The ARM system is not really more efficient in tests such as gimp, kdenlive, web basemark and probably glmark as well because whilst it uses ~60% of the power it's performance is sometimes over 40% slower. It's probably more efficient in the pure number crunching tests such as sysbench CPU but those tests are the least likely to be representative of real world use. I also suspect from the icons on the left panel that the standard ubuntu installation might have more programmes installed than the Rockchip specific ubuntu installation.
The RK3588 definitely has its use cases, especially on low power applications, mobile solutions, or where the system is being powered by solar/battery.
@@pikaskew I can also confirm that I bought a Chuwi Mini-PC with 12 GB of RAM and a 512 GB NVMe N100 for 129 dollars, and it is simply fantastic. Previously, I had a Raspberry Pi 5, which was too slow for me and had poor compatibility with certain containers that were x86-based. Now I'm using this little powerhouse, and it is so much better than the Raspberry Pi 5. I was even able to install QEMU and run Windows 11 in a VM under Ubuntu Linux. It’s really great, man. The N100 impresses. Moreover, it is extremely efficient: at maximum load, it only consumes 21 watts, and in normal operation, it’s even just 9 watts.
The biggest takeaway for me, is the fact that there is an Ubuntu that works very nicely with Rock chips. That makes the Rock 5 B look attractive again.
Great comparison. I'm always amazed by my n-100 rig, the thing never gets too hot. That's one great board, even with a little more wattage. Thanks Chris!
Another great video Chris. It is interesting to see how fat the RISC processors have come. I have an older Intel N5105 Brick PC which is a older and a little slower than the N100, but seems fine for basic tasks and web browsing. All in all, an interesting comparison. Thanks for all your hard work Chris!
It is actually amazing to see how 7W of higher energy consumption can lead to such an increase in performance. Having a 22W PC capable of 4K streaming is actually quite impressive. I am currently experimenting with a Fujitsu Futro S520, which has a 7W total power consumption and is capable of 720p @ 60 fps video streaming outside of browsers via the mpv. I realized that using my desktop PC (better berformance although old) with its 160W of power consumption is kind of wasted if used for browsing and videostreaming.
I wouldn't be surprised if for the 4K stream it would use only 15W. Because it's using the hardware decoder, the CPU is barely used (certainly not in 100% as it is in sysbench)
And it's not just AMD's monolithics. I'm always impressed when I see my parents' i3-12100 play 1080p YT video (WUXGA screen) at less than 4 W of indicated package power even at elevated speed. I once set it to 5 W PL1 and 15 W PL2 for grins and it did slow down noticeably when loading things then (while video playback moved closer to 3.5 W), but I think you would barely notice anything in normal use at say 15 W PL1 and 28 W PL2. I did have to do some tweaking to C-states and ASPM settings and such to get idle power down if memory serves - the CPU makes it to C6 at about 0.6 W then, with the whole system at about 16 W from the wall (the board has some onboard USB hubs to complement the B760 chipset and runs XMP RAM at 1.35 V so you could probably shave off another 1.5 W or so if you insisted, and several watts more with a PicoPSU setup instead of a conventional 400W ATX PSU). I can get the CPU to 72-73 W at the absolute max with Prime95 Small FFTs using AVX2 (4.1 GHz allcore), but you can limit power to half that and the drop in clocks is not exactly dramatic. You can probably guess that with a midrange tower cooler and in a midtower case, the system normally runs very, very quiet. Still a very ITX-friendly processor no doubt, while being a lot faster and more capable than even the N100, though obviously at a price. If your board / OS of choice doesn't let you play with the power limits, you may consider a T SKU (PL1 = 35 W instead of 60 W). My own system (i7-11700 on Z590) is not nearly as efficient - while I did manage to get it to 14.5 W from the mains at idle, power under load is a different story, with CPU package power during YT playback being at 5-6 W at normal speed and 9-12 W at elevated speed if you 're lucky (and I've seen up to 21-24 W). I guess that's what twice the cores and more L3 cache on a much older process (14nm+++) gets you. And it's not like it feels any faster either, quite the contrary (Alder Lake at 4.3 GHz slightly beats Rocket Lake at 4.9 GHz thanks to much larger L2 caches), and I'm normally running it with reduced turbo multipliers to boot. Yes, it does still have more _oomph_ in heavily multithreaded workloads, but the system doesn't see those very often outside of benchmarks.
Glmark2 is great for testing desktop manager performance as well, using 'glmark -s 1920x1080' to run a hidef test on the destop, then 'glmark --fullscreen' to run a native hidef test and comparing the numbers.
