For those of you who say the OS were not fair, take a look at my other video, where I ran the same tests on Raspbian Jessie Lite: czcams.com/video/Xwum6kHczJc/video.htmlm21s The sysbench numbers for that are run on Raspbian Jessie Lite. Raspbian Jessie runs 32-bit on all Pis to ensure backward compatibility. Running Ubuntu Mate allows you to use the 64-bit capabilities of the processor. When I ran these tests, I had no HDMI plugged in, so the OS would not run any GUI resources and waste CPU cycles on that. Each score also was from 5 trials, each taken 5 minutes between to ensure there was no effect of thermal throttling.
QuickPi see benchmark wise the ripi has been beaten at a similar price point multiple times, that's not why the ripi is still for some (like me) a better purchase because the ripi has so much more support so if I need a nes case or a 3D print I can be sure with the ripi that I will get it, also with the orangepi and other boards I almost bought I found out the GPIO didn't even work?! Anyway if you need power I think the rock is best if you don't mind not having %100 confidence that if you run into a problem some can help.
+OPGameZ if you are doing server based applications, you need a better processor to handle loads. If you are doing something with GPIO or etc, I definitely suggest Raspberry Pi. I did also state this in the conclusion. Beginners and newbies should stick to the Pi to learn the overall Linux and learn bash. People with good experience and knowledge can keep moving on to better boards like the Rock, OrangePi, etc.
can you use the EMMC for primary boot and still use the SD card for storage? and how do you get a boot image onto the EMMC ? i HATE the rpi 3 (all r pi models actually ,i tried the zero and zero w, same exact issues) it is worthless it corrupts my card ever 2 or 3 boots sometimes even after 1 reboot the card is corrupted sometimes on initial boot it corrupts the rpi 3 is useless garbage i do not recommend it ,it is trash i am looking to get a board with EMMC because r pi boards are worthless garage that corrupt SD's constantly might as well take your money and flush it directly down the toilet
Before the smartphone era DC-barrel-plugs were the industri standard - so calling it proprietary made me giggle :-) Micro-USB was NOT designed with main priority to delivering power, but DC-barrel-plugs were. Also the USB-3 spec demands the host can deliver 1.5 A, the USB-1&2 only demands 0.5 A, which means the Rock64 needs 2.5 A headroom for 3 USB ports against the 2 A headroom for the Raspberry Pi's 4 USB ports. Personally I am VERY happy with the choise of a DC-barrel-plug on the Rock64 - micro-USB is sub optimal for 1 A and above.
the barrel jack for 5v is in no way a down side, I wish the Pi had a barrel jack because uUSB connectors are not good at all for carrying above 1A of current without introducing voltage drop or in the worst case over heating due to the tiny pins.
a barrel jack power supply is an upgrade over the pi. Who isn't sick of seeing that insufficient power indicator in Raspbian. and trying to power anything with the usb ports on a pi was a lost cause.
+QuickPi - doesn't seem to be a fore-gone conclusion, however my RPi3 kit was from Element14 and the adapter is 2.5A, 5.1V I've searched some forums for discussions on this but haven't found a final definitive response...just many opinions. Couple of recurring points seems to be regarding number of attached peripherals and voltage drops (including through cables), and the question of over-clocking.
I have the Rock64 (4GB and a 16GB emmc), so just a few points I have found from your video. Power supply, I currently have it running from a 2.1A supply, with a 4TB USB3 HD attached and get no problems. I bought a couple of cheap USB to barrel connector leads from ebay, ran perfectly well from a cheap USB power bank, and from a 1Amp supply. Using a barrel connector means that the jack is much more secure in place than using a flimsy micro USB connector (how many broken ones do we have?) . Booting up with nothing attached, I got a maximum power draw of about .5A. So, your comments about the power supply being a limiting factor are a bit wrong as it is a common connector, and very common power supply. The 3A is the rating for the maximum it can handle (actually the max is 3.5A), allowing the Rock64 to power attached USB devices, you note how the Pi3 wants you to use a powered hub for attaching items. On the forums people are reporting having 3 USB discs attached with no problem. I am using mine as a network server running SAMBA, owncloud and minidlna, taking over from an Orange Pi Plus (which wimps out with 3 discs attached due to the 2.1A limit meaning I need a powered hub.) That impressed with the Rock64 , just ordered another one.
