Reference systems · Kuala Lumpur

Complete systems, configured per order.

Seraya supplies servers, workstations, storage and networking as complete systems. Single GPU cards and SXM modules are quoted separately on the GPU line, because they are not interchangeable with a finished system.

15 reference configurations · 5 categories 8 GPU models · 14 versions are listed on the GPU line

Buying single GPUs in bulk?

Single cards and SXM modules are quoted on the GPU line, separately from the complete systems on this page. A module needs an HGX baseboard; a card needs a free PCIe slot.

Systems supply

Seraya supplies complete systems — workstations, servers, storage and networking — on a B2B quotation basis. Configuration is confirmed per order. Prices are not published.

AI Servers

Seraya supplies AI servers configured to your workload. Card count, chassis, power and cooling are agreed per order and stated in the quotation.

Open line

AI Workstations

Workstations built around one or two accelerators, for teams that need the work off shared infrastructure and quiet enough for an office.

Open line

Gaming Systems

Built for resellers, retailers and esports venues buying in volume. This is a B2B supply line, not retail single-unit sales.

Open line

Storage & Networking

Storage and network infrastructure sized to match the compute it serves — pools and capacity tiers for the data, leaf/spine switching or InfiniBand for the fabric.

Open line

How a systems order runs

  1. 1Submit RFQ
  2. 2Confirm supply and quote
  3. 3Agree terms and payment
  4. 4Dispatch and delivery

Use the systems RFQ form, or email felix.mm@seraya.systems with the subject tag [Systems RFQ].

Reference configurations

Every configuration on this page is an example, not a validated build. Compatibility, firmware and availability are confirmed per order.

Compute

General-purpose rack nodes for virtualization, databases and mixed workloads.

SR-C220 — Dual-socket rack node Reference image
SR-C220 Dual-socket rack node

1U/2U rack server · Reference

CPU
2× Xeon Silver 4410Y (12C)
RAM
256 GB DDR5 ECC
Disk
2× 1.92 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-C450 — Virtualization host Reference image
SR-C450 Virtualization host

2U rack server · Reference

CPU
2× EPYC 9354 (32C)
RAM
768 GB DDR5 ECC
Disk
4× 3.84 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-C880 — High-density HPC node Reference image
SR-C880 High-density HPC node

Rack server · Reference

CPU
2× EPYC 9654 (96C)
RAM
1.5 TB DDR5 ECC
Disk
8× 7.68 TB NVMe
New / ITAD Quoted per item Get a quote →

GPU & AI servers

Complete servers configured for inference, fine-tuning and large-model training.

SR-G410 — Inference & fine-tune entry Reference image
SR-G410 Inference & fine-tune entry

GPU server · 2 cards · Reference

CPU
1× Xeon Gold 6448Y
GPU
2× L40S 48 GB
RAM
256 GB DDR5
Disk
2× 3.84 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-G780 — Multi-GPU training server Reference image
SR-G780 Multi-GPU training server

GPU server · 4× H100 NVL cards · Reference

CPU
2× EPYC 9474F
GPU
4× H100 NVL 94 GB (PCIe, NVLink-bridged pairs)
RAM
1 TB DDR5 ECC
Disk
4× 7.68 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-G990 — Large-model training node Reference image
SR-G990 Large-model training node

GPU server · 8 SXM modules · Reference

CPU
2× EPYC 9654
GPU
8× H200 SXM 141 GB (NVLink)
RAM
2 TB DDR5 ECC
Disk
8× 7.68 TB NVMe + IB HDR
New / ITAD Quoted per item Get a quote →

Storage

All-flash pools, capacity tiers and backup targets, NVMe through SATA.

SR-S240 — NVMe shared pool Reference image
SR-S240 NVMe shared pool

Shared all-flash pool node · Reference

Role
Shared block/file pool for VM and database workloads
Media
NVMe SSD, TLC, mixed-use endurance
Host
1× Xeon Silver 4410Y · 128 GB RAM
Drives
24× 7.68 TB NVMe
Raw capacity
≈184 TB
Interface
NVMe over PCIe 4.0; 100 GbE host ports
Redundancy
Agreed per order (RAID or erasure coding)
New / ITAD Quoted per item Get a quote →
SR-S560 — All-flash archive + replica Reference image
SR-S560 All-flash archive + replica

Capacity all-flash node · Reference

Role
Capacity tier and replica target for the primary pool
Media
NVMe SSD, read-intensive endurance
Host
2× EPYC 9334 · 512 GB RAM
Drives
36× 15.36 TB NVMe
Raw capacity
≈550 TB
Interface
NVMe over PCIe; 100/400 GbE host ports
Redundancy
Agreed per order (RAID or erasure coding)
New / ITAD Quoted per item Get a quote →
SR-S980 — High-density object / backup Reference image
SR-S980 High-density object / backup

High-density disk shelf node · Reference

Role
Object store and backup target for cold or regulatory data
Media
SATA HDD (nearline)
Host
2× Xeon Gold 6438N · 256 GB RAM
Drives
60× 22 TB SATA
Raw capacity
≈1.3 PB
Interface
SATA/SAS backplane; 25/100 GbE host ports
Redundancy
Agreed per order (parity or erasure coding)
New / ITAD Quoted per item Get a quote →

Networking

25 / 100 / 400 GbE switching and InfiniBand fabric for AI clusters.