Personally I would always pick the N100 system, mainly due to the expansion slot and hardware virtualisation capabilities. The performance is also nice for the intended use case: NAS, firewall, etc. It also has AES acceleration but unfortunately no Quick Assist (which most users won't miss anyway). The ~10W increase in power usage is only when the unit is at full load. Most devices run at under 10% load, so that's not a huge problem. It would have been nice to also measure the temperature but that's for future videos - fingers crossed :)
Awesome video. I really appreciate the work that went into the benchmarks. Thank you also for highlighting the repo for the Rockchip-optimized Ubuntu builds. I had no idea that was a thing. :) It was interesting to see that it was the graphics performance area where the ARM SBC really fell down. I think this would be an interesting topic to revisit if Rockchip ever ups their iGPU game, but in the meantime, especially for home server use, I think the more powerful Intel iGPU makes a big difference. Once the Intel N30x chips/boards drop in price a bit more, that will be an interesting comparison to the Rockchip RK3588 as well, as they're (mostly) more evenly matched on cores, even if the difference in the Rockchip architecture complicates that. This is a big leap forwards, but the graphics and compatibility on the N100 seem to give it the edge unless maximizing power efficiency is critical for the application.
The computing power the ARM provides with 15 watts is nothing short of astounding! I'm especially surprised how well it performed in the "brute" CPU task. Thanks for the great comparison!
If only integrated GPUs in the ARM world were better. I don’t know if it the hardware, the software support, or both but it’d be fantastic to see better iGPUs in that world.
Thanks Chris for an interesting comparison test, I thought that the RK3588 with the extra cores & threads would beat the the N100 easily, I personally still favour the N100 even though it’s a bit more expensive & has a higher power consumption it would suit my requirements! It was good to see lots of Ducks & Geese in the video tests :) 🦆🦆🦆🦆🦆🦆👍
The RK3588 wins in synthetic benchmarks, the N100 wins in practical benchmarks despite having 4 fewer cores and it also wins in having more software available to it both user software and the operating systems the user would want to run as the N100 is not restricted to special builds that must be specifically built for it like with the RK3588.
Excellent and fair review of these two interesting products! I really like seeing ARM tapping into another marketshare, the desktop environment here. Such competition can only encourage further improved value and quality in the market! Cheers Christopher!
This was a really good test video, and it looks like ARM is hoding its own. Hopefully it won't be too many years before you can make a legit test of RISC v ARM v X86 test.
Regardless of the individual test scores I think it's great to see that relatively inexpensive and low power systems such as these can still offer usable performance in many everyday computing tasks. In particular I appreciate the advances modern processors have made in high resolution video playback!
Appreciate the expected game between 3588 and N100, great one, Chris. I've said it could be similar price with half performance and half power consumption in the comments of previous video. Now, it seems the 30% price diff came from AsRock, and this can be eliminated if the N100 was bought from Shenzhen. And the similar idle power… I think it comes from the various interfaces of the 3588-ITX (compared with pi 5B), so this wasn't considered in my mind, sorry. Keep that going, Chris, stay safe and bring more videos.
Although I would pick the N100 every time, I'm amazed at how far Arm processors have come in the last few years. Processing power per £ or $ it's so close now.
All tricked out, 250 for board, 90 for case v21 TT, 2x 90ea NVME, 120 mem 32GB maxxed. Had Wifi/BT m.2 card, BT mouse 30, keyb ps2.Approaching a thousand with capture 250, and pcie 16x multi m2 120. Canadian dollars.Looking at Elgato 'mixer' 350. Yeah. I was thinking how cheap it would be.
Another masterpiece. Thanks for all the info. I actually had a dream last night that I was watching one of your videos, so you literally made my dream come true!😊
Excellent, thorough, and very fair comparison. Your standard benchmark suite is really very insightful for many different computers and operating systems.
Another excellent video as always. Here’s hoping this if four in a row with over 100,000 views. I would be curious to know the maths in terms of how much the electrical cost difference would be between the two systems over one years time.
I've been looking forward to a head to head with these chips, and this did not disappoint. I bet some software optimization could help those real world benchmarks in the future. Some day...
Very interesting. The RK3588 won on five measures and the N100 won on four measures with a draw on TH-cam 4K rendering. Where the N100 excelled it was significantly ahead of the RK3588. In most cases where the RK3588 won there was very little between them. Overall they both did everything required. Ultimately this shows that RISC-V boards are already a very viable alternative and are becoming ever more capable.
People say that the N100 completely crushes the RK3588 thanks to the software compatibility, but they forget there's a number of applications that require various embedded functionalities that the N100 doesn't have - such as GPIO support (with various communication buses), CSI camera support or even the built-in NPU. That's completely non-existent on the N100, where you're forced to use external MCUs for GPIO, external NPUs/TPUs/GPUs for AI acceleration and are limited to USB and HDMI when it comes to cameras and displays. N100 computers or SBCs are great for casual users, who simply want to have a low-power desktop or server, but for makers I believe they're out of the question. If you wanna tinker a little with simple robotics, internet of things or edge AI, the RK3588 wins. The only thing we need now is good software support, which, I admit, sucks very much (with the exception of Joshua's work).
I had been using a 2012 AMD FX-6300 system until 2021, which was somewhat sluggish but not unusable for basic tasks at the time. It's quite spectacular that the N100 apparently significantly outperforms that CPU with a TDP nearly 16 times lower (Geekbench).
Thanks Christopher for the comparison. I ordered an N100 board after seeing your previous video. Just waiting for it to arrive and build a silent itx build. 🙂
Thanks for another great video. Quite amusing and refreshing to see the two systems side by side. Despite the relative immaturity of the Rock5 image, it still performs quite close to the N100, with lower power consumption. I do have a rock 5b already, fitted with an nvme drive, but it's runding armbian. I believe I'll test the ubuntu image later this week to see if it performs like the full rk3588 ITX variant.