+stealth banana wow that’s impressive, I just switched my whole home Pi to a Rock64 and seeing a lot of improvements. I might switch to an eMMC module just to ensure there is no downtime from SD card enduarance
Just remember that not everything is running correctly on the Rock64 yet, some people reporting really bad performance on some distros, but headless ubuntu minimal works fine for me. I did have the USB ports power getting switched off after logging into a GUI, but that is solved now as far as I know. I did not do any speed tests of the emmc v sd card, but it runs fast enough for what I want. I did order a 64GB emmc, and swapped it in my pinebook and that is nice and nippy.
+issamu2k I’m assuming the direct USB, it has 3 USB ports. If he’s trying to get USB 3.0 out of all drives, he could be using a USB 3.0 hub to get max data speeds
Andrew's point is valid. If you are doing benchmarks, you should use the same OS, and the 10x difference can't be explained by the better hardware. You are comparing a fine-tuned OS from the maker of one board (Pine's flavor of ARMbian) with a generic distribution ported to the Pi. It would have been fairer to use Raspbian or even a generic ARMbian on the Pi, even if it was 32-bit. Running a 64-bit OS on a system with less than 4 GB of RAM does not give any real advantage (you can address a lot of memory, but you will crash into swap long before filling the 32-bit 4 GB address space). Of course, the Rock64 is more powerful than the Pi 3. But ten times? The processor is based on the same core (A53) at 50% greater speed (1.8 vs. 1.2 GHz), and the faster RAM would only give a moderate advantage (say 30-50%) in memory-intensive operations. All in all, I'd guess the Rock64 is about twice as fast as the Pi 3, not ten times. There has to be a hidden problem with the benchmark (maybe the versions of Sysbench for Pine's Debian and Ubuntu are compiled with different optimization flags, or with different versions of GCC).
My guess, RPi 3 bottlenecked with ARMv6 code in ARMv7 mode vs Rock64 native ARMv8. See: forum.armbian.com/index.php?/topic/751-rfc-support-cortex-a53arm64/&do=findComment&comment=12462
Something that might be useful to note is the 2nd GPIO strip includes headers for a second Ethernet port @ 10/100. Useful if you want to make a vpn tunnel Router as it can be hard wired rather than depending on the slow RPi WiFi in addition to the onboard Ethernet. If you really want decent WiFi then you have to add an external one anyway. Also the RPi shares the usb bus for it’s Ethernet connection so would be interested in a comparison of throughput when using usb devices (maybe an external HDD or SSD).
Thanks for that. Your comments really helped and perhaps I might look at this for Plex as either a server or client, though clients are available for most devices now so a low cost portable server might be a good use with the USB3 for a large Media drive. The 1GB ethernet looks good too!
Has anyone completed any Ethernet speed testing with Rock64 board? I'm curious how much of the 1Gbps it will give you on a bandwidth test. Great video!!! Nice job
i run as low as 2GB card and even then it gets corrupted after every single boot no matter what card i use what size or brand and no matter what power supply i use the pi's corrupt the cards i tried 3 amp wall warts , 2 amps wall warts , a PC power supply from and old 486 used the 5v rail and tried 8 different SD cards , 2 different pi 3 model b's , 5 different pi zero 1.3 's and 3 pi w 's, all of these boards suck and will corrupt any card after just a few boots , how did the pi get so popular it is an utterly worthless piece of shit ,you need to reburn the OS and start over from scratch practically after every single reboot usually around 2 or 3 boots and your card is fucked up ,what fuckkng garbage these things are i am going to just break down and buya board that has EMMC now, fuck this pi shit
Do you know if any of the hat boards such as those from Suptronics will work with the Pine64? It looks at first glance that the GPIO pass through is in the same location as the Pi..
that connector they're using for the eMMC should really be made standard on these things. what's this connector actually called, I can never find them. also, it'd make more sense if the eMMC Enable jumper were integrated into that connector.
