Performance Index
Ethernet
| Claim | Technologies | Specification Details or Systems Measured | Measurement | Claim/Test Date |
|---|---|---|---|---|
| PWR1: Intel Ethernet Network Adapter E835-CQDA2 is 1.9× higher performance per watt vs Nvidia ConnectX-6 DX (CX614106A) & ~1.4× higher performance per watt vs Broadcom BCM957508-P2100G under load given line rate bidirectional performance | Intel Ethernet Network Adapter E835-CQDA2 Nvidia ConnectX-6 DX (CX614106A) Broadcom BCM957508-P2100G | The system under test (SUT) was configured as a dual-socket Xeon 6 B2B platform using a Supermicro SuperServer SYS-222HA-TN with BIOS version 1.4 (July 2025). The system consisted of six nodes, each with two sockets populated by Intel® Xeon® 6960P processors (72 cores at 2.7 GHz with 432 MB cache). Hyper-threading was disabled while turbo mode remained enabled. Each node was equipped with 196 GB of memory using 12x16 GB DDR 4800 MT/s Kingston SRx8 RDIMMs. The system ran Rocky Linux 9.7 with kernel version 5.14.0-611.5.1.el9_7.x86_64. Operating system optimizations included disabling the firewall and user space IRQ balancer, while maintaining the default cpupower profile. CPU and power management settings were tuned for latency optimization, with virtual NUMA disabled, Energy Performance Bias set to extreme performance, SpeedStep enabled, turbo mode enabled, hardware P-states in native mode, and deeper C-states disabled (C6 and C1 demotion disabled). Testing was conducted in April 2026 using Netperf 2.7.0 with custom scripts, including n-perf_multi.sh for scaling thread counts and n-perfmon_2Port.py for command execution with statistics monitoring; these scripts were available via a git repository. Each traffic item used eight client threads, and TCP stream/MAERTS tests were run with a 64 KB message size. Power was monitored using Intel® DG Monitor PMC version 3.0.0.2 (internal), while temperature data was collected through the /sys/class/hwmon interface. For IP forwarding validation, driver and firmware configurations were verified against both Rocky Linux 9.7 inbox drivers and the latest vendor releases. The Intel® E835-CQDA2 adapter used NVM version 1.00 (0x00180E0) with the ice driver version 2.5.4 (26R0 PC). The NVIDIA ConnectX-6 DX adapter (CX614106A) used firmware version 22.48.1000 with driver version 26.01-1.0.0. The Broadcom Thor BCM957508-P2100G used firmware version 228.1.111.0 (package 236.1.153.0) with driver version 1.10.3-236.1.155.0. No driver tuning was applied, including CPU affinity adjustments, and the adapter was placed on NUMA node 3 in a PCIe Gen 5 x16 slot. Connectivity was established using QSFP56 1-meter direct attach cables in a back-to-back configuration between ports, with no switch in the data path. Performance measurements were conducted under bidirectional traffic conditions, achieving approximately 370 Gbps payload throughput per adapter at line rate using eight netperf threads in each direction. Idle power measurements were recorded after the system had been powered on for five minutes, and incremental power was calculated as the difference between loaded and idle conditions. Temperature data was collected using the system hardware monitoring interface. Test by Intel as of April 2026. | • Intel® Ethernet Network Adapter E835-CQDA2 | 32 Gbps/W | Baseline • Nvidia ConnectX-6 DX (CX614106A) | 17 Gbps/W | Intel = 1.88× higher • Broadcom BCM957508-P2100G | 23 Gbps/W | Intel = 1.39× higher | Apr-26 |
| PWR2: Intel Ethernet Network Adapter E835-CQDA2 consumes ~47% lower power than Nvidia ConnectX-6 DX (CX614106A) and ~28% lower power than Broadcom BCM957508-P2100G under load at full 200G line rate | Intel Ethernet Network Adapter E835-CQDA2 Nvidia ConnectX-6 DX (CX614106A) Broadcom BCM957508-P2100G | The system under test (SUT) was configured as a dual-socket Xeon 6 B2B platform using a Supermicro SuperServer SYS-222HA-TN with BIOS version 1.4 (July 2025). The system consisted of six nodes, each with two sockets populated by Intel® Xeon® 6960P processors (72 cores at 2.7 GHz with 432 MB cache). Hyper-threading was disabled while turbo mode remained enabled. Each node was equipped with 196 GB of memory using 12x16 GB DDR 4800 MT/s Kingston SRx8 RDIMMs. The system ran Rocky Linux 9.7 with kernel version 5.14.0-611.5.1.el9_7.x86_64. Operating system optimizations included disabling the firewall and user space IRQ balancer, while maintaining the default cpupower profile. CPU and power management settings were tuned for latency optimization, with virtual NUMA disabled, Energy Performance Bias set to extreme performance, SpeedStep enabled, turbo mode enabled, hardware P-states in native mode, and deeper C-states disabled (C6 and C1 demotion disabled). Testing was conducted in April 2026 using Netperf 2.7.0 with custom scripts, including n-perf_multi.sh for scaling thread counts and n-perfmon_2Port.py for command execution with statistics monitoring; these scripts were available via a git repository. Each traffic item used eight client threads, and TCP stream/MAERTS tests were run with a 64 KB message size. Power was monitored using Intel® DG Monitor PMC version 3.0.0.2 (internal), while temperature data was collected through the /sys/class/hwmon interface. For IP forwarding validation, driver and firmware configurations were verified against both Rocky Linux 9.7 inbox drivers and the latest vendor releases. The Intel® E835-CQDA2 adapter used NVM version 1.00 (0x00180E0) with the ice driver version 2.5.4 (26R0 PC). The NVIDIA ConnectX-6 DX adapter (CX614106A) used firmware version 22.48.1000 with driver version 26.01-1.0.0. The Broadcom Thor BCM957508-P2100G used firmware version 228.1.111.0 (package 236.1.153.0) with driver version 1.10.3-236.1.155.0. No driver tuning was applied, including CPU affinity adjustments, and the adapter was placed on NUMA node 3 in a PCIe Gen 5 x16 slot. Connectivity was established using QSFP56 1-meter direct attach cables in a back-to-back configuration between ports, with no switch in the data path. Performance measurements were conducted under bidirectional traffic conditions, achieving approximately 370 Gbps payload throughput per adapter at line rate using eight netperf threads in each direction. Idle power measurements were recorded after the system had been powered on for five minutes, and incremental power was calculated as the difference between loaded and idle conditions. Temperature data was collected using the system hardware monitoring interface. Test by Intel as of April 2026. | • Intel® Ethernet Network Adapter E835-CQDA2 | 11.68W | Baseline • Nvidia ConnectX-6 DX (CX614106A) | 22.15W | Intel = ~47% lower power • Broadcom BCM957508-P2100G | 16.19W | Intel = ~28% lower power | Apr-26 |