Join the future of AI

Live in the now

Scale for the future

8U Universal GPU Systems

Keeping your AI workload running swiftly requires lots of CPU cores and hard to find NVIDIA accelerated computing cards. Coastline Micro has powerful hardware solutions to run your project at every level, from entry to enterprise.  We have solutions architects waiting to will design the system perfect for your unique needs. We also build, test, and ship faster than the competition, with 90% of solutions shipping within a week.

Contact us today to bring your AI project to life!

AS -8125GS-TNHR (H13 Delta Next)

Server for AI

Optimized for machine learning through massive I/O capacity:

  • End-to-end nonblocking I/O to keep GPUs hydrated with data
  •  2-socket design supporting 4th Gen AMD EPYC™ Processors
  • Up to 24 DIMMs for up to 6 TB of DDR5-4800 memory
  • Flexible PCI-E 5.0 options for I/O and networking
  • Dual-zone cooling optimized for performance and lower operating costs
  • Titanium-Level efficiency power supplies

Accelerator Support

  • 8x H100 SXM5 GPU with NVLink

Processors Support

  • Dual EPYC GENOA Socket SP5 processors, up to 192 cores/384 threads

Memory Capacity

  • 24x DIMM slots, ECC DDR5 supports up to 4800MHz

PCIe Expansion

  • 8x PCIe 5.0 x16 low profile
  • (Optional) 4x PCIe 5.0 x16 FHHL

Drive Bays

  • 8x PCIe5.0 x4 NVMe U.2
  • (optional) 8x PCIe5.0 x4 NVMe U.2
  • 1x onboard NVMe M.2
  • 2x SATA 2.5″

System I/O

  • 1x VGA, 2x USB3.0, 1x Dedicated IPMI RJ45 port

System Cooling

  • 5x Front and 5x Rear counter-rotating fans with optimal fan speed control

System Power

  • 6x or 8x 3000-watt N+N Redundant Titanium Level Power Supplies

System Management

  • Built-in Server Management Tools (IPMI 2.0, KVM/media over LAN) with Dedicated LAN Port

Server Dimensions

  • 14.0″ (H) x 17.2″ (W) x 33.2″ (D)

SYS-821GE-TNHR (LCX13 Delta Next)

System Key Features

  • Modular by Design for Flexibility and Serviceability
  • Improved Thermal Capability
  • Rear I/O and Front I/O version (OCP option)

Power Savings

  • On system level, we can save up to 10-15% power when compared to air cooling
  • PUE lower to 1.1
  • Cooling cost in Datacenters will be reduced

Accelerator Support

  • 8 X NVIDIA HGX H100 80GB HBM3 SXM5


Processors Support

  • Dual Intel® 4th Gen Xeon® Scalable processors LGA-4677

Memory Capacity

  • Up to 48 DIMM slots, supporting DDR4 ECC RDIMM, LRDIMM, and Intel® Optane™ Persistent Memory modules

PCIe Expansion

  • Up to 14 PCI-E 4.0 slots for GPU or networking expansion

Drive Bays

  • 24 x 2.5″ hot-swap
  • 2 x M.2 NVMe


  • Dual 10GbE LAN ports with Intel® X550 controller, optional InfiniBand support

System Power

  • 4x 3000W Redundant Power Supplies, Titanium Level



Management (IPMI)

  • PMI 2.0 with virtual media over LAN and KVM-over-LAN support

Management Software

  • OOB Management Package (SFT-OOB-LIC)