EMMC should be made standard period my pi need to have the image reburned after every boot almost SD sucks all these pi boards suck and constantly corrupt SD images with out EMMC they are worthless, then again i only tried r pi boards they are all garbage ,every single model , maybe other SBC's that boot from SD are better do not have that problem but with EMMC you are not going to have to worry about reburning images constantly
performance looks promising. But what about NAS-performance? How fast and reliable can the rock64 reads data from usb3 and deliver it versa ethernet? I had a disappointing experience with the odroid-c1, which was only capable of delivering a steady stream of 8 mb/sec from a connected usb2-hdd. I was using OpenMediaVault. A benchmark in this regards is much appreciated.
ucity metalhead haven't used it myself but I have used the Raspberry Pi. The Raspberry Pi is really good for emulation but kind of sucks when it comes to Nintendo 64 games. The pine64 might perform slightly better though because of its specs but you're going to have to deal with the lack of Internet unless you're using ethernet. Also your resolution possibility would be higher because it's able to Output 4K with the pine but I would just stick to the Raspberry Pi because most of the software out there runs fine for it and there's a huge community of people that could help with any issues you might have with the pi but if you can figure out a way to enable wireless on the pine64 I'd say give it a try
I have bought a second rock64 just for gaming, not had time to set it up yet. On the Pine64 community forums, people have been reporting it running significantly faster on emulation on Android, and on Linux than the Pi3 runs. Some emulators are not available in 64 bit (yet). N64 is supposed to run nigh on flawlessly on an Android emulation, as does PSX.
If it only had Wifi, for portability, it would be INCREDIBLE, especially considering the comment posted here about using a USB power supply with a common connector, given that the 3A (or 3.5...) is the maximum, not the minimum. The power is very impressive for the price.
I was able to get a 128gb micro SD card to work with pi 3 but I think results very depending on brand and make. The Rock 64 looks to be a great board but would really like it too have at least Bluetooth for wireless gaming.
The 1GB Rock64 is on Amazon right now for $16.99 bundled with a 5V 3A power supply, a Sandisk Ultra 16GB TF card and a 16x2 line LCD display. www.amazon.com/gp/product/B0868WSTXH/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1
These are good benchmarks but like every single rock64 comparison video the youtuber fails to mention weather or not they used emmc or microsd on the rock64. There is a huge difference in process time between the two cards.
Hey bro I saw you videos and I want ask you if you can make a video installing and configuring the broadlink mr with the plugin for Homebridge I can’t configure that if you can help with the configure.json for that plugin
I’ve been using my Rock64 for transcoding. Been doing a better job than my Pi ever did. The CPU is just more beefy. Just add a good heatsink and a small fan and you should be fine.
Hy i got me the raspberry pi 3 now i dont have internet or wifi in my home and really i want to get on line do u think using wifi from my phone it work using the hotspot.?
Do you think you could do a "Best Replacement For A Rasberry Pi" taking the title literally? I don't mean which SBC is a better SBC over all. I mean which SBC is the best and most powerful 100% software and hardware (that connector with all the pins on it) compatible replacement. IE: A Rasberry Pi on steroids. lol Thank you.
OK thanks. I was confused because of the storage type recommended at the end. I thought if using something as memory then it would fundametely affect but I guess it was not using it like ram. Thanks again
If you used Raspbian Jessie instead of Ubuntu Mate, the 10X difference would come down to a 1.5X difference. There is a video somewhere on youtube, where a guy compared performance of a Pi running Raspbian Jessie and Ubuntu Mate, and Mate was like 8 times slower.
I don't know what Ubuntu did but I switched to Mint because each Ubuntu release got slower on all the machines I have. Debian of course is faster so I think maybe Ubuntu has fallen into the Windows Vista trap of just adding bloat for no useful purpose :)
Lost interest after seeing lack of Wifi and BT. Not to mention losing a USB port for no good reason. Its 2018 and there is no excuse for not having WiFi and BT on the main board.
I have tried it with Debian Jessie Lite and the sysbench score does not change. To see the sysbench results with Raspbian, here is a link to my other video where I did similar geekbench tests: czcams.com/video/Xwum6kHczJc/video.htmlm21s
This might be the sloppiest review/comparison video Ive seen to date. It takes about 10 min reading through the spec sheets of the two SOCs in question to determine they are within 30% of each other in terms of processing power and 2 min of Googling why Sysbench gives such different results
10:47 not same- ubuntu suck. why not use debian on rpi 3 too. to make more fun. make test not other way use ubuntu on rock64 lol WOOPS anyone did 5seconds lol
If you can use the same OS on each board, then use the same OS on each board. Otherwise, your data is askew and should be junked. Accessory boards to the Raspberry Pi 3 are not "backwards" compatible to the Rock64, they are compatible to the Rock64. Some of the Accessory boards to the Raspberry Pi 3 are backwards compatible to the Raspberry Pi 2.
