CWWK/Topton/... Nxxx quad NIC router

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

fta

Active Member
Feb 19, 2017
155
163
43
94
sometimes I build pkgs in poudriere (a lot of pkgs) for an update and it make the load almost max for hours.
Just know that your routing performance will be pretty bad during this time.

IIRC, PL2 caps the short bursts(turbo) while PL1 limits the TDP wattage over extended loads. If the benchmark lasts more than a few seconds, higher PL1 values should contribute more to higher scores.
This is correct. A PL1 of 8 or 10W is ridiculously low. The nominal TDP of the N305 is 15W.
 
  • Like
Reactions: Stovar and lucker

idle_user

Member
Jun 24, 2023
76
44
18
Yeah I installed promox and ran a quick vm with no issues. I ran the memtest after because I forgot to do it before getting started.

So now that I see the errors, I'm not feeling so confident about continuing :confused:

Follow-up:

I let it run memtest while I was away and came back to a crashed system and the screen read that the RAM temp was 96C. Uh-oh?
 

lucker

Member
May 28, 2023
29
26
13
Just know that your routing performance will be pretty bad during this time.
Thanks for heads up, haven't thought about that. I'm moving my VPS installation and have never bothered myself with such an issue. Usually, all services have been performing without any hiccups during building-time. I'll try to nice it or something similar. Thanks once again!

IIRC, PL2 caps the short bursts(turbo) while PL1 limits the TDP wattage over extended loads. If the benchmark lasts more than a few seconds, higher PL1 values should contribute more to higher scores.
This is correct. A PL1 of 8 or 10W is ridiculously low. The nominal TDP of the N305 is 15W.
So, you are saying that it's pretty safe to try PL1 up to 15, is my take on this correct? What would be a safe limit for PL2?
I'm learning a lot from the thread, thank you guys, thank you.
 

fta

Active Member
Feb 19, 2017
155
163
43
94
So, you are saying that it's pretty safe to try PL1 up to 15, is my take on this correct? What would be a safe limit for PL2?
I'm learning a lot from the thread, thank you guys, thank you.
The case has some amount of heat it can dissipate, and the only way to know is to test. Set it to 15W, run the prime95/small fft, and watch the temp. If the temp stops rising, the case can dissipate that much. If the temp continues to rise until you thermal throttle (might even take hours to do it), it can't. Personally, I think running a CPU no higher that 50 or 60C is a waste of the CPU.
 
  • Like
Reactions: lucker

lucker

Member
May 28, 2023
29
26
13
Ran passmark and got a 10821. Still don't know what to do about the memtest failures though.
As I understand, memtests are designed to reveal the issues with RAM. The pages on Aliexpress for similar diveices (at least the Topton ones) warn than not all RAM and SSD are compatible with such motherboards. That's the reason people here share the exact SKU numbers for SO-DIMMs and SSDs. I risked with Apacer SSD no one tried before and got lucky.

There are at least two possibilities in your case:
  • RAM module is faulty (it happens, though rarely, that maybe this kind of luck)
  • This particule module or the SKU in general has compatibility issue with the MB (some chips are picky to the MBs)
Test the RAM module on another machine, even better on several different. Or return the chip to the seller as a faulty one with screenshot of failed memtest (they normally should test it for a decision). Or Buy another one that was mentioned in this thread, in this case you have untested spare RAM module for future endeavors.

Personally, I have encountered issue with RAM once, but the RAM SKU was cheap (perhaps noname, don't remember) and MB was MSI or something (not top hardware, it was long time ago).
 

idle_user

Member
Jun 24, 2023
76
44
18
As I understand, memtests are designed to reveal the issues with RAM. The pages on Aliexpress for similar diveices (at least the Topton ones) warn than not all RAM and SSD are compatible with such motherboards. That's the reason people here share the exact SKU numbers for SO-DIMMs and SSDs. I risked with Apacer SSD no one tried before and got lucky.

There are at least two possibilities in your case:
  • RAM module is faulty (it happens, though rarely, that maybe this kind of luck)
  • This particule module or the SKU in general has compatibility issue with the MB (some chips are picky to the MBs)
Test the RAM module on another machine, even better on several different. Or return the chip to the seller as a faulty one with screenshot of failed memtest (they normally should test it for a decision). Or Buy another one that was mentioned in this thread, in this case you have untested spare RAM module for future endeavors.

Personally, I have encountered issue with RAM once, but the RAM SKU was cheap (perhaps noname, don't remember) and MB was MSI or something (not top hardware, it was long time ago).
Is there a compiled list of compatible SKU available? I reached out to the seller to ask what they recommend, but no response so far.