When I was a kid in the 1960's, the height of technology was our black & white 405 line valve TV, and my father's radiogram! Wind forwards 60 years, and I have a completely flat TV, a flat handheld computer, a box that answers my questions and turns my lights on, a pocket computer that takes photos and makes phone calls, and a PC that fits in the palm of my hand! One wonders what the future holds? Particularly after we're no longer around?
Absolutely loved it, since it confirms my studies. NCM (N 100 M) or the Nice Cube Machine is getting an upgrade, so we won't give you comparative stats yet. 4x HDMI capture coming on Tuesday and heat management (a fan).
I got an N100 based mini-PC in a 12x12x5 cm case (with external power supply) that weights less than half of a kilogram. Frankly, it's quite usable since it's performance is comparable with i5-7500 I had years ago (i5-7500 has a TDP of 65W, while N100 in my system has a limited power budget of 6W). Neat little PC that you can snap on a back of a monitor, or connect to a TV, or put into a bag to carry.
Thanks Chris and say Hi to the Crystal Rabbit! I thought you going to take a page off the movie “Toy Story “ in Woody vs the Etch-A-Sketch and then draw their guns at 1080p on Blender…😂! Thanks for the comparison video and in off grid places such as camping or a cabin the extra wattage might be a deciding factor for potential purchasers. Have a great day!
You're going to make a me move to a mini ITX and Linux yet...I think it might be time to replace my 7 year old NUC computer that has served me extremely well, but 4K is making it chock these days....another smashing video Mr Barnett.
CHRIS, LOVE UR STUFF! Because these platforms seemed pretty even in your tests, but with different outcomes in various areas, might it have been useful to give your opinion which model to buy depending on your needs? For example, if I wanted a machine for gaming which one would be more practical? How about for business work? i.e. spreadsheets and word processing. Which machine for programming? For graphics designers? Your 1 million viewers value your opinion.. Many thanks for all your detailed work!
Great questions. The N100 system would be better for gaming, and for anything where a breadth of software support is required. However, if all that is required are browsers, office apps, an e-mail client, media players, etc, then the Arm system has its merits -- not least in terms of power use, and security. There is not a lot of Arm desktop malware out there! This said, running Linux on an x86 system like the N100 also offers major security benefits. I will ponder on your comment and how to include a discussion of the same in future content. Thanks! :)
RK3588 with the upcoming Panthor Mali drivers will be a beast i think, i bought an Orange Pi 5 at pre-release and regretted it ever since.. barely usable, i really adopted way too soon. But it's looking better now
nice video again chris. its kio on my other account. I'm not surprised the n100 beat the rk3588 a lower end x86 can compete and beat an arm based unit. the architectural designs of the chips are not equal in terms of performance in regards to ghz and power. x86 handles more complex computations and arm is simple. I've shown a few videos showing how real world emulation like ps2 on the radxa x2l kicked the crap out of pi5 and orange pi 5 in a linux environment in regards to emulation and other various tests. the geek bench scores and other measurable results mean nothing. great video again as always. I've been trying to get the word out to people that these sbcs by just hardware don't mean anything if its not paired with a great os to compliment it. these real world test results are the best way to decide whats what
Thak you very much. It was very interesting. I thought the N100 would win easily, and got surprised when the RK took a lead on the synthetic tests. The N100 took back the crown on the apps, but it ended being much more tied than I anticipated . Best regards mate.
I too think that both PCs have performed very well, so I can't just choose one winner. :) Also, yaaaaaaay! Ducks _and_ geese in the same video!! Quack quack. 🦆❤️
ran glmark2 myself out of curiosity. I scored 8792 points. Core i5-4590 paired with GTX 1050 (non-Ti) 4GB. It's not mini-ITX, but it is small form factor. 10 year old platform, with nearly 8 year old GPU tech, as well.
The result surprised me. I think if the extra price paid for N100 was used to include a better GPU in the RK3588, the RK2588 could be a winner in all aspects. Thanks Chris!
The drivers aren't good. The Mali GPU should have a bit more power but we're in the mini-pc area here so... I go with a Ryzen 9 8945HS + 96 GB DDR5 5600 or a Ryzen 7 87840HS + 32 GB DDR5 5600 depending on what I intend to do with it. Gen4 SSDs and USB4 alone are worth it for me. Edit: I should mention that I actuall have a bunch of those systems running here witch Arch Linux & Debian.
@@DJDocsVideosFair point, BUT; The N100 (and the L chip) are supposed to be low cost, low-power computers that offer an acceptable UX, and not performance beasts like you describe. It also seems to me like the N100 is probably a little more oriented to office work, hence the more powerful graphics than you usually get with processors in that price range.
@@DJDocsVideos The specs you mentioned are different class. Like comparing to a Ferrari to a Honda. I am looking forward to further refinement of the RISC-V and Mali GPU as time goes on. I am currently using Ryzen 9 5950X with Radeon XFX 6900 XT video card and 128 gig RAM in a custom built PC that I put together myself. I use it mostly for remote desktop work, e-mail and web surfing so the RISC-V based mini-pc would be perfect for me as a daily driver. Although I occasionally run Handbrake which is where my Ryzen 9 shines.