The power adapter is junk, or the one I got. If the cord or board moved just a bit it would lose connection and reboot when it did. I ended up getting a barrel x USB-A cable to cure this ill. I have the 4GB which performs well as a portable desktop.
For beginners?! Raspberry Pi is used by everyone from beginners to experts. It's the commumity support that allows beginners to use it with ease as well as those willing to "put in the work" to complete mpre advanced projects. You didn't even mention what other capabilities the Rock 64 has other than using it as a media server. I couldn't even find it for the low price you quoted and that was starting at 2GB and before adding wifi and BT which comes standard on the Raspberry Pi. I'm not convinced.
There's something very wrong with your benchmarks. There's no way the 600 Mhz difference in CPU frequency between the two boards would lead to a 10x difference in performance.
I have tested multiple programs and I am seeing vastly better performance. I transcode video for HomeKit, my previous delay in the stream was approximately 20-30 seconds with the raspberry pi. With the Rock64, the delay in the stream is just under a second and I have no skipped frames. This is just one example. I have tested multiple Raspberry Pi and Rock64 to ensure my results aren't out of the ordinary...
It's not just the 600MHz CPU clock frequency that makes the difference. Note that the RAM on the Rock64 is SDDR3 @ 1600MHz. However, the pertinent information here would be: What is the sysbench command line that was used to generate these results?
This was hardly a scientific comparison. To make an accurate speed assessment, ALL variables must be the same. Same OS on both, same amount of RAM, etc. If this can't be done, then general comparison benchmarks are basically useless: apples to oranges. I didn't buy into the speed disparity right off the bat because different flavors of OS may contain different overhead processing that is unaccounted for in a general head-to-head comparison. As such, your tests cannot actually say that the hardware of the one is faster than the other, because it could actually be software that accounts for much of the difference. Please do a video that compares setups that are as completely matched and measured with appropriate benchmarks as precisely as you can. You haven't done that here.
Yeah...I was a backer for Pine64....I will never give money to a company that uses the community for it's entire software support, again....nope not for me!
For those of you who say the OS were not fair, take a look at my other video, where I ran the same tests on Raspbian Jessie Lite: czcams.com/video/Xwum6kHczJc/video.htmlm21s
The sysbench numbers for that are run on Raspbian Jessie Lite. Raspbian Jessie runs 32-bit on all Pis to ensure backward compatibility. Running Ubuntu Mate allows you to use the 64-bit capabilities of the processor. When I ran these tests, I had no HDMI plugged in, so the OS would not run any GUI resources and waste CPU cycles on that. Each score also was from 5 trials, each taken 5 minutes between to ensure there was no effect of thermal throttling.
QuickPi see benchmark wise the ripi has been beaten at a similar price point multiple times, that's not why the ripi is still for some (like me) a better purchase because the ripi has so much more support so if I need a nes case or a 3D print I can be sure with the ripi that I will get it, also with the orangepi and other boards I almost bought I found out the GPIO didn't even work?! Anyway if you need power I think the rock is best if you don't mind not having %100 confidence that if you run into a problem some can help.
+OPGameZ if you are doing server based applications, you need a better processor to handle loads. If you are doing something with GPIO or etc, I definitely suggest Raspberry Pi. I did also state this in the conclusion. Beginners and newbies should stick to the Pi to learn the overall Linux and learn bash. People with good experience and knowledge can keep moving on to better boards like the Rock, OrangePi, etc.
QuickPi Agreed (mostly)
be nice to actually see a running video
can you use the EMMC for primary boot and still use the SD card for storage?
and how do you get a boot image onto the EMMC ? i HATE the rpi 3 (all r pi models actually ,i tried the zero and zero w, same exact issues) it is worthless it corrupts my card ever 2 or 3 boots sometimes even after 1 reboot the card is corrupted sometimes on initial boot it corrupts the rpi 3 is useless garbage i do not recommend it ,it is trash i am looking to get a board with EMMC because r pi boards are worthless garage that corrupt SD's constantly might as well take your money and flush it directly down the toilet
Before the smartphone era DC-barrel-plugs were the industri standard - so calling it proprietary made me giggle :-)
Micro-USB was NOT designed with main priority to delivering power, but DC-barrel-plugs were. Also the USB-3 spec demands the host can deliver 1.5 A, the USB-1&2 only demands 0.5 A, which means the Rock64 needs 2.5 A headroom for 3 USB ports against the 2 A headroom for the Raspberry Pi's 4 USB ports.
