Mango DPU

Bottleneck freeJBOF system

BoostX™ NRT

Name: BoostX™ NRT
Brand: MangoBoost

The ultimate DPU for NVMe-oF target storage

Product Inquiry

Driving the next era of scalable storage.

BoostX™ NRT delivers line-rate NVMe-oF RoCE target performance. Embedded ARM core and integrated NVMe controller lead to headless eBoF or eBoD solutions.

Product Brief Spec Sheet

Complete system solution

NVMe-oF RoCE Target. Integrated NVMe controller. Local control plane via embedded ARM.

Extreme TCO Savings

Headless solution delivers 200Gbps throughput in 75W.

Delivering Scale

Supports up to 32 NVMe SSD drives.

Complete system solution

NVMe-oF RoCE Target. Integrated NVMe controller. Local control plane via embedded ARM.

Extreme TCO Savings

Headless solution delivers 200Gbps throughput in 75W.

Delivering Scale

Supports up to 32 NVMe SSD drives.

CPU free. Bottleneck free.

BoostX™ NRT eliminates need for host CPU and system DRAM.

We took out the system’s host CPU and DRAM from the data path.
BoostX™ NRT drives pure line-rate 200Gbps directly to your SSDs.

Before

CPU handles all data I/O, causing bottlenecks.

After ｜ With BoostX™ NRT

Achieves Headless JBOF by eliminating CPU/DRAM.

The leanest path from Ethernet to Flash.

BoostX™ NRT runs the NVMe over RoCEv2 target in hardware: Standard Ethernet in, PCIe Gen5 NVMe out, and no server CPU or DRAM in between.

Standard Ethernet in, NVMe out, and a headless target in between. Cost falls at every stage of the path.

Standard Ethernet

RoCEv2 · in

any switchany NICany optic

KESAR™ Enclosure · Headless JBOF · 2U

BoostX™ NRT

target · canister A

BoostX™ NRT

target · canister B

1.6Tbps · HA

Server CPUDRAM

no server CPU or DRAM

NVMe SSD

26 × PCIe Gen5 · out

Standard Ethernet

RoCEv2 runs on standard Ethernet, so you are free to choose any switch, NIC, and optics vendor. That avoids the InfiniBand premium and the single-vendor lock-in it carries.

Network fabric cost · illustrative

Lower

Higher

NRT (whitebox Eth)InfiniBand

Headless Offload

BoostX™ NRT offloads the NVMe over RoCEv2 target into hardware, so the enclosure carries no server CPU, DRAM, or software stack. It is a true headless JBOF, immune to memory price swings.

Compute overhead at same capacity · illustrative

Lower

Higher

NRT (headless)Server target

NVMe Efficiency

PCIe Gen5 NVMe attaches directly, with no SAS HBA, expander, or RAID layer. At equal performance it draws far less power in far less rack space.

Power at same performance · illustrative

Lower

Higher

NRT (NVMe)SAS target

All figures are illustrative deltas: whitebox Ethernet vs InfiniBand, Gen5 NVMe vs 24G SAS at equal performance, and server BOM estimates with CPU tax verified at zero.

Performance

From microseconds to petabytes, One drive for every workload.

Full-Duplex architecture. Read and Write, simultaneously unleashed. A 5:5 mixed workload boosts IOPS by 73% and throughput by 83% beyond single-direction limits.