Updated 2016 PfSense build with 4 Intel LAN,8GB RAM,120GB SSD, 10 watts - $257

[dsm]

re: rates. There is a big difference when the cpu supports AES-NI because the hw offloads a lot of the encryption work which would otherwise been done with many instructions using a software library. The i5-5250 (2015) is just a hair slower than the i7-4500u (2013). You can see here the i7-4500u can do 299Mbps. https://forum.pfsense.org/index.php?topic=115673.0. I've got 225Mbps service and I'd like to know if the i5-5250 could keep up too.

 

[sfx2000]

N3700 should get you close - but the i5-5250 is probably a very good choice - should get wireline speed, or close to it on a 255Mbps symmetric service...

N3700 numbers... i5-5250 will be even faster... the only concern I would have with the i5 - under full load, it might start throttling, so a good thermal solution there is pretty much a need - J1900 is a 10W TDP, N3700 is 6W, the 5250U is a 15W...

AES - Intel N3700

type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
aes-128-gcm     124969.99k   235895.98k   323049.81k   358767.96k   369295.36k
aes-256-gcm     113344.15k   207811.56k   274003.88k   300459.35k   307882.67k

 

[dsm]

Oh, that's nice, I'm only trying to hit 225Mbps and it's probably a bit cheaper. Is there an N3700 bare bones box available with more than one ethernet port? I did a quick google and couldn't find one. If the J1900 had AES-NI it would be perfect .

 

[sfx2000]

The i5 is likely a good choice....

Here's some 1037U numbers - which is a celeron - smaller cache, no AESNI - and it's big cores, e.g. IvyBridge, rather than the intel small cores - the newer small Intel cores do perform well though...

1037U's and J1800/1900's - check alibaba or the amazon - QOTOM is a good keyword I guess (I have nothing to do with them)

No AES - Intel 1037U (big cores, low power - IvyBridge)

aes-128-gcm      55538.44k    63789.50k   201876.38k   220349.44k   225894.40k
aes-256-gcm      42510.78k    47149.55k   155853.74k   171946.67k   176248.15k
aes-128-cbc     157385.09k   179165.99k   183619.50k   186146.47k   186933.25k
aes-256-cbc     116930.84k   129919.51k   132131.33k   133703.34k   134113.96k
sha256           24903.27k    63643.31k   120950.44k   156428.97k   171024.38k
sha512           17044.49k    68922.05k   121715.71k   184331.95k   217240.92k

gnutls-cli --benchmark-ciphers
Checking cipher-MAC combinations, payload size: 16384
        SALSA20-256-SHA1 168.88 MB/sec
        AES-128-CBC-SHA1 126.41 MB/sec
        AES-128-CBC-SHA256 90.64 MB/sec
             AES-128-GCM 101.67 MB/sec
       CHACHA20-POLY1305 0.22 GB/sec

Checking MAC algorithms, payload size: 16384
            SHA1 0.38 GB/sec
          SHA256 172.12 MB/sec
          SHA512 0.21 GB/sec

Checking ciphers, payload size: 16384
                3DES-CBC 12.83 MB/sec
             AES-128-CBC 186.97 MB/sec
             ARCFOUR-128 0.21 GB/sec
             SALSA20-256 0.30 GB/sec

 

[occamsrazor]

I received my Qotom Q355G4 with i5-5250U CPU, 64GB SSD, & 8GB RAM and have installed pfSense 2.4.
All seems to be working well. My connection is only currently 10MB so this machine is way overspecced but I knew that in advance

I have PowerD enabled and am using the "adaptive" setting. This lets the CPU drop from the base frequency 1.6Ghz to as low as 500Mhz speed for power saving when not busy which is mine most of the time. When you do something e.g. with the webGUI it quickly rises to 1.6Ghz.

Even on 500Mhz the CPU rarely goes above 10% and when it's at base 1.6Ghz it's about 2-3%. The temperature of the CPU is reported at around 45-53 degrees C depending on use.

I am trying to think of a way to stress-test the CPU just (out of curiosity) to watch it rise above 1.60Ghz in Turbo Boost but I can't get anywhere close.