Personally I am VERY happy with the choise of a DC-barrel-plug on the Rock64 - micro-USB is sub optimal for 1 A and above.
the barrel jack for 5v is in no way a down side, I wish the Pi had a barrel jack because uUSB connectors are not good at all for carrying above 1A of current without introducing voltage drop or in the worst case over heating due to the tiny pins.
a barrel jack power supply is an upgrade over the pi. Who isn't sick of seeing that insufficient power indicator in Raspbian. and trying to power anything with the usb ports on a pi was a lost cause.
I was sick of that indicator until I bought a good quality power supply, and never saw it again.
Indeed, seems the RPi3 requires a 5.1v supply - apparently that 0.1v makes all the difference.
+George Kaimakis interesting... never seen that before
+QuickPi - doesn't seem to be a fore-gone conclusion, however my RPi3 kit was from Element14 and the adapter is 2.5A, 5.1V
I've searched some forums for discussions on this but haven't found a final definitive response...just many opinions.
Couple of recurring points seems to be regarding number of attached peripherals and voltage drops (including through cables), and the question of over-clocking.
I have no problems with mine
I have the Rock64 (4GB and a 16GB emmc), so just a few points I have found from your video. Power supply, I currently have it running from a 2.1A supply, with a 4TB USB3 HD attached and get no problems. I bought a couple of cheap USB to barrel connector leads from ebay, ran perfectly well from a cheap USB power bank, and from a 1Amp supply. Using a barrel connector means that the jack is much more secure in place than using a flimsy micro USB connector (how many broken ones do we have?) . Booting up with nothing attached, I got a maximum power draw of about .5A. So, your comments about the power supply being a limiting factor are a bit wrong as it is a common connector, and very common power supply. The 3A is the rating for the maximum it can handle (actually the max is 3.5A), allowing the Rock64 to power attached USB devices, you note how the Pi3 wants you to use a powered hub for attaching items. On the forums people are reporting having 3 USB discs attached with no problem.
I am using mine as a network server running SAMBA, owncloud and minidlna, taking over from an Orange Pi Plus (which wimps out with 3 discs attached due to the 2.1A limit meaning I need a powered hub.) That impressed with the Rock64 , just ordered another one.
+stealth banana wow that’s impressive, I just switched my whole home Pi to a Rock64 and seeing a lot of improvements. I might switch to an eMMC module just to ensure there is no downtime from SD card enduarance
Just remember that not everything is running correctly on the Rock64 yet, some people reporting really bad performance on some distros, but headless ubuntu minimal works fine for me. I did have the USB ports power getting switched off after logging into a GUI, but that is solved now as far as I know. I did not do any speed tests of the emmc v sd card, but it runs fast enough for what I want. I did order a 64GB emmc, and swapped it in my pinebook and that is nice and nippy.
+stealth banana I’m on stock Debian and haven’t encountered any problems so far so I’m happy with it
stealth banana what are you using to plug the hds on USB?
+issamu2k I’m assuming the direct USB, it has 3 USB ports. If he’s trying to get USB 3.0 out of all drives, he could be using a USB 3.0 hub to get max data speeds
why did you not use Raspian on your review? this is also a Debian Jessie variant and would have been more equal than Ubuntu Mate
+Andrew Hayes the Ubuntu mate is the current distribution that is pure 64-bit only, giving higher performance.
Andrew's point is valid. If you are doing benchmarks, you should use the same OS, and the 10x difference can't be explained by the better hardware.
You are comparing a fine-tuned OS from the maker of one board (Pine's flavor of ARMbian) with a generic distribution ported to the Pi. It would have been fairer to use Raspbian or even a generic ARMbian on the Pi, even if it was 32-bit. Running a 64-bit OS on a system with less than 4 GB of RAM does not give any real advantage (you can address a lot of memory, but you will crash into swap long before filling the 32-bit 4 GB address space).
Of course, the Rock64 is more powerful than the Pi 3. But ten times? The processor is based on the same core (A53) at 50% greater speed (1.8 vs. 1.2 GHz), and the faster RAM would only give a moderate advantage (say 30-50%) in memory-intensive operations. All in all, I'd guess the Rock64 is about twice as fast as the Pi 3, not ten times. There has to be a hidden problem with the benchmark (maybe the versions of Sysbench for Pine's Debian and Ubuntu are compiled with different optimization flags, or with different versions of GCC).
