Fortinet added two models to its FortiGate G Series: the 3500G for high-density data centres and the 400G for the enterprise edge. Both run FortiOS 8.0 on Fortinet’s NP7 and SP5 ASICs.
The numbers
The 3500G is rated at 595 Gbps firewall throughput, 163 Gbps IPsec VPN, 105 Gbps threat protection, and 179 million concurrent sessions. The 400G comes in at 164 Gbps firewall, 55 Gbps IPsec, 13 Gbps threat protection, and 28 million concurrent sessions. Both add hardware-enforced firmware validation, and both carry native shadow AI detection, giving visibility into unsanctioned AI application use on the network.
Why this shows up in security designs
FortiGate sits at the edge of a large share of Canadian mid-market and municipal networks, which means it is often the box that video, access control, and building-system traffic must traverse to reach monitoring centres or the cloud. Two line items here matter for those designs. Enhanced encrypted-traffic inspection means the firewall can keep visibility as camera-to-cloud and access-to-cloud traffic goes fully TLS. And the session capacity matters because video systems are session-heavy in ways ordinary office traffic is not; a few hundred cameras streaming, plus health-check chatter, consumes connection state fast.
The read
For integrators, the practical takeaway is to size the firewall for threat-protection throughput, not the headline firewall number, because that is the figure that applies once inspection is actually turned on. The gap between the two on any vendor’s datasheet is where undersized deployments come from.