the sliders so they don't spin and maybe making some sort of heat shield

nielo360 · « **Reply #55 on:** April 16, 2015, 05:45:23 pm »

I think iv figured out why the Q9 is heating up and loosing battery, it seems that after running a game or emu, the cpu gets stuck at 1.8ghz and doesnt clock down.

No frills cpu reports that with interactive mode cfq. Once a game is loaded the current cpu stays at 1.8ghz, in fact it skips 1.2ghz 1.4ghz and just boosts to 1.8ghz for no reason. Im not sure if this is normal but if anyone has more knowlege about this it could be helpful.

For now forcing the max clock to 1.2ghz helps and reduces heats greatly, (battery still drops and shuts down at 9% or 4%. Im open to any suggestions.

ppsspp runs pretty much the same at 1.2ghz (testing dragons aria)

EDIT: NOPE still gets hot, pretty hot

(Played strangers Wrath at 1.2ghz 27mins)

eragon2890 · « **Reply #56 on:** April 16, 2015, 07:40:39 pm »

Quote from: nielo360 on April 16, 2015, 05:45:23 pm

I think iv figured out why the Q9 is heating up and loosing battery, it seems that after running a game or emu, the cpu gets stuck at 1.8ghz and doesnt clock down.

No frills cpu reports that with interactive mode cfq. Once a game is loaded the current cpu stays at 1.8ghz, in fact it skips 1.2ghz 1.4ghz and just boosts to 1.8ghz for no reason. Im not sure if this is normal but if anyone has more knowlege about this it could be helpful.

For now forcing the max clock to 1.2ghz helps and reduces heats greatly, (battery still drops and shuts down at 9% or 4%. Im open to any suggestions.

ppsspp runs pretty much the same at 1.2ghz (testing dragons aria)

EDIT: NOPE still gets hot, pretty hot (Played strangers Wrath at 1.2ghz 27mins)

That?s probably because during all the ton of gaming I did on the archos gamepad 2 (rk3188) it only got really hot when using the GPU. Strangers wrath obviously uses the GPU a lot. Clocking down the CPU isn?t going to help one bit for that :-/

I always had my gamepad 2 in performance governor, nice and snappy, and it still had couple of hours battery life, so perfectly usable, and didn?t get warm at all almost, untill you used to GPU. I assume it?s the same here (your post indicates so as well). The mali - t764 is the fastest ARM GPU in existence on the planet right now, so it runs hot. It?s faster than some bit older laptop nvidia cards...

More importantly for me, I am guessing it did not show any instability after those 30 minutes right? No artifacts, no lag? Because I noticed, rk3188 devices get hot enough to BBQ on, especially if you also put the charger in while you are playing an intensive 3d game (try it, you can probably cook breakfast egg!) but they never seem to loose any stability whatsoever and even after a year of regularly doing that in hour-long sessions, it doesn?t show any sign of breaking down. Except for that pesky non-rotating-by-design analog stick, but that?s separate issue

nielo360 · « **Reply #57 on:** April 17, 2015, 02:24:56 am »

Quote from: eragon2890 on April 16, 2015, 07:40:39 pm

Quote from: nielo360 on April 16, 2015, 05:45:23 pm
I think iv figured out why the Q9 is heating up and loosing battery, it seems that after running a game or emu, the cpu gets stuck at 1.8ghz and doesnt clock down.

No frills cpu reports that with interactive mode cfq. Once a game is loaded the current cpu stays at 1.8ghz, in fact it skips 1.2ghz 1.4ghz and just boosts to 1.8ghz for no reason. Im not sure if this is normal but if anyone has more knowlege about this it could be helpful.

For now forcing the max clock to 1.2ghz helps and reduces heats greatly, (battery still drops and shuts down at 9% or 4%. Im open to any suggestions.

ppsspp runs pretty much the same at 1.2ghz (testing dragons aria)

EDIT: NOPE still gets hot, pretty hot (Played strangers Wrath at 1.2ghz 27mins)

That?s probably because during all the ton of gaming I did on the archos gamepad 2 (rk3188) it only got really hot when using the GPU. Strangers wrath obviously uses the GPU a lot. Clocking down the CPU isn?t going to help one bit for that :-/

I always had my gamepad 2 in performance governor, nice and snappy, and it still had couple of hours battery life, so perfectly usable, and didn?t get warm at all almost, untill you used to GPU. I assume it?s the same here (your post indicates so as well). The mali - t764 is the fastest ARM GPU in existence on the planet right now, so it runs hot. It?s faster than some bit older laptop nvidia cards...

More importantly for me, I am guessing it did not show any instability after those 30 minutes right? No artifacts, no lag? Because I noticed, rk3188 devices get hot enough to BBQ on, especially if you also put the charger in while you are playing an intensive 3d game (try it, you can probably cook breakfast egg!) but they never seem to loose any stability whatsoever and even after a year of regularly doing that in hour-long sessions, it doesn?t show any sign of breaking down. Except for that pesky non-rotating-by-design analog stick, but that?s separate issue

All games including emulators use gpu to some degree. Sure it will not always get hot, but uncomfortably warm at times. Youv mentioned the rk3188, and while I personally have never felt it get it hot (on gpf g7 overclocked) I think its all about design and placement of the chipset.. In my personal opinion a dedicated gaming device even if it gets hot/warm should never ever be felt by the end user since its a dedicated gaming system and not a phone (in phones its more acceptable) hence I judge them as that.

Its simply poor design decisions that create the problem, not the cpu or gpu as they are doing their job and staying stable. If placed properly on a gaming device the end user would barely notice this stuff.

nielo360 · « **Reply #58 on:** April 17, 2015, 05:14:27 am »

Added some bechmarks taken with antutu and gpu benchmark. In comparison to tegra 4 shield portable even @720p the gpu is quiet weak.

Melquiades · « **Reply #59 on:** April 17, 2015, 06:40:11 am »

The benchmark does not really surprise, it shows a small increase compared to 3188 devices.

Anyway, what governor are you using? At what speed? Also, is it Antutu or Antutu X? May I assume the governor you are using for testing the Shield is the same?

Author Topic: GPD Q9 Video Review Emulation Android tests.Benchmarks & Tegra 4 comparison! (Read 69234 times)

nielo360 (OP)

nielo360 (OP)

nielo360 (OP)

nielo360 (OP)

nielo360 (OP)

nielo360 (OP)