My guess, RPi 3 bottlenecked with ARMv6 code in ARMv7 mode vs Rock64 native ARMv8. See: forum.armbian.com/index.php?/topic/751-rfc-support-cortex-a53arm64/&do=findComment&comment=12462
Something that might be useful to note is the 2nd GPIO strip includes headers for a second Ethernet port @ 10/100. Useful if you want to make a vpn tunnel Router as it can be hard wired rather than depending on the slow RPi WiFi in addition to the onboard Ethernet. If you really want decent WiFi then you have to add an external one anyway. Also the RPi shares the usb bus for it’s Ethernet connection so would be interested in a comparison of throughput when using usb devices (maybe an external HDD or SSD).
Not sure where the 64GB limitation info for the Pi came from. I'm running Octoprint with a 128GB MicroSD.
Very fair review thanks for see the pros and cons of both.
Thanks for that. Your comments really helped and perhaps I might look at this for Plex as either a server or client, though clients are available for most devices now so a low cost portable server might be a good use with the USB3 for a large Media drive. The 1GB ethernet looks good too!
Has anyone completed any Ethernet speed testing with Rock64 board? I'm curious how much of the 1Gbps it will give you on a bandwidth test. Great video!!! Nice job
I ran a 128 GB card on the Raspberry Pi 3.
Yes it does work, but it is not 'officially' supported
i run as low as 2GB card and even then it gets corrupted after every single boot no matter what card i use what size or brand and no matter what power supply i use the pi's corrupt the cards i tried 3 amp wall warts , 2 amps wall warts , a PC power supply from and old 486 used the 5v rail and tried 8 different SD cards , 2 different pi 3 model b's , 5 different pi zero 1.3 's and 3 pi w 's, all of these boards suck and will corrupt any card after just a few boots , how did the pi get so popular it is an utterly worthless piece of shit ,you need to reburn the OS and start over from scratch practically after every single reboot usually around 2 or 3 boots and your card is fucked up ,what fuckkng garbage these things are i am going to just break down and buya board that has EMMC now, fuck this pi shit
Do you know if any of the hat boards such as those from Suptronics will work with the Pine64? It looks at first glance that the GPIO pass through is in the same location as the Pi..
that connector they're using for the eMMC should really be made standard on these things. what's this connector actually called, I can never find them. also, it'd make more sense if the eMMC Enable jumper were integrated into that connector.
Sarreq Teryx sorry late reply but its called an emmc to microsd adapter. You can get it at ameridroid dot com for $4
EMMC should be made standard period
my pi need to have the image reburned after every boot almost
SD sucks all these pi boards suck and constantly corrupt SD images with out EMMC they are worthless, then again i only tried r pi boards they are all garbage ,every single model , maybe other SBC's that boot from SD are better do not have that problem but with EMMC you are not going to have to worry about reburning images constantly
Good analysis. Any guess why Rockchip is better in implementation than Broadcom. I though Broadcom should be a natural good implementer of ARM chips.
Are this good for a 64 slot Minecraft, or a 80 slot GMod Pedobear= murder server?
How about a comparison between the Rock 64 and the Asus Tinker Board ?
I wish I could purchase the RK3399 model of this, but it's not available.
performance looks promising. But what about NAS-performance? How fast and reliable can the rock64 reads data from usb3 and deliver it versa ethernet? I had a disappointing experience with the odroid-c1, which was only capable of delivering a steady stream of 8 mb/sec from a connected usb2-hdd. I was using OpenMediaVault. A benchmark in this regards is much appreciated.
I've seen another video stating it can saturate the gigabit Lan at 100 MBps transfers.. I just ordered one for just this
What about game emulation how well does that work on the rock 64?
ucity metalhead haven't used it myself but I have used the Raspberry Pi. The Raspberry Pi is really good for emulation but kind of sucks when it comes to Nintendo 64 games. The pine64 might perform slightly better though because of its specs but you're going to have to deal with the lack of Internet unless you're using ethernet. Also your resolution possibility would be higher because it's able to Output 4K with the pine but I would just stick to the Raspberry Pi because most of the software out there runs fine for it and there's a huge community of people that could help with any issues you might have with the pi but if you can figure out a way to enable wireless on the pine64 I'd say give it a try
Sourcefed Nerd yeah and being a newbie to these things the pi would be a better choice for me too.
