Hardware Sizing

MTA Server Hardware & CPU Capacity Calculator

Size CPU cores, RAM buffers, and NVMe IOPS required for PowerMTA, KumoMTA, or Postfix high-speed mail spooling.

Target Hourly Dispatch Volume 250,000 / hr

Average Message Size (KB) 45 KB

MTA Daemon Engine PowerMTA v5.5 (C++ Async)

TLS Encryption Enforcement 100% Mandatory TLS 1.3

Recommended CPU Cores

8 vCPU

3.2+ GHz Recommended

Dedicated RAM

16 GB

Queue in-memory buffers

Spool Disk IOPS Required

3,500 IOPS

Direct NVMe SSD required

Outbound Bandwidth

25.0 Mbps

Sustained wire egress

Hardware Profile Specification

For 250,000 msgs/hr with PowerMTA, allocate an 8-core Xeon/EPYC instance with 16GB RAM and NVMe storage. Format the spool volume with ext4 (noatime, nodiratime) for maximum file descriptor velocity.

Deploy Pre-Optimized Bare-Metal MTA Clusters

AutoSMTP provisions custom bare-metal nodes tuned for 1M+ emails/hour out of the box.

Deploy Dedicated MTA →

Tuning Operating System Kernels for High-Speed SMTP Spooling

Standard Linux server distributions are configured out of the box for general-purpose web workloads, not asynchronous high-volume email queuing. An MTA pushing hundreds of messages per second will rapidly exhaust file descriptors and TCP socket buffers if not tuned.

File Descriptors (nofile)

Every concurrent SMTP socket and spool file requires an active file descriptor. Set nofile limits to at least 100,000 in /etc/security/limits.conf.

Filesystem Mount Options

Disable file access timestamp updates on your MTA spool partition by mounting with noatime,nodiratime to reduce disk write amplification by up to 40%.

TCP Sockets & Ephemeral Ports

Expand the local port range net.ipv4.ip_local_port_range = 1024 65535 and enable net.ipv4.tcp_tw_reuse = 1 to prevent TIME_WAIT socket exhaustion.