I think on the cpu test you shouldve done a single thread test as well, since unless you are doing certain computation heavy tasks (which you likely wont be doing on a mini computer anyway), the vast majority of tasks and processes are single thread. So if indeed the ARM based system scored better in the cpu test simply because it has more cores, it would have been good to see a single core vs single core matchup.
I pre-subscribed to make sure I'd be subscribed, and I pre-liked the video because I knew I'd love it to pieces, and I pre-wrote this comment right after watching the video to make sure I didn't just waltz away without shining 2 watts of verbiage on yet another video of the quality I for one greatly pre-enjoy, and in fact, also simply enjoy. :)
Thanks for another quality comparison video. It would be good if you also clock in the power usage for each of the tests for both systems and give a score per watt index. This would give more clarity into the efficiency of both systems, which is what these CPUs are touted for.
I think higher power usage under stress is a win, not lose. It's good to have more power available when needed. 200W kettle is not better than 1kW kettle if needs 5 times more time to boil the same amount of water. N100 is a clear winner if can render 100% faster using 30% more power
This pretty much matches my expectations. Choosing one architecture type over the other really depends on how impatient you are. RISC-based architectures are generally slower and partly because they use less power. In a lot of these graphics capabilities tests, more cores is an advantage, but clearly a faster chip will overpower that advantage as shown here. So it'd be nice to see a test with an older x86 chip that more closely matches the speed and core count of the RK3588 and see how that changes the result. However, given the low requirements of startup processes, the higher core count gives an advantage here. Although, once optimizations have caught up, maybe RISC-based architectures will beat x86, and that would be interesting.
Thanks, another great video. Tho, the final results could be better. Hard to interpret . Please add percentage differences. Just highlighting in yellow is not enough
Oooh, hardware. My favorite. (settles in with tea and crumpets * ) Very interesting. The RK is a little better for other stuff, but the N100 blows it away on anything graphics-intensive. ( * I didn't have any crumpets, so it's really a peanut butter sandwich)
I came here to watch the Crystal Rabbit. Did not disappoint.
It's a very good rabbit too!
@@Praxibetel-Ix Unless it is from Caerbannog
I'm thinking crystal hare my self. ps I'm hooked!
n100 is a monster. i can run crysis on it.
The winner here is us, the customers. Both boards performed really well. x86-64 still has a lot more compatibility, and arm is super efficient.
Agreed! :)
The ARM system is not really more efficient in tests such as gimp, kdenlive, web basemark and probably glmark as well because whilst it uses ~60% of the power it's performance is sometimes over 40% slower. It's probably more efficient in the pure number crunching tests such as sysbench CPU but those tests are the least likely to be representative of real world use. I also suspect from the icons on the left panel that the standard ubuntu installation might have more programmes installed than the Rockchip specific ubuntu installation.
n100 is a monster. i can run crysis on it.
The RK3588 definitely has its use cases, especially on low power applications, mobile solutions, or where the system is being powered by solar/battery.
n100 is a monster. i can run crysis on it.
@@mal-avcisi9783 absolutely love the n100. It’s the backbone of my custom NAS. Sips power, but has a phenomenal processing power to consumption ratio.
@@pikaskew I can also confirm that I bought a Chuwi Mini-PC with 12 GB of RAM and a 512 GB NVMe N100 for 129 dollars, and it is simply fantastic. Previously, I had a Raspberry Pi 5, which was too slow for me and had poor compatibility with certain containers that were x86-based. Now I'm using this little powerhouse, and it is so much better than the Raspberry Pi 5. I was even able to install QEMU and run Windows 11 in a VM under Ubuntu Linux. It’s really great, man. The N100 impresses. Moreover, it is extremely efficient: at maximum load, it only consumes 21 watts, and in normal operation, it’s even just 9 watts.
The biggest takeaway for me, is the fact that there is an Ubuntu that works very nicely with Rock chips. That makes the Rock 5 B look attractive again.
This is very true -- and all due to the work of the Ubuntu Rockchip project. They have done an amazing job.
After hours of exhausting tests, the conclusion - they both work fine.
Amazing how far both platforms have come in terms of performance, efficiency and low cost. The 4K video playback in particular was amazing!
You present a well organized review.. well done.. I so look forward to Sunday morning to watch EC! Thank you ever so much for the great content!
As I am very new to the ARM architecture it was very interesting to see. Thank you for the video!
Great comparison. I'm always amazed by my n-100 rig, the thing never gets too hot. That's one great board, even with a little more wattage. Thanks Chris!
Yes! the good old Apples to Apples test. Goes a long way to clearing any confusion in the air. Thank you for this.
Another great video Chris. It is interesting to see how fat the RISC processors have come. I have an older Intel N5105 Brick PC which is a older and a little slower than the N100, but seems fine for basic tasks and web browsing. All in all, an interesting comparison. Thanks for all your hard work Chris!