I have bought a second rock64 just for gaming, not had time to set it up yet. On the Pine64 community forums, people have been reporting it running significantly faster on emulation on Android, and on Linux than the Pi3 runs. Some emulators are not available in 64 bit (yet). N64 is supposed to run nigh on flawlessly on an Android emulation, as does PSX.
Why no links for the power supply and emmc for the rock pi?
Thank you for the information!
What about power consumption and heat between both boards? Thanks!
hey can you upload SHOWDOWN videos on ( RaspberryPi 3B+ Vs Rock64, & TinkerBoard S Vs RaspberryPi 3B+ & RaspberryPi 3B+ Vs RaspberryPi 3)
Sure
Thanks......Can you tell me the date when you are going to upload that #showdown video...
At 4:45 audio and microphone? That's news to me. Is this a Pi3 specific feature I never received the memo for?
It can support audio and mic thru the 3.5mm jack
It’s been there for a while and you just have to change it in video settings, from composite to audio and mic
Then I shall defer to your superior knowledge on the subject.
hey, give a compare of Raspberry pi 3B+ Vs Rock64 4GB.... It will help me to choose one of it for my project....
make comparison of raspberry pi 4B and Rock pro
You did not include the cost of a USB wifi adapter which is standard on the Pi3+, which can also boot from the USB bus.
Can it run retropie?
Yes.
If it only had Wifi, for portability, it would be INCREDIBLE, especially considering the comment posted here about using a USB power supply with a common connector, given that the 3A (or 3.5...) is the maximum, not the minimum. The power is very impressive for the price.
I was able to get a 128gb micro SD card to work with pi 3 but I think results very depending on brand and make. The Rock 64 looks to be a great board but would really like it too have at least Bluetooth for wireless gaming.
Excellent presentation, thanks!
Great video! How about comparing LattePanda with Rock64...
Maybe if someone can sponser that haha
no mic in on the 3.5mm on rasbpi
nice info, please compare opencv on the Rock64 and Pi3. thanks.
should have had a USB-C port, even if it just powered it.
The 1GB Rock64 is on Amazon right now for $16.99 bundled with a 5V 3A power supply, a Sandisk Ultra 16GB TF card and a 16x2 line LCD display. www.amazon.com/gp/product/B0868WSTXH/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1
Are u be able to play sega Saturn, psp, sega dreamcast, N64, ps2, and GameCube
yes if you install retropie on the Rock64
will the rock run retro pi??
Power source makes the Rock64 loose on portability. I can run Rasp Pie from a pocket juice.
These are good benchmarks but like every single rock64 comparison video the youtuber fails to mention weather or not they used emmc or microsd on the rock64. There is a huge difference in process time between the two cards.
SD card on both devices. Same SD card model on both as well. Sorry for missing out the detail.
Hey bro I saw you videos and I want ask you if you can make a video installing and configuring the broadlink mr with the plugin for Homebridge I can’t configure that if you can help with the configure.json for that plugin
+Sergio Guadron yes it’s a plan
So which one would give the best performance as a Plex/Kodi/ media server? Output would be to a 1080 HDMI 52-56?
Thanks
I’ve been using my Rock64 for transcoding. Been doing a better job than my Pi ever did. The CPU is just more beefy. Just add a good heatsink and a small fan and you should be fine.
QuickPi Thanks for the info. I'll be looking in to it.
Hy i got me the raspberry pi 3 now i dont have internet or wifi in my home and really i want to get on line do u think using wifi from my phone it work using the hotspot.?
Yes that should work
Now install Imica in it and you have small scale Amiga that is really Fast.
ARM performance depends on RAM performance. DDR3 instead of DDR2 makes all the difference.
whats the rock64 like for cooling?
I suggest installing a heatsink of somekind. It definitely helps with cooling and decreases the chance of thermal throttling.
The power supply input is much better as it correct power adapter not stupid USB for phones
Do you think you could do a "Best Replacement For A Rasberry Pi" taking the title literally? I don't mean which SBC is a better SBC over all. I mean which SBC is the best and most powerful 100% software and hardware (that connector with all the pins on it) compatible replacement. IE: A Rasberry Pi on steroids. lol
Thank you.
I think the main reason for the drastic speed difference is the memory used, should of tested with sd cards in both though to check that out
Both boards had the same SD card. Also the sysbench tests rely on pure CPU computing power, nothing storage based.