SR-N100 — 25/100 G top-of-rack Reference image
SR-N100 25/100 G top-of-rack

1U leaf switch · Reference

Ports
32× 100 GbE QSFP28
Speeds
100 / 50 / 40 / 25 / 10 GbE per port
Switching capacity
6.4 Tb/s, cut-through
Protocols
Ethernet (RoCEv2 capable); L2/L3; VXLAN; MLAG
Redundancy
Dual PSU, redundant fans, MLAG pair
New / ITAD Quoted per item Get a quote →
SR-N400 — 400 G spine switch Reference image
SR-N400 400 G spine switch

1U/2U spine switch · Reference

Ports
32× 400 GbE QSFP-DD
Speeds
400 / 200 / 100 / 50 GbE per port
Switching capacity
25.6 Tb/s, non-blocking
Protocols
Ethernet (RoCEv2); BGP EVPN; VXLAN; PFC/ECN for lossless
Redundancy
Dual PSU, redundant fans, spine pair
New / ITAD Quoted per item Get a quote →
SR-N-IB — InfiniBand HDR / NDR fabric Reference image
SR-N-IB InfiniBand HDR / NDR fabric

InfiniBand switch and HCA set · Reference

Ports
40× NDR 400 Gb/s
Speeds
NDR 400 / HDR 200 Gb/s InfiniBand
Protocols
InfiniBand; SHARP in-network compute; GPUDirect RDMA
Cabling
Active optical or DAC, agreed per order
Redundancy
Dual-rail fabric and redundant PSU
New / ITAD Quoted per item Get a quote →

Workstation

Engineering and content workstations, quiet enough for the office.

SR-W310 — Engineering workstation Reference image
SR-W310 Engineering workstation

Single-socket tower · Reference

CPU
Threadripper 7960X (24C)
GPU
RTX 5000 Ada 32 GB
RAM
128 GB DDR5
Disk
2× 2 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-W620 — Creation & simulation flagship Reference image
SR-W620 Creation & simulation flagship

Dual-GPU tower · Reference

CPU
Threadripper Pro 7995WX (96C)
GPU
2× RTX 6000 Ada 48 GB
RAM
512 GB DDR5 ECC
Disk
4× 4 TB NVMe
New / ITAD Quoted per item Get a quote →
SR-W180-EDGE — Silent edge node Reference image
SR-W180-EDGE Silent edge node

Compact workstation · Reference

CPU
Ryzen 9 7950X
GPU
RTX 4000 SFF Ada 20 GB
RAM
64 GB DDR5
Disk
1× 4 TB NVMe
New / ITAD Quoted per item Get a quote →

Capacity shown is raw installed media (drive count × drive capacity), not usable host capacity. Usable capacity depends on the redundancy and file-system layout agreed for the order.

Get a quote

Give us the system type, how many systems, and the cards per system if that applies. The two counts are quoted separately.

What this enquiry covers

Requirement
Buy a complete system

Filled in from the page you came from, so you only complete the items below.

Model, configuration, or the requirement itself if you are still scoping it.

Optional, and separate from the number of systems.

Additional information optional

* required. The rest is optional and only shortens the back-and-forth.

Prefer email? Write to felix.mm@seraya.systems with the subject tag [Buyer RFQ]. Attachments are handled by reply — this form does not accept file uploads.

Two supply routes — and they are not priced the same

Comparison point New ITAD / certified pre-owned
Cost basis List or allocation pricing Recovered cost — normally a discount to new
Performance & life Full service life, current firmware Same silicon; shorter remaining life, older firmware
Lead time Driven by allocation and configuration Driven by what has been recovered and tested
Payment & warranty Deposit or credit terms Set case by case against the grading report

What changes the quote

  • System complexity

    A single node and a rack-scale rollout do not quote the same way.

  • Stock status

    In stock, on order and on allocation carry different lead times.

  • Goods condition

    New, refurbished and ITAD are priced, paid for and warranted differently.

All three are stated in writing in every quote.