Memorised by the crystal rabbit! Great to confirm the N100 is pretty fast at graphical tasks!
didn’t expect to see the Sega expert here
It is actually amazing to see how 7W of higher energy consumption can lead to such an increase in performance. Having a 22W PC capable of 4K streaming is actually quite impressive. I am currently experimenting with a Fujitsu Futro S520, which has a 7W total power consumption and is capable of 720p @ 60 fps video streaming outside of browsers via the mpv. I realized that using my desktop PC (better berformance although old) with its 160W of power consumption is kind of wasted if used for browsing and videostreaming.
Just go for a 15W zen 3 in laptop or mini pc. Almost feels like a 65W desktop part in most scenarii.
I wouldn't be surprised if for the 4K stream it would use only 15W. Because it's using the hardware decoder, the CPU is barely used (certainly not in 100% as it is in sysbench)
And it's not just AMD's monolithics. I'm always impressed when I see my parents' i3-12100 play 1080p YT video (WUXGA screen) at less than 4 W of indicated package power even at elevated speed. I once set it to 5 W PL1 and 15 W PL2 for grins and it did slow down noticeably when loading things then (while video playback moved closer to 3.5 W), but I think you would barely notice anything in normal use at say 15 W PL1 and 28 W PL2. I did have to do some tweaking to C-states and ASPM settings and such to get idle power down if memory serves - the CPU makes it to C6 at about 0.6 W then, with the whole system at about 16 W from the wall (the board has some onboard USB hubs to complement the B760 chipset and runs XMP RAM at 1.35 V so you could probably shave off another 1.5 W or so if you insisted, and several watts more with a PicoPSU setup instead of a conventional 400W ATX PSU). I can get the CPU to 72-73 W at the absolute max with Prime95 Small FFTs using AVX2 (4.1 GHz allcore), but you can limit power to half that and the drop in clocks is not exactly dramatic.
You can probably guess that with a midrange tower cooler and in a midtower case, the system normally runs very, very quiet. Still a very ITX-friendly processor no doubt, while being a lot faster and more capable than even the N100, though obviously at a price. If your board / OS of choice doesn't let you play with the power limits, you may consider a T SKU (PL1 = 35 W instead of 60 W).
My own system (i7-11700 on Z590) is not nearly as efficient - while I did manage to get it to 14.5 W from the mains at idle, power under load is a different story, with CPU package power during YT playback being at 5-6 W at normal speed and 9-12 W at elevated speed if you 're lucky (and I've seen up to 21-24 W). I guess that's what twice the cores and more L3 cache on a much older process (14nm+++) gets you. And it's not like it feels any faster either, quite the contrary (Alder Lake at 4.3 GHz slightly beats Rocket Lake at 4.9 GHz thanks to much larger L2 caches), and I'm normally running it with reduced turbo multipliers to boot. Yes, it does still have more _oomph_ in heavily multithreaded workloads, but the system doesn't see those very often outside of benchmarks.
yeah this is exactly why im looking into a low consuming mini pc running on a light OS. so much energy wasted.
Glmark2 is great for testing desktop manager performance as well, using 'glmark -s 1920x1080' to run a hidef test on the destop, then 'glmark --fullscreen' to run a native hidef test and comparing the numbers.
those are some real nice tips there
Just as I was thinking about ARM and x86, and I mean this very day, you upload this. Talk about timing.
I was on the edge of my seat during the boot competition!
This is great to hear!
I really like compact builds AND YES I was among those ones interested in this computer 😎👍.
Personally I would always pick the N100 system, mainly due to the expansion slot and hardware virtualisation capabilities. The performance is also nice for the intended use case: NAS, firewall, etc. It also has AES acceleration but unfortunately no Quick Assist (which most users won't miss anyway). The ~10W increase in power usage is only when the unit is at full load. Most devices run at under 10% load, so that's not a huge problem. It would have been nice to also measure the temperature but that's for future videos - fingers crossed :)
Awesome video. I really appreciate the work that went into the benchmarks. Thank you also for highlighting the repo for the Rockchip-optimized Ubuntu builds. I had no idea that was a thing. :)
It was interesting to see that it was the graphics performance area where the ARM SBC really fell down.
I think this would be an interesting topic to revisit if Rockchip ever ups their iGPU game, but in the meantime, especially for home server use, I think the more powerful Intel iGPU makes a big difference.
Once the Intel N30x chips/boards drop in price a bit more, that will be an interesting comparison to the Rockchip RK3588 as well, as they're (mostly) more evenly matched on cores, even if the difference in the Rockchip architecture complicates that.
This is a big leap forwards, but the graphics and compatibility on the N100 seem to give it the edge unless maximizing power efficiency is critical for the application.
very useful and interesting. it seemed like a lot of work on your part. thanks for posting.
The computing power the ARM provides with 15 watts is nothing short of astounding! I'm especially surprised how well it performed in the "brute" CPU task.
Thanks for the great comparison!
If only integrated GPUs in the ARM world were better. I don’t know if it the hardware, the software support, or both but it’d be fantastic to see better iGPUs in that world.
Thank you for this brilliant suite of benchmarks. I see a few new tests I plan to adopt for my own experiments.
