CCR 1036 one cpu reaching 100, is it a problem?

Hardware MikroTik hardware general
1

How much traffic can a 1036 handle?
I have about 3gbps traffic and sometimes 1 of the cpus reaches 100.
If 1 cpu reaches 100, does this mean I need to upgrade the router?

> system resource cpu print detail 
 0 cpu="cpu0" load=23% irq=23% disk=0% 

 1 cpu="cpu1" load=11% irq=11% disk=0% 

 2 cpu="cpu2" load=11% irq=11% disk=0% 

 3 cpu="cpu3" load=11% irq=11% disk=0% 

 4 cpu="cpu4" load=28% irq=28% disk=0% 

 5 cpu="cpu5" load=12% irq=12% disk=0% 

 6 cpu="cpu6" load=12% irq=12% disk=0% 

 7 cpu="cpu7" load=5% irq=5% disk=0% 

 8 cpu="cpu8" load=11% irq=11% disk=0% 

 9 cpu="cpu9" load=25% irq=25% disk=0% 

10 cpu="cpu10" load=54% irq=47% disk=0% 

11 cpu="cpu11" load=23% irq=22% disk=0% 

12 cpu="cpu12" load=41% irq=41% disk=0% 

13 cpu="cpu13" load=27% irq=27% disk=0% 

14 cpu="cpu14" load=10% irq=10% disk=0% 

15 cpu="cpu15" load=54% irq=51% disk=0% 

16 cpu="cpu16" load=31% irq=31% disk=0% 

17 cpu="cpu17" load=42% irq=42% disk=0% 

18 cpu="cpu18" load=10% irq=10% disk=0% 

19 cpu="cpu19" load=7% irq=7% disk=0% 

20 cpu="cpu20" load=47% irq=8% disk=0% 

21 cpu="cpu21" load=7% irq=7% disk=0% 

22 cpu="cpu22" load=47% irq=47% disk=0% 

23 cpu="cpu23" load=23% irq=23% disk=0% 

24 cpu="cpu24" load=29% irq=29% disk=0% 

25 cpu="cpu25" load=53% irq=13% disk=0% 

26 cpu="cpu26" load=10% irq=10% disk=0% 

27 cpu="cpu27" load=22% irq=8% disk=0% 

28 cpu="cpu28" load=28% irq=27% disk=0% 

29 cpu="cpu29" load=14% irq=14% disk=0% 

30 cpu="cpu30" load=28% irq=10% disk=0% 

31 cpu="cpu31" load=4% irq=4% disk=0% 

32 cpu="cpu32" load=42% irq=2% disk=0% 

33 cpu="cpu33" load=11% irq=11% disk=0% 

34 cpu="cpu34" load=25% irq=11% disk=0% 

35 cpu="cpu35" load=49% irq=49% disk=0%




,

 > interface monitor-traffic vlan922
                         name:    vlan922
        rx-packets-per-second:     99 763
           rx-bits-per-second:  721.4Mbps
     fp-rx-packets-per-second:     99 763
        fp-rx-bits-per-second:  721.4Mbps
          rx-drops-per-second:          0
         rx-errors-per-second:          0
        tx-packets-per-second:    106 690
           tx-bits-per-second:  802.0Mbps
     fp-tx-packets-per-second:          0
        fp-tx-bits-per-second:       0bps
          tx-drops-per-second:          0
    tx-queue-drops-per-second:          0
         tx-errors-per-second:          0

 > interface monitor-traffic sfp-sfpplus1
                         name:  sfp-sfpplus1
        rx-packets-per-second:        89 164
           rx-bits-per-second:     904.5Mbps
     fp-rx-packets-per-second:        89 164
        fp-rx-bits-per-second:     904.5Mbps
          rx-drops-per-second:             0
         rx-errors-per-second:             0
        tx-packets-per-second:        44 032
           tx-bits-per-second:      59.4Mbps
     fp-tx-packets-per-second:        44 032
        fp-tx-bits-per-second:      59.4Mbps
          tx-drops-per-second:             0
    tx-queue-drops-per-second:             0
         tx-errors-per-second:             0

 > interface monitor-traffic sfp-sfpplus2
                         name:  sfp-sfpplus2
        rx-packets-per-second:       166 671
           rx-bits-per-second:     916.4Mbps
     fp-rx-packets-per-second:       166 671
        fp-rx-bits-per-second:     916.4Mbps
          rx-drops-per-second:             0
         rx-errors-per-second:             0
        tx-packets-per-second:       219 333
           tx-bits-per-second:     1860.3...
     fp-tx-packets-per-second:       219 333
        fp-tx-bits-per-second:     1860.3...
          tx-drops-per-second:             0
    tx-queue-drops-per-second:             0
         tx-errors-per-second:             0
2

check using tools profile to spot the source of cpu usage
cpu-usage-profile.png

3

Here is my tool profile. It shows cpu18 is using 51%, but what it does not show what is causing this cpu to use this amount.
tool profiel.PNG

4

looks like firewall (filter, mangle, nat, raw) and routing are the main contributors

in those sections are your chances to optimize cpu usage

if i read well yout interface total traffic is around 2.0 gbps, maybe there is some chance to improve

5

But this is the confusing part, My firewall rules are very simple and few. My routing table consists of only 400,835 routes.
My firewall:
Filter: Default Splynx filter rules
NAT: Default Splynx blocking rules only and src-nat for 4 x /27 client addresses
zero mangle rules or raw rules, zero layer 7 rules.