OK thanks. I was confused because of the storage type recommended at the end. I thought if using something as memory then it would fundametely affect but I guess it was not using it like ram. Thanks again
Woooow! Great!!! Thanks!!!
i like my pi3 for retrogaming... if something runs good for what you do why change ;)
Have you even tried retro gaming on the Rock64?
If you used Raspbian Jessie instead of Ubuntu Mate, the 10X difference would come down to a 1.5X difference. There is a video somewhere on youtube, where a guy compared performance of a Pi running Raspbian Jessie and Ubuntu Mate, and Mate was like 8 times slower.
I don't know what Ubuntu did but I switched to Mint because each Ubuntu release got slower on all the machines I have. Debian of course is faster so I think maybe Ubuntu has fallen into the Windows Vista trap of just adding bloat for no useful purpose :)
Do the chip vs raspberry pi 3 plz
+WolfMan YT I already have a CHIP vs Raspberry Pi Zero video...
QuickPi raspberry pi 3 i want to know
+WolfMan YT not really a big difference and also pi3 and chip are different league SBCs. One is 9 bucks while the other is $35
QuickPi ok thx
Lost interest after seeing lack of Wifi and BT. Not to mention losing a USB port for no good reason.
Its 2018 and there is no excuse for not having WiFi and BT on the main board.
UFOhunter its a security feature not to have wireless
Make another video of this and put Debian Jessie Lite on both of them and see the performance of the Raspberry PI 3.
I have tried it with Debian Jessie Lite and the sysbench score does not change. To see the sysbench results with Raspbian, here is a link to my other video where I did similar geekbench tests: czcams.com/video/Xwum6kHczJc/video.htmlm21s
take a look at my pinned comment...
This might be the sloppiest review/comparison video Ive seen to date. It takes about 10 min reading through the spec sheets of the two SOCs in question to determine they are within 30% of each other in terms of processing power and 2 min of Googling why Sysbench gives such different results
10:47 not same- ubuntu suck. why not use debian on rpi 3 too. to make more fun. make test not other way use ubuntu on rock64 lol WOOPS anyone did 5seconds lol
If you can use the same OS on each board, then use the same OS on each board. Otherwise, your data is askew and should be junked.
Accessory boards to the Raspberry Pi 3 are not "backwards" compatible to the Rock64, they are compatible to the Rock64. Some of the Accessory boards to the Raspberry Pi 3 are backwards compatible to the Raspberry Pi 2.
As well as.....
The power adapter is junk, or the one I got. If the cord or board moved just a bit it would lose connection and reboot when it did. I ended up getting a barrel x USB-A cable to cure this ill. I have the 4GB which performs well as a portable desktop.
you voice is so light, its honestly quite annoying but good video
his mother was a hobbit
For beginners?! Raspberry Pi is used by everyone from beginners to experts. It's the commumity support that allows beginners to use it with ease as well as those willing to "put in the work" to complete mpre advanced projects. You didn't even mention what other capabilities the Rock 64 has other than using it as a media server. I couldn't even find it for the low price you quoted and that was starting at 2GB and before adding wifi and BT which comes standard on the Raspberry Pi. I'm not convinced.
There's something very wrong with your benchmarks. There's no way the 600 Mhz difference in CPU frequency between the two boards would lead to a 10x difference in performance.
I have tested multiple programs and I am seeing vastly better performance. I transcode video for HomeKit, my previous delay in the stream was approximately 20-30 seconds with the raspberry pi. With the Rock64, the delay in the stream is just under a second and I have no skipped frames. This is just one example. I have tested multiple Raspberry Pi and Rock64 to ensure my results aren't out of the ordinary...
It's not just the 600MHz CPU clock frequency that makes the difference. Note that the RAM on the Rock64 is SDDR3 @ 1600MHz.
However, the pertinent information here would be: What is the sysbench command line that was used to generate these results?
This was hardly a scientific comparison. To make an accurate speed assessment, ALL variables must be the same. Same OS on both, same amount of RAM, etc. If this can't be done, then general comparison benchmarks are basically useless: apples to oranges. I didn't buy into the speed disparity right off the bat because different flavors of OS may contain different overhead processing that is unaccounted for in a general head-to-head comparison. As such, your tests cannot actually say that the hardware of the one is faster than the other, because it could actually be software that accounts for much of the difference.
Please do a video that compares setups that are as completely matched and measured with appropriate benchmarks as precisely as you can. You haven't done that here.
Yeah...I was a backer for Pine64....I will never give money to a company that uses the community for it's entire software support, again....nope not for me!