Hard to believe a brand new stick of Corsair RAM is faulty. And unfortunately I don't have any other system that supports ddr5 sodimm to test it on.

I do have a Crucial stick arriving tomorrow. Hopefully no issues with that one. And if there are issues, then it's definitely not the RAM.. so I'd have to deal with the seller over a faulty chip or buggy BIOS.
 

Becks0815

Well-Known Member
Oct 15, 2022
225
284
63
As I understand, memtests are designed to reveal the issues with RAM. The pages on Aliexpress for similar diveices (at least the Topton ones) warn than not all RAM and SSD are compatible with such motherboards. That's the reason people here share the exact SKU numbers for SO-DIMMs and SSDs. I risked with Apacer SSD no one tried before and got lucky.
Another possible reason, but take it with a big grain of salt....

As far as I know the GPU shares the memory with the rest of the machine, and either the GPU or the CPU can run into allocating the wrong memory block if you use a too big RAM module, resulting in a random crash at some point, because the same memory block gets allocated for both (GPU and the OS/programs), I remember I read something about these errors in another thread.

My machine with 16GB is rock stable. Since the last reboot of the FW I pushed 20TB data without issues around, and I haven't had any unwanted reboot or crash since I bought the machine.
 
Last edited:
  • Like
Reactions: filister and lucker

blunden

Active Member
Nov 29, 2019
502
162
43
Hard to believe a brand new stick of Corsair RAM is faulty. And unfortunately I don't have any other system that supports ddr5 sodimm to test it on.
New sticks being faulty can happen from time to time, or it could simply be that the stick doesn't work properly with the timings, subtimings and other related settings that the motherboard selects. After all, the proper values are dependent on the individual memory stick, the IMC and the motherboard's memory trace layout. It could also be due to high temperatures.

The large motherboard manufacturers spend quite a bit of resources optimizing the memory trace layout and BIOS/UEFI for optimal memory performance and support. I highly doubt that these smaller manufacturers spend the same development time to their boards and firmware.
 
  • Like
Reactions: idle_user

lucker

Member
May 28, 2023
29
26
13
Is there a compiled list of compatible SKU available? I reached out to the seller to ask what they recommend, but no response so far.
The closest thing to this is this thread. You could compile one from the listed SKUs here and make a post, for instance, then there would be one :)
My guess is that manufacturer doesn't test much of RAM modules and SSDs to begin with, that's probably why Topton recommend to use etsablished brands lins Samsung, Crucial, etc (please refer to the Topton device page for details), I suppose that Topton somehow get some info from CWWK or around.
 
Last edited:

idle_user

Member
Jun 24, 2023
76
44
18
Alright so I'm back with an update.

Recap:
Ordered an i3-n305 baremetal from topton on AE. Placed in a 1tb nvme and Corsair 16gb (CMSX16GX5M1A4800C40) RAM and got it to boot easily. Installed promox and ran a small vm no problem. Issues started when I decided to run memtest86 for peace of mind.

Memtest86 on the CMSX16GX5M1A4800C40 spit out errors within 5 minutes (under 60c), eventually it overheated, and froze. Did more tests but same issues.

Do not have another machine to test the RAM to blame it, so decided to order a new stick from a different brand.

Follow up:
32gb Crucial ddr5 came in (CT32G48C40S5), cleared CMOS, and booted up memtest86 on a USB. Ran fine for about 15 minutes but then the box restarted. This happened about 3 times, that's when I tracked that it happened exactly at the start of memtest86 Test 8. Temps were fine, and I even blew a fan on it one run to make sure of that.

Knowing of the consistent restarts, I configured the test to only run Test 8. And yep, immediate reboots from the start of it.

I then configured for each test to run EXCEPT for Test 8. Ran 2 successful passes before I stepped away. So everything looks good there.

So not sure what to take from this. Maybe the first stick of RAM was truly faulty, or it just didn't play well with the firmware. I think I'm better off with the Crucial RAM since it didn't spit out errors, but the reboots on memtest86 Test 8 is worrisome.


I will probably consolidate all my previous comments and make a new thread about my experience later.
 

ohartl

New Member
Jul 16, 2023
1
0
1
Did anyone check or know a compatible LTE/5G adapter that works on the internal M.2 E slot?
 

lucker

Member
May 28, 2023
29
26
13
32gb Crucial ddr5 came in (CT32G48C40S5), cleared CMOS, and booted up memtest86 on a USB. Ran fine for about 15 minutes but then the box restarted. This happened about 3 times, that's when I tracked that it happened exactly at the start of memtest86 Test 8. Temps were fine, and I even blew a fan on it one run to make sure of that.

