Ok, I guess I will wait a few more years then. I don't even need faster ports, my use case is literally a router, everything goes through a managed switch, and the 200/100Mbit speeds we get from our ISP is perfectly enough.
Hmm, these chips seem to get hotter with every new generation. My J5005/J4125 are pretty cool. N5105 is quite a bit warmer. My N100 is running quite hot (X86-P5 V1, should probably redo thermal paste) and now you say the N150 units run even hotter?One thing to note is the N150/N355 cpus run a tad hotter then their previous N100/N350s. A few NAS boxes out there I suspect saw the possible thermal issues and stuck with the N100 over the N150 like the lincstation N2. I believe the early prototype or specs mentioned a possible N150 but then went with the N100 (lincstation N2).
My GMKtec N150 ran so hot it literally melted and died, but it was a very badly built and still is faulty product imo. This was with 2x noctua 120mm blowing on top and bottom with lots of mogding though.
Think its better to look for something with better performance and especially more pci-e lanes, even if it does use more wattage since you end up with more headroom for performance down the road.
Yeah tad hotter, I think all intel did was rebadged the N100 by boosted the cpu and gpu clocks a tad for the N150/350, so it just ran more hotter. If they added more pci-e lanes that would have been an worthy upgrade, maybe next time.Hmm, these chips seem to get hotter with every new generation. My J5005/J4125 are pretty cool. N5105 is quite a bit warmer. My N100 is running quite hot (X86-P1 V1, should probably redo thermal paste) and now you say the N150 units run even hotter?
Probably should think about how to use the heat, cooking coffee maybe? Or maybe these units are all just cheap shit.
My Topton/CWWK fanless N100 arrived with thermal paste that looked like it had been applied with a tablespoon. Way too much of it and not very good at that. First thing I did was remove it all and re-apply my own with just enough to do the job. Temps dropped by at least 10C at idle and probably more like 20C under load.My N100 is running quite hot (X86-P1 V1, should probably redo thermal paste)
Yeah, going to do this. Every other unit I repasted, but the N100 was never in real production here, just a testing device. Recently built an OPNsesne config on it and realized how hot this thing get. Well, these cWWk X86-Px cases aren't the best case for passive cooling either.My Topton/CWWK fanless N100 arrived with thermal paste that looked like it had been applied with a tablespoon. Way too much of it and not very good at that. First thing I did was remove it all and re-apply my own with just enough to do the job. Temps dropped by at least 10C at idle and probably more like 20C under load.
Yeah, the worst unit I had was a Qotom J4125 with 5 network ports, I had to add a 0.5mm shim. The best was a Qotom Kaby Lake Celeron 3865U.There was a lot of talk on early units (both mini pc & itx) of the cpu die to copper shim/plate being too big a gap for everyday thermal paste. Quality paste helped but some were getting big gaps. Paste intends <= 0.2mm. Depending on the unit design some people fixed that with copper shims or filing mounting studs.
Probably fast and cheap manufacturing process and therefore large tolerances. In short: avoid first revision, repaste cpu and add a pad or a shim if necessary. And if they are broken, then they are broken, don't cry over them.It's not clear if the design & build quality improved over time. It might simply be a poor consistency problem in the manufacturing.
I started with one 0.2mm honeywell PTM layer first and when I screw down the heatsink or heatsink router type case I then undo those same screws and lift it up and see if its left an imprint on the honeywell pad or not.What thickness of the Arctic TP-3 did you take, 0.5, 1 or 1.5mm?
Going to test this on my N100.
Thanks, going to order a 0.5 arctic and stack if needed. The Honeywell is not available in retail here, maybe from ali or somewhere but not in local online shops.I started with one 0.2mm honeywell PTM layer first and when I screw down the heatsink or heatsink router type case I then undo those same screws and lift it up and see if its left an imprint on the honeywell pad or not.
If no imprint than just build it up from there so add another 0.2mm pad on top. Some use the bright torch method but I find it fiddly so just check for the cpu core imprint just to be 100% sure.
I think even if you used an 0.5 artic pad it would squash down pretty well anyhow, you could monitor idle/load cpu temps by running prime95 or intelburn in to check for stability afterwards also.
drive.google.com
drive.google.com
| CPU Thermal State Core 0: 36°C | Core 1: 36°C | Core 2: 36°C | Core 3: 36°C |
| Cooling |
| Fan Speed(s) Fan3: 900 RPM |
Not to ask the same question again, but have you solved anything regarding these issues? I'd also had no problems since December of last year, and on September 4 I started developing random reboots on my proxmox server that's also a Topton model. Ran memtest for a few hours and never had any errors - and I've been using a meanwell PSU the entire time.Did you ever end up getting this resolved? After 9 solid months of operation, my N100 unit running proxmox randomly reboots throughout the day as well. Sometimes I stay up for 2 days, other times 4 hours. I replaced the power supply with a genuine meanwell. Waiting on new ram to try that too.
what ram you using.. i had to mod bios to get corsair ram stableNot to ask the same question again, but have you solved anything regarding these issues? I'd also had no problems since December of last year, and on September 4 I started developing random reboots on my proxmox server that's also a Topton model. Ran memtest for a few hours and never had any errors - and I've been using a meanwell PSU the entire time.
I'm using G.Skill Ripjaws - I think it's 5600 mt/s but I'll have to open it up to double check. It's just strange that it's been working flawlessly since December and now all of a sudden it's all weird.what ram you using.. i had to mod bios to get corsair ram stable
hmm no in my case i always had random crashed with corsair ram.. sometime multiple a day sometimes after 2 weeks..I'm using G.Skill Ripjaws - I think it's 5600 mt/s but I'll have to open it up to double check. It's just strange that it's been working flawlessly since December and now all of a sudden it's all weird.
Hm - I'd be willing to tinker with the BIOS and see if that's causing it - can you point me in the right direction?hmm no in my case i always had random crashed with corsair ram.. sometime multiple a day sometimes after 2 weeks..
I had to modify the bios to unlock the ram options as they were not available in mine.. then once done i set speed to 4400 instead of 4800.. used this to mod bios [Guide] Usage of AMI’s AptioV UEFI Editor + FPT Flash MethodHm - I'd be willing to tinker with the BIOS and see if that's causing it - can you point me in the right direction?