Thanks Chris for an interesting comparison test, I thought that the RK3588 with the extra cores & threads would beat the the N100 easily, I personally still favour the N100 even though it’s a bit more expensive & has a higher power consumption it would suit my requirements! It was good to see lots of Ducks & Geese in the video tests :) 🦆🦆🦆🦆🦆🦆👍
I truly love the n100 CPU. For me it has replaced the Raspberry Pi SBCs for the foreseeable future. ❤😊
Just installed Ubuntu on one of my machines for first time in years, wow, I am impressed. Tempted to swap my mint installs over to it also.
The RK3588 wins in synthetic benchmarks, the N100 wins in practical benchmarks despite having 4 fewer cores and it also wins in having more software available to it both user software and the operating systems the user would want to run as the N100 is not restricted to special builds that must be specifically built for it like with the RK3588.
Back on the road again Mr. Barnatt, just got home Sunday afternoon and enjoying you Explain Computer to me!! Love ya Man!!!
Excellent and fair review of these two interesting products! I really like seeing ARM tapping into another marketshare, the desktop environment here. Such competition can only encourage further improved value and quality in the market! Cheers Christopher!
Agreed!!!!
I wish AMD had a greater presence in this market.
This was a really good test video, and it looks like ARM is hoding its own. Hopefully it won't be too many years before you can make a legit test of RISC v ARM v X86 test.
Hopefully by 2026. :)
Regardless of the individual test scores I think it's great to see that relatively inexpensive and low power systems such as these can still offer usable performance in many everyday computing tasks. In particular I appreciate the advances modern processors have made in high resolution video playback!
I totally agree.
Greetings. This video certainly delivers what you've promised last week and week before. A comprehensive comparative video.
15:56 Came for the ducks, stayed for the awesome review!
Greetings!
Duckies are the cutest 😊
Do not forget the rabbit!
@@dorinp007 Bunny 🐇🐰🤗
Appreciate the expected game between 3588 and N100, great one, Chris. I've said it could be similar price with half performance and half power consumption in the comments of previous video.
Now, it seems the 30% price diff came from AsRock, and this can be eliminated if the N100 was bought from Shenzhen. And the similar idle power… I think it comes from the various interfaces of the 3588-ITX (compared with pi 5B), so this wasn't considered in my mind, sorry.
Keep that going, Chris, stay safe and bring more videos.
Thanks for the explanations and update on your progress and the differences Mike
Arm is so good at power efficiency. We only need native arm software ports now to make the switch possible.
Although I would pick the N100 every time, I'm amazed at how far Arm processors have come in the last few years. Processing power per £ or $ it's so close now.
I like those mini PC, really cheap.
Great aren't they! 😄
All tricked out, 250 for board, 90 for case v21 TT, 2x 90ea NVME, 120 mem 32GB maxxed. Had Wifi/BT m.2 card, BT mouse 30, keyb ps2.Approaching a thousand with capture 250, and pcie 16x multi m2 120. Canadian dollars.Looking at Elgato 'mixer' 350. Yeah. I was thinking how cheap it would be.
Another fantastic video Mr Barnatt, you've definitely made my Sunday evening more enjoyable.
Молодец. Отличное видео.😊
The whole channel is cool.
Another masterpiece. Thanks for all the info. I actually had a dream last night that I was watching one of your videos, so you literally made my dream come true!😊
:)
Excellent, thorough, and very fair comparison. Your standard benchmark suite is really very insightful for many different computers and operating systems.
Another excellent video as always. Here’s hoping this if four in a row with over 100,000 views. I would be curious to know the maths in terms of how much the electrical cost difference would be between the two systems over one years time.
I've been looking forward to a head to head with these chips, and this did not disappoint. I bet some software optimization could help those real world benchmarks in the future. Some day...
There is a way that nature speaks, that land speaks. Most of the time we are simply not patient enough, quiet enough, to pay attention to the story.
Cheers Chris excellent comparison. Disappointed not to see Mr Scissors or Stan the knife
they were being sharpened for future cuts ...
Even cutting tools need their rest.
@@jyvben1520 priceless humour
@@Praxibetel-Ix lol
This test series shows us something we already knew, ARM GPUs need to be improved significantly.
Very interesting. The RK3588 won on five measures and the N100 won on four measures with a draw on TH-cam 4K rendering. Where the N100 excelled it was significantly ahead of the RK3588. In most cases where the RK3588 won there was very little between them. Overall they both did everything required. Ultimately this shows that RISC-V boards are already a very viable alternative and are becoming ever more capable.
Thanks for the comparison. I enjoyed it. Looking forward to your next video!
Thanks Perry. :)
Spectacular. Thank you for your hard work and many years of reviews, tests, development and your time bring in IT.
People say that the N100 completely crushes the RK3588 thanks to the software compatibility, but they forget there's a number of applications that require various embedded functionalities that the N100 doesn't have - such as GPIO support (with various communication buses), CSI camera support or even the built-in NPU. That's completely non-existent on the N100, where you're forced to use external MCUs for GPIO, external NPUs/TPUs/GPUs for AI acceleration and are limited to USB and HDMI when it comes to cameras and displays. N100 computers or SBCs are great for casual users, who simply want to have a low-power desktop or server, but for makers I believe they're out of the question. If you wanna tinker a little with simple robotics, internet of things or edge AI, the RK3588 wins. The only thing we need now is good software support, which, I admit, sucks very much (with the exception of Joshua's work).