Knowing of the consistent restarts, I configured the test to only run Test 8. And yep, immediate reboots from the start of it.
The first search result Memtest86 Failing Test 8 where guys are troubleshooting the similar error. See the path they went and try to replicate to narrow down the issue.
I'm not an engineer, but seems like you excluded the RAM module fault (statistically speaking), so the most probable cause is the MB. Becks0815 suggested another possibility. What I would do is to flash BIOS anew, re-download and re-make memstick with Memtest86+ and test again. If the same outcomes occur then I would grudgingly contact the seller to replace the faulty box (or die trying).

P.S. Depending on your use case you may decide to proceed with the existing issue, I frankly do not know the possible troubles down the road you may encounter. For my use case the issue would be unacceptable.
P.P.S. Just in case, examine carefully the SO-DIMM slot, once I have a tiny hair in it when upgrading an acient desktop and it was acting weirdly sometimes, BSODing and other peculiar things. Probably not the case for you but worth to check.
 
Last edited:

lucker

Member
May 28, 2023
29
26
13
I did contact CWWK for the BIOS files, and they sent me a link - I asked if this will work for i3-N305 because the file names say N100, and chat confirmed it would. I don't know how long this link will work for so I have of course saved it elsewhere, I can't upload it here because it is 7.7mb and the forum throws an error saying it is too large for the site to process.
http://www.7immy.com:5244/d/主板BIOS/英特尔系列BIOS/CW-12代主机-转接板款-N95-N100-N200-N305/CW-N100.rar?sign=xqCktRzniFvuJFAIVtnk9q19MSXOzeDDH9_MQ3iISyI=:0
Did anyone flush the BIOS from the link abdullah has mentioned? I would like to try the 4xNVMe board that came with ordered box, but that involves flushing special version of BIOS. In case the PCB doesn't work or doesn't meet the expectations I would very much prefer to roll back to the basics.

Now my box has BIOS 2.22.1287 (arrived with it). What is the version by the mentioned link?
 
Last edited:

idle_user

Member
Jun 24, 2023
76
44
18
The first search result Memtest86 Failing Test 8 where guys are troubleshooting the similar error. See the path they went and try to replicate to narrow down the issue.
I'm not an engineer, but seems like you excluded the RAM module fault (statistically speaking), so the most probable cause is the MB. Becks0815 suggested another possibility. What I would do is to flash BIOS anew, re-download and re-make memstick with Memtest86+ and test again. If the same outcomes occur then I would grudgingly contact the seller to replace the faulty box (or die trying).
Yeah I'm trying to rule out everything I can. I have attempted to reach out to the TopTon on AE but I just keep getting the automated messages. I will most likely begin a dispute with AE.

I will wait for the seller to provide troubleshooting instructions, and hopefully a new BIOS update. I do not want to gamble damaging the system even more because then I'd be at fault.
 

sos_nz

New Member
Mar 24, 2023
7
5
3
I'm watching this thread with interest, since I've just ordered an N100 passively cooled box from AE to run bare metal OPNsense.

https://www.aliexpress.com/item/1005005593329224.html

It's an older looking case, but says it supports DDR5 - so fingers are crossed it works as a replacement for my current HP 6300SFF w/i5-3470T and T4 i350 NIC. I have a Crucial P3 nvme drive and 16GB Crucial DDR5 4800 ready to install.
 

gordinho

New Member
Jul 17, 2023
2
0
1
Hi everyone... Read the thread to catch up but there's so many links and info, it's hard to keep track of everything. I'd like to replace my old pfsense box with an n100 since I need PPPoE 1gb/s. I'll probably promox it so I can put home assistant on it too. I'll probably eventually move the HA elsewhere as I'd like to avoid single point of failure.

Anyway, at Ali express seems like I now see from Topton a cheaper n100 and a more expensive one where the difference is just the chassis, I believe commonly called C type here? Now, for the the same price (300cad) of the more expensive one seems like there is now a 5 port eth that is already setup to host two nvme inside, which would be my preference as I'd probably raid 1 it for backup purposes. Alternatively, can promox backup to a cloud provider on a regular basis?

I also don't know anymore... But is ordering directly from cwwk better? I'm in Canada and ordering from a place where customs aren't a hassle would be great . I'm hoping to order during vacation time since I suspect it'll take weeks...

Part of me really wants just a two eth nic device able to push up to 2gb/s PPPoE to replace my current pfsense box but on the other hand love the idea of running a few qol VMs in a promox environment...
 

ohm

Member
Jun 10, 2023
43
28
18
I'm watching this thread with interest, since I've just ordered an N100 passively cooled box from AE to run bare metal OPNsense.
If you're planning to run OPNsense bare metal, you may wish to wait for a future release of OPNsense with at least FreeBSD 13.2 due to this issue.
 
  • Like
Reactions: Stovar