On site test facilities available for your use. New and improved technologies are emerging all the time, and this can be a daunting situation for customers planning their future projects. Making the right decision about new hardware is a difficult proposition, made even harder when clients are unable to test and understand the hardware first before making their purchase.
In some instances, the only way to know if a product or solution meets your needs is to try it. Our technical team are ready to help build full solution stacks on cloud-based lab systems or even on loan hardware that can be brought to your environment to ensure the right product is purchased for the right task. Our R&D lab - Boston Labs hosts a wide variety of the latest hardware and software solutions ready for evaluation either remotely or hand in hand with one of our experienced field application engineers
Our experienced team can assist with bring up and optimisation of both hardware and software elements of solutions. Advice on the best product selections and tuneables delivering the best performance, power consumption and cost is on tap when you book a session at Boston labs.
|New Product Development & POC
|Collaboration Testing and Development
|Support Centre and Debug
|Test bed for new products and solutions
|Latest current technologies
|Facility for Customers and suppliers to bollaborate, test, debug, and develop products
|As well as development, support and debug scenarios facilitated
|Proof of concept sandbox environment for multi-layer solutions
|Unreleased technologies (2+ years in advance)
VIEW COMPONENTS AVAILABLE TO TEST
SORT BY WORKLOAD
Remotely accessible 2U dual-node rackmount solution based on AMD EPYC or AMD Ryzen Threadripper PRO for the most demanding workstation applications.
2U dual-node rackmount chassis
AMD EPYC 7002/7003 or Threadripper PRO 5000WX series
Up to 2TB of DDR4-3200 memory per node
Up to 3 x RTX A6000 per node
IPMI Remote Management with iKVM and Virtual Media
This sleek workstation is capable of tackling most AI workloads at the convenience of your desk. Lower your footprint, increase your compute, and research using this system.
W-3400 and W-2400 Intel Sapphire Rapids CPU
6 single slot GPUs
Perfect for media & entertainment, life sciences, financial services, energy & geo-sciences, data science & AI development, and engineering
The Boston Stream RT is a 2U rack mount system optimised for a high volume of broadcasts with a smaller number of viewers demanding high-density transcoding.
AVC & HEVC supported
Up to 14x 4Kp60 Ultra-HD channels
Subdivide resolutions to support up to 336 channels
VQ equivalent to x264 & x265 Faster
Optimised for; Safe City, Social Video, eSports, Live Video
The AS-1125HS-TNR is a high-performance system tailor made for majority of HPC workloads like virtualisation, AI/ML, and cloud computing among other things.
4th Generation AMD EPYC™ 9004 Series CPU
24x DIMM slots
up to 6 TB
Flexible networking options
This system uses a Building Block design that features a single socket motherboard paired with an Ampere Altra CPU with up to 128 cores per server.
Single Socket Ampere® Altra® Processor or Max Processor
Up to 4TB ECC
10x 2.5" U.2 NVMe Hot-Swap Drive Bays
Based on Supermicro's 1U Ultra server range, this solution is a liquid-cooled server utilising 2-phase direct-to-chip technology and sporting AMD's EPYC™ 7003 range of processors, based on Zen 3 architecture.
Zutacore Hypercool 2-phase direct-to-chip liquid cooling
Dual AMD EPYC 7003 series CPU
Up to 8TB
Dual-port 10GBase-T networking
Based on Supermico's Big Twin server range, this solution is a liquid-cooled 4-node server, utilising 2-phase direct-to-chip technology and supports AMD EPYC™ 7003 range of processors.
AMD EPYC 7003 Series CPU
Up to 4TB
DDR4-3200 ECC memory
6 hot-swap drives per node
Liquid cooling technology is slowly but surely becoming more necessary thanks to increasing thermal demands or high-performing components. With liquid cooling you can reach higher TDPs while maintaining a low TCO.
Natural convection driven circulation
Climate independent with warm water cooling ability
Heat reuse potential optimised
Monitoring & control integrated
Technical 60kW, typical 32kW cooling capacity
Based on Supermicro's legendary BigTwin architecture, a 4-node, high density system featuring 3rd Gen Intel Xeon Scalable CPUs.
Dual 3rd Gen Intel® Xeon® Scalable CPUs
2U Rackmount System with 4 hot-pluggable nodes
Up to 4TB of DDR4 ECC Memory
2x M.2 Support
Up to 2 PCIe x16 4.0 HHHL expansion slots
This system supports up to 8 double-width GPUs Direct connect for maximum performance and is ideal for AI/Deep Learning, highlighting the power of the H100 GPU.
Supports up to 10 double-width GPUS
Dual AMD EPYC™ 7002 Series Processors
8TB Registered ECC DDR4 3200MHz SDRAM in 32 DIMMs
Up to 24 Hot-swap 2.5" drive bays; 4x 2.5" SATA drives and 4x 2.5" NVMe drives supported with included H/W
The Bow Pod16 is ideal for exploration. It provides all the power, performance and flexibility required to fast track the prototyping stage, rapidly moving to production.
16 x GC200 Mk2 IPUs
4 x IPU-M2000
4 petaFLOPS FP16.16, 1 petaFLOPS FP32
23,552 IPU Cores
The SYS-221BT-HNTR is an all-flash NVMe Hyperconverged infrastructure that is perfect to act as a container-as-a-Service, and application accelerator. It also works as a diskless HPC clustre.
4th Gen Intel® Xeon® Scalable processors
Up to 16 DIMMs ECC DDR5 Slots
Optional M.2 NVMe Boot Controller
The SYS-121H-TNR is perfect for virtualisation, AI/ML, software-defined storage, and act as an enterprise server. It is more than capable of handling most high-performance workloads.
Dual 4th Gen Intel® Xeon® Scalable processors
32 DIMM slots
up to 8TB of memory
Flexible networking options
RDIMMs up to DDR5-4800
Perfect for embarking on your AI journey, the new Boston ANNA Ampere S2 server features the power of up to 3 NVIDIA A30 double width GPUs.
2U Dual-node rackmount chassis
Single AMD EPYC™ 7003/7002 series processor (per node)
Up to 2TB registered ECC DDR4 3200MHz SDRAM in 8 DIMMs
Up to 3 NVIDIA® A30 GPUs (double width) (per node)
2 front hot-swap 2.5" U.2 NVMe Gen4 drive bays (per node)
NVIDIA's DGX H100 is their all-in-one AI and HPC solution. The system you can test at our labs specifically has eight H100 SXM5 NVLink GPUs.
32 petaFLOPS of FP8 compute
640GB of total GPU memory
2TB of system memory
8x NVIDIA H100 Tensor Core GPUs
PCIe 5.0 lanes
The highly versatile, configurable platform harnesses the power of the AMD Ryzen™ Threadripper™ PRO Processor with up to 64 cores and 128 PCI-E 4.0 lanes.
AMD Ryzen™ Threadripper™ PRO
64 Cores/128 threads
4.4 GHz Max Boost
4 x Dual Width Graphics Cards
PCIe Gen 4 support and M.2 storage
This system is a high density 4-node dual-processor server based on AMD’s high performance EPYC processors
dual SP3 socket motherboard
Up to 64 cores and 128 threads
4TB of DDR4-3200 ECC memory
Supports 4 hot-swap U.2 NVMe SSDs and 2 SATA drives
2x M.2 PCIe 4.0 interface per node
To help our clients make informed decisions about new technologies, we have opened up our research & development facilities and actively encourage customers to try the latest platforms using their own tools and if necessary together with their existing hardware. Remote access is also available
The annual GPU Technology Conference is returning once again in San Jose, California. See what awaits you!