Quantum Force CPU cores <=> virtual

[Wolfgang]

For some of the new Quantum force appliances curious virtual cores are shown in the datasheet. 10 physical and only 16 virtual. Should be 20 or not ?

9200 => 1 CPU, 4 physical cores, 8 virtual 

9300 => 1 CPU, 10 physical cores, 16 virtual

9400 => 1 CPU, 14 physical cores, 20 virtual

9700 => 1 CPU, 16 physical cores, 32 virtual

[Bob_Zimmerman]

One important note: the 6800 has ECC memory. While the 9300's processor supports ECC, it looks like the device as a whole does not:

[Expert@SomeFirewall:0]# dmidecode -t 16
# dmidecode 3.0
Getting SMBIOS data from sysfs.
SMBIOS 3.6.0 present.
# SMBIOS implementations newer than version 3.0 are not
# fully supported by this version of dmidecode.

Handle 0x003F, DMI type 16, 23 bytes
Physical Memory Array
	Location: System Board Or Motherboard
	Use: System Memory
	Error Correction Type: None
	Maximum Capacity: 128 GB
	Error Information Handle: Not Provided
	Number Of Devices: 4

[Timothy_Hall]

Would you mind posting an ethtool -i & ethtool -S against an onboard interface, and doing another ethtool -i & ethtool -S against one of your CPAC-8-1/10F-D line card interfaces?  Thanks!

[Bob_Zimmerman]

[Expert@SomeFirewall:0]# ethtool -i eth1
driver: net_e1000_igb
version: DPDK 20.11.7.4.0 (12 Jan 24)
firmware-version: 1.63, 0x800009fb
expansion-rom-version: 
bus-info: 0000:02:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

[Expert@SomeFirewall:0]# ethtool -S eth1
NIC statistics:
     ifs_ibytes_hi: 0
     ifs_ibytes_lo: 0
     ifs_obytes_hi: 0
     ifs_obytes_lo: 0
     ifs_ipackets: 0
     ifs_opackets: 0
     ifs_imcasts: 0
     ifs_omcasts: 0
     ifs_noproto: 0
     ifs_ibcasts: 0
     ifs_obcasts: 0
     ifs_linkchanges: 0
     ife_ierrors: 0
     ife_oerrors: 0
     ife_iqdrops: 0
     ife_oqdrops: 0
     iee_rx_missed: 0
     rx_q0_packets: 0
     rx_q1_packets: 0
     rx_q2_packets: 0
     rx_q3_packets: 0
     rx_q4_packets: 0
     rx_q5_packets: 0
     rx_q6_packets: 0
     rx_q7_packets: 0
     tx_q0_packets: 0
     tx_q1_packets: 0
     tx_q2_packets: 0
     tx_q3_packets: 0
     tx_q4_packets: 0
     tx_q5_packets: 0
     tx_q6_packets: 0
     tx_q7_packets: 0

[Expert@SomeFirewall:0]# ethtool -g eth1
Ring parameters for eth1:
Pre-set maximums:
RX:		4096
RX Mini:	0
RX Jumbo:	0
TX:		4096
Current hardware settings:
RX:		1024
RX Mini:	0
RX Jumbo:	0
TX:		1024

[Expert@SomeFirewall:0]# ethtool -i eth1-05
driver: net_ice
version: DPDK 20.11.7.4.0 (12 Jan 24)
firmware-version: 4.20 0x800178e2 1.3346.0
expansion-rom-version: 
bus-info: 0000:01:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: yes

[Expert@SomeFirewall:0]# ethtool -S eth1-05
NIC statistics:
     ifs_ibytes_hi: 0
     ifs_ibytes_lo: 0
     ifs_obytes_hi: 0
     ifs_obytes_lo: 0
     ifs_ipackets: 0
     ifs_opackets: 0
     ifs_imcasts: 0
     ifs_omcasts: 0
     ifs_noproto: 0
     ifs_ibcasts: 0
     ifs_obcasts: 0
     ifs_linkchanges: 0
     ife_ierrors: 0
     ife_oerrors: 0
     ife_iqdrops: 0
     ife_oqdrops: 0
     iee_rx_missed: 0
     rx_q0_packets: 0
     rx_q1_packets: 0
     rx_q2_packets: 0
     rx_q3_packets: 0
     rx_q4_packets: 0
     rx_q5_packets: 0
     rx_q6_packets: 0
     rx_q7_packets: 0
     rx_q8_packets: 0
     rx_q9_packets: 0
     tx_q0_packets: 0
     tx_q1_packets: 0
     tx_q2_packets: 0
     tx_q3_packets: 0
     tx_q4_packets: 0
     tx_q5_packets: 0
     tx_q6_packets: 0
     tx_q7_packets: 0
     tx_q8_packets: 0
     tx_q9_packets: 0
     tx_q10_packets: 0

[Expert@SomeFirewall:0]# ethtool -g eth1-05
Ring parameters for eth1-05:
Pre-set maximums:
RX:		4096
RX Mini:	0
RX Jumbo:	0
TX:		4096
Current hardware settings:
RX:		2048
RX Mini:	0
RX Jumbo:	0
TX:		512

Eleven transmit queues? Weird. Incidentally, the PCI topology for the interfaces is also kind of weird. Here it is sorted by PCIe address:

eth1-05  01:00.0  8086:1593 (rev 02)  16GT/s, Width x16
eth1-01  01:00.1  8086:1593 (rev 02)  16GT/s, Width x16
eth1-06  01:00.2  8086:1593 (rev 02)  16GT/s, Width x16
eth1-02  01:00.3  8086:1593 (rev 02)  16GT/s, Width x16
eth1-07  01:00.4  8086:1593 (rev 02)  16GT/s, Width x16
eth1-03  01:00.5  8086:1593 (rev 02)  16GT/s, Width x16
eth1-08  01:00.6  8086:1593 (rev 02)  16GT/s, Width x16
eth1-04  01:00.7  8086:1593 (rev 02)  16GT/s, Width x16

eth1     02:00.0  8086:1521 (rev 01)  5GT/s, Width x4
eth5     02:00.1  8086:1521 (rev 01)  5GT/s, Width x4
eth2     02:00.2  8086:1521 (rev 01)  5GT/s, Width x4
eth6     02:00.3  8086:1521 (rev 01)  5GT/s, Width x4

Sync     03:00.0  8086:1533 (rev 03)  2.5GT/s, Width x1

Mgmt     04:00.0  8086:1533 (rev 03)  2.5GT/s, Width x1

eth3     07:00.0  8086:1521 (rev 01)  5GT/s, Width x4
eth7     07:00.1  8086:1521 (rev 01)  5GT/s, Width x4
eth4     07:00.2  8086:1521 (rev 01)  5GT/s, Width x4
eth8     07:00.3  8086:1521 (rev 01)  5GT/s, Width x4

Evidently the onboard ports are split left-from-right, and arranged vertically like a switch. This isn't a problem, of course, just a little odd. The 1/10g card is arranged similarly, but it counts bottom to top rather than top to bottom.

All the PCIe links are huge. That 1/10g card is PCIe 4, and 16 lanes gives it 256g link rate to cover 80g of possible port link rate. Similarly, the onboard ports have 20g of PCIe link rate for 4g of possible port link rate. Other than the lack of ECC support, the 9300's internals seem excellent. Sure, it only has so much processor performance, but that should be the only performance limit.