 

[sfx2000]

if you just want to thermal stress the CPU --- ssh into the router, and go to the shell - (option 8)

type in the following -- yes > /dev/null

This will basically put one of the cores into a hard loop, and you'll see the load go to max on whichever core it lands on - to exit, just hit crtl-c

 

[occamsrazor]

if you just want to thermal stress the CPU --- ssh into the router, and go to the shell - (option 8)

type in the following -- yes > /dev/null

This will basically put one of the cores into a hard loop, and you'll see the load go to max on whichever core it lands on - to exit, just hit crtl-c

Nice tip! I tried that and could see the CPU usage rise in the pfsense dashboard. But do you know any way to verify the actual current CPU speed being used? I'd like to test that Turbo Boost actually does happen when under high CPU load. All the specs indicate the i5-5250U is much faster than the J1900 but if Turbo Boost isn't actually used then you are looking at a base frequency of 1.6Ghz only. Or am I getting confused?

 

[occamsrazor]

Nice tip! I tried that and could see the CPU usage rise in the pfsense dashboard. But do you know any way to verify the actual current CPU speed being used? I'd like to test that Turbo Boost actually does happen when under high CPU load. All the specs indicate the i5-5250U is much faster than the J1900 but if Turbo Boost isn't actually used then you are looking at a base frequency of 1.6Ghz only. Or am I getting confused?

Did some reading....

powerd -v

....shows the current speeds. It won't show actual Turbo Boost speeds though. But apparently the fact it shows speed as max+1, which in my case was 1601Mhz - this indicates Turbo Boost is in use.

Another way is to show the capabilities...

sysctl dev.cpu. | grep freq

which in my case gave:

dev.cpu.0.freq_levels: 1601/15000 1600/15000 1500/13791 1400/12754 1300/11738 1200/10742 1100/9626 1000/8676 900/7744 800/6833 700/5806 600/4936 500/4086

 

[sfx2000]

Nice tip! I tried that and could see the CPU usage rise in the pfsense dashboard. But do you know any way to verify the actual current CPU speed being used? I'd like to test that Turbo Boost actually does happen when under high CPU load. All the specs indicate the i5-5250U is much faster than the J1900 but if Turbo Boost isn't actually used then you are looking at a base frequency of 1.6Ghz only. Or am I getting confused?

Make sure your powerplan is set to "hiadaptive" vs. "adaptive"

Should see if the kernel has enabled est there...

grep -i speedstep /var/run/dmesg.boot​

Remember, turboboost only kicks in of the cpu state is in p0, and then only if thermal conditions and the conf (see above) allow it to.

 

[occamsrazor]

Make sure your powerplan is set to "hiadaptive" vs. "adaptive"

Does it have to be? Shouldn't adaptive also enter TB albeit maybe slower to do so?
Below is the result of "powerd -v" with "adaptive" and in the middle I enter a "yes > /dev/null" command:

load 15%, current freq 500 MHz (12), wanted freq 500 MHz
load 12%, current freq 500 MHz (12), wanted freq 500 MHz
load 0%, current freq 500 MHz (12), wanted freq 500 MHz
load 7%, current freq 500 MHz (12), wanted freq 500 MHz
load 10%, current freq 500 MHz (12), wanted freq 500 MHz
load 62%, current freq 500 MHz (12), wanted freq 826 MHz
changing clock speed from 500 MHz to 900 MHz
load 104%, current freq 1000 MHz ( 7), wanted freq 3202 MHz
changing clock speed from 1000 MHz to 1601 MHz
load 114%, current freq 1601 MHz ( 0), wanted freq 3202 MHz
load 104%, current freq 1601 MHz ( 0), wanted freq 3202 MHz
load 101%, current freq 1601 MHz ( 0), wanted freq 3202 MHz
load 101%, current freq 1601 MHz ( 0), wanted freq 3202 MHz
load 100%, current freq 1601 MHz ( 0), wanted freq 3202 MHz

I think this means it entered TB.... right?

Should see if the kernel has enabled est there...

grep -i speedstep /var/run/dmesg.boot​

Result of that is:

root: grep -i speedstep /var/run/dmesg.boot
est0: <Enhanced SpeedStep Frequency Control> on cpu0
est1: <Enhanced SpeedStep Frequency Control> on cpu1
est2: <Enhanced SpeedStep Frequency Control> on cpu2
est3: <Enhanced SpeedStep Frequency Control> on cpu3

Thanks for all the help...

 

[ikomrad]

Does this device support Gigabit ethernet speeds . I'm on the Centurylink Fiber plan.