I want the N100 just for making an arcade/emulation (PS2, SWITCH, etc)/multimedia machine. Do you recommend the N100 or the RK3588S?
i have 3 mini pc with n100, 2.5gb/s ethernet and Proxmox installed over them (with ceph) outstanding devices
Excellent comparison! Thank you Christopher!
I had been using a 2012 AMD FX-6300 system until 2021, which was somewhat sluggish but not unusable for basic tasks at the time. It's quite spectacular that the N100 apparently significantly outperforms that CPU with a TDP nearly 16 times lower (Geekbench).
Thanks Christopher for the comparison. I ordered an N100 board after seeing your previous video. Just waiting for it to arrive and build a silent itx build. 🙂
Was looking forward to a rousing tech review of just this nature. Much thanks.
You should call the Kentucky Derby next year. The enthusiasm is awesome.
Thanks for another great video. Quite amusing and refreshing to see the two systems side by side. Despite the relative immaturity of the Rock5 image, it still performs quite close to the N100, with lower power consumption. I do have a rock 5b already, fitted with an nvme drive, but it's runding armbian. I believe I'll test the ubuntu image later this week to see if it performs like the full rk3588 ITX variant.
Great comparisons, Chris. Should do really well for the channel. Thanks for sharing with us.
Thanks Brian. :)
Another good explanatory video, thanks
Interesting comparison of the two machines. There's not a lot between them so I think that the ultimate end use will determine which one to buy
Great vid! Long time question finally answered. Thanks for doing the comparison
When I was a kid in the 1960's, the height of technology was our black & white 405 line valve TV, and my father's radiogram!
Wind forwards 60 years, and I have a completely flat TV, a flat handheld computer, a box that answers my questions and turns my lights on, a pocket computer that takes photos and makes phone calls, and a PC that fits in the palm of my hand!
One wonders what the future holds? Particularly after we're no longer around?
Absolutely loved it, since it confirms my studies. NCM (N 100 M) or the Nice Cube Machine is getting an upgrade, so we won't give you comparative stats yet. 4x HDMI capture coming on Tuesday and heat management (a fan).
I got an N100 based mini-PC in a 12x12x5 cm case (with external power supply) that weights less than half of a kilogram. Frankly, it's quite usable since it's performance is comparable with i5-7500 I had years ago (i5-7500 has a TDP of 65W, while N100 in my system has a limited power budget of 6W). Neat little PC that you can snap on a back of a monitor, or connect to a TV, or put into a bag to carry.
Informative comparison. Now looking forward to further ARM desktop adventures, most notably your assessment of the state of Windows on ARM.
I like your Tests very much.
Thank you.
Another hit. They just keep coming! Thanks.
Just excellent Chris. Well done. Thanks.
Thanks Chris and say Hi to the Crystal Rabbit! I thought you going to take a page off the movie “Toy Story “ in Woody vs the Etch-A-Sketch and then draw their guns at 1080p on Blender…😂!
Thanks for the comparison video and in off grid places such as camping or a cabin the extra wattage might be a deciding factor for potential purchasers.
Have a great day!
You're going to make a me move to a mini ITX and Linux yet...I think it might be time to replace my 7 year old NUC computer that has served me extremely well, but 4K is making it chock these days....another smashing video Mr Barnett.
A great comparison, thank you very much for the effort you put in making these videos!
CHRIS, LOVE UR STUFF! Because these platforms seemed pretty even in your tests, but with different outcomes in various areas, might it have been useful to give your opinion which model to buy depending on your needs? For example, if I wanted a machine for gaming which one would be more practical? How about for business work? i.e. spreadsheets and word processing. Which machine for programming? For graphics designers? Your 1 million viewers value your opinion.. Many thanks for all your detailed work!
Great questions. The N100 system would be better for gaming, and for anything where a breadth of software support is required. However, if all that is required are browsers, office apps, an e-mail client, media players, etc, then the Arm system has its merits -- not least in terms of power use, and security. There is not a lot of Arm desktop malware out there! This said, running Linux on an x86 system like the N100 also offers major security benefits. I will ponder on your comment and how to include a discussion of the same in future content. Thanks! :)
I always dance the robot every time the theme music plays.
yep OG has beat us all in tech since the 90s or 80s for sure mad respected❤🔥❤🔥❤🔥
That was such a nice. well-made test. Nicely done.
RK3588 with the upcoming Panthor Mali drivers will be a beast i think, i bought an Orange Pi 5 at pre-release and regretted it ever since.. barely usable, i really adopted way too soon.
But it's looking better now
they have 6.1 bsp kernels with the Panthor driver, scores much better in glmark2 i just got score of 2470 with performance governer set.
nice video again chris. its kio on my other account. I'm not surprised the n100 beat the rk3588 a lower end x86 can compete and beat an arm based unit. the architectural designs of the chips are not equal in terms of performance in regards to ghz and power. x86 handles more complex computations and arm is simple. I've shown a few videos showing how real world emulation like ps2 on the radxa x2l kicked the crap out of pi5 and orange pi 5 in a linux environment in regards to emulation and other various tests. the geek bench scores and other measurable results mean nothing. great video again as always. I've been trying to get the word out to people that these sbcs by just hardware don't mean anything if its not paired with a great os to compliment it. these real world test results are the best way to decide whats what
Thak you very much. It was very interesting.
