Moore Threads logo

Moore Threads MTT SGX5000

rack
8× MTT S5000
Eight single-port 400Gb InfiniBand HDR or Ethernet CX7 adapters, one per accelerator, for scale-out; storage on a dual-port 100Gb RoCE adapter and management on gigabit Ethernet. Moore Threads publishes no scale-up fabric bandwidth for the chassis.
Power
11.4
kW Total
Memory
—
GB Total

Moore Threads' eight-card training and inference server, and the hardware underneath the KUAE cluster product. An 8U rack chassis holds eight OAM compute modules alongside two x86 or C86 processors, up to 2 TB of host DDR5-5600 RDIMM in a typical training build of 32 x 64GB or 24 x 96GB, two 960 GB SATA boot drives and four 3.84 TB PCIe Gen4 NVMe data drives. Each accelerator gets its own 400Gb InfiniBand or Ethernet port, so scale-out runs over standard fabrics. Moore Threads rates the machine at 11.4 kW and 100 kg. The specification table identifies the accelerator only as an OAM compute module, but Moore Threads publishes the same inference benchmark footnote on two pages, describing the test hardware once as six SGX5000 eight-card servers and once as six S5000 eight-card servers, which is the vendor equating the two directly. No throughput figure is shown because Moore Threads publishes none for either the server or the S5000: its public claims are relative to unnamed international flagship parts, and the absolute TFLOPS values on the S5000 page are reference baselines for other vendors' products. Total accelerator memory is likewise absent, since the vendor states no system total and the S5000's own capacity is not published on the Moore Threads site. The 11.4 kW is a rated figure, not a measured draw.

We will point you at suppliers who have it. Free, and no signup.

System Details

GPU Configuration

GPU Count: 8 GPUs
Architecture: MUSA 4th Gen
Interconnect: Eight single-port 400Gb InfiniBand HDR or Ethernet CX7 adapters, one per accelerator, for scale-out; storage on a dual-port 100Gb RoCE adapter and management on gigabit Ethernet. Moore Threads publishes no scale-up fabric bandwidth for the chassis.

System Specifications

Form Factor: rack
Total Power: 11.4 kW
Total Memory: —
Memory Bandwidth: 12.8 TB/s

Powered by Moore Threads MTT S5000

This system utilizes 8 × Moore Threads MTT S5000 GPUs, each delivering exceptional performance for AI and HPC workloads.

Per GPU TDP

—

Per GPU Memory

80 GB

Process Node

—

Architecture

MUSA 4th Gen

Frequently asked questions

How much does the MTT SGX5000 cost?

Moore Threads sells the MTT SGX5000 on quote through OEMs and resellers, and Flopper has no live rental listing for its MTT S5000 accelerator to derive an hourly figure from. System rental pricing →

How much power does the MTT SGX5000 require?

Moore Threads rates the MTT SGX5000 at 11.4 kW, which is the figure to plan power delivery and cooling around; per-GPU draw is listed on the accelerator's own page.

How many GPUs are in the MTT SGX5000?

8. The MTT SGX5000 carries 8 Moore Threads MTT S5000 accelerators in a rack form factor, and every throughput figure on this page is the total across them. MTT S5000 specifications →

Documentation & Resources

Need this system?

Tell us the quantity, timeframe and region. We reply by email with who can supply it.

Flopper Spec Sheet

Moore Threads MTT SGX5000 specifications, generated from our database. Printable.

Vendor product page

Moore Threads MTT SGX5000 documentation

View ↗

Moore Threads MTT S5000 parameters disclosed: 80GB, 1000 TFLOPS single-card AI compute

Moore Threads • 2026-02-12

View Document ↗

Typical Use Cases

AI/ML Training
High-Performance Computing
Data Analytics

The Moore Threads MTT SGX5000 runs 8× MTT S5000 GPUs over Eight single-port 400Gb InfiniBand HDR or Ethernet CX7 adapters, one per accelerator, for scale-out; storage on a dual-port 100Gb RoCE adapter and management on gigabit Ethernet. Moore Threads publishes no scale-up fabric bandwidth for the chassis..

© 2026 Flopper.io - Compare the GPUs Powering AI