I thought the N100 would win easily, and got surprised when the RK took a lead on the synthetic tests. The N100 took back the crown on the apps, but it ended being much more tied than I anticipated .
Best regards mate.
Greetings. I like this comparison in the frontier between SBCs and Mini PCs. It reminds me of the frontier between FPGA and CPLD devices.
lovely video. thank you a lot chris for compare theese two teeny tiny machines 😊 !
Thank you for the comparison, and years of videos.
I too think that both PCs have performed very well, so I can't just choose one winner. :)
Also, yaaaaaaay! Ducks _and_ geese in the same video!! Quack quack. 🦆❤️
ran glmark2 myself out of curiosity. I scored 8792 points. Core i5-4590 paired with GTX 1050 (non-Ti) 4GB. It's not mini-ITX, but it is small form factor. 10 year old platform, with nearly 8 year old GPU tech, as well.
The result surprised me.
I think if the extra price paid for N100 was used to include a better GPU in the RK3588, the RK2588 could be a winner in all aspects.
Thanks Chris!
The drivers aren't good. The Mali GPU should have a bit more power but we're in the mini-pc area here so...
I go with a Ryzen 9 8945HS + 96 GB DDR5 5600 or a Ryzen 7 87840HS + 32 GB DDR5 5600 depending on what I intend to do with it. Gen4 SSDs and USB4 alone are worth it for me.
Edit: I should mention that I actuall have a bunch of those systems running here witch Arch Linux & Debian.
@@DJDocsVideos I totally forgot about the drivers.
Thanks for the info
@@DJDocsVideosFair point, BUT; The N100 (and the L chip) are supposed to be low cost, low-power computers that offer an acceptable UX, and not performance beasts like you describe. It also seems to me like the N100 is probably a little more oriented to office work, hence the more powerful graphics than you usually get with processors in that price range.
@@DJDocsVideos The specs you mentioned are different class. Like comparing to a Ferrari to a Honda. I am looking forward to further refinement of the RISC-V and Mali GPU as time goes on. I am currently using Ryzen 9 5950X with Radeon XFX 6900 XT video card and 128 gig RAM in a custom built PC that I put together myself. I use it mostly for remote desktop work, e-mail and web surfing so the RISC-V based mini-pc would be perfect for me as a daily driver. Although I occasionally run Handbrake which is where my Ryzen 9 shines.
LOVE THESE "TECH DUELS".
Great video as always! Suggest changing the 4k video playback to 60fps.
I think on the cpu test you shouldve done a single thread test as well, since unless you are doing certain computation heavy tasks (which you likely wont be doing on a mini computer anyway), the vast majority of tasks and processes are single thread. So if indeed the ARM based system scored better in the cpu test simply because it has more cores, it would have been good to see a single core vs single core matchup.
I pre-subscribed to make sure I'd be subscribed, and I pre-liked the video because I knew I'd love it to pieces, and I pre-wrote this comment right after watching the video to make sure I didn't just waltz away without shining 2 watts of verbiage on yet another video of the quality I for one greatly pre-enjoy, and in fact, also simply enjoy. :)
Thanks for another quality comparison video.
It would be good if you also clock in the power usage for each of the tests for both systems and give a score per watt index. This would give more clarity into the efficiency of both systems, which is what these CPUs are touted for.
I think higher power usage under stress is a win, not lose. It's good to have more power available when needed.
200W kettle is not better than 1kW kettle if needs 5 times more time to boil the same amount of water.
N100 is a clear winner if can render 100% faster using 30% more power
RK3588 has excellent low power performance. Intel graphics tough to beat on less expensive systems.
Thanks for another Sunday info vid.
U R the "Un-biased Benchmark Whisperer". I really like that about you. Cheers from my comfee chair.
Thanks. :)
This pretty much matches my expectations. Choosing one architecture type over the other really depends on how impatient you are. RISC-based architectures are generally slower and partly because they use less power. In a lot of these graphics capabilities tests, more cores is an advantage, but clearly a faster chip will overpower that advantage as shown here. So it'd be nice to see a test with an older x86 chip that more closely matches the speed and core count of the RK3588 and see how that changes the result. However, given the low requirements of startup processes, the higher core count gives an advantage here. Although, once optimizations have caught up, maybe RISC-based architectures will beat x86, and that would be interesting.
Awesome video again!
I think I would prefer the RK3588.
Thanks, another great video. Tho, the final results could be better. Hard to interpret . Please add percentage differences. Just highlighting in yellow is not enough
Oooh, hardware. My favorite. (settles in with tea and crumpets * )
Very interesting. The RK is a little better for other stuff, but the N100 blows it away on anything graphics-intensive.
( * I didn't have any crumpets, so it's really a peanut butter sandwich)
Excellent comparison. My choice would be the N1000