Category top banner badge image

AMD Instinct Servers

Heterogeneous server solutions featuring AMD EPYC and AMD Instinct™ for HPC and AI applications.

value propositon

Vega 7nm Technology

Accelerators featuring the Industry’s first 7nm GPU design dramatically improve performance per watt over previous generation products.

value propositon

Wide AMD Platform Selection

Exxact offers a wide selection of server and infrastructure platforms to meet the compute needs of each customer's unique use case.

value propositon

Standard 3 Year Warranty

Have peace of mind, focus on what matters most, knowing your system is backed by a 3 year warranty and support.

Find the Right AMD Instinct Platform for Your Workload

Solution image
Solution value property image
Highly customizable
Solution value property image
Scalable for GPU compute
Base Specs
CPU1x AMD EPYC 7003 Series Processor
GPUUp to 4x AMD Instinct GPUs
STO8x 2.5"/3.5" Hot-swap Drive Bays
STO2x U.2 NVMe Drive Bays
Solution image
Solution value property image
For demanding HPC applications
Solution value property image
Balanced GPU & GPU performance
Base Specs
CPU2x AMD EPYC 7003 Series Processors
MEMUp to 1TB System Memory
GPUUp to 4x AMD Instinct GPUs
STO8x 3.5" Hot-Swap Drive Bays
Solution image
Solution value property image
Maximum GPU compute
Solution value property image
Excellent for AI & HPC
Base Specs
CPU2x AMD EPYC 7003 Series Processors
MEMUp to 4TB System Memory
GPUUp to 8x AMD Instinct GPUs
STO8x 2.5" Hot-swap Drive Bays

Not sure what you need?

Let us know what kind of project you have planned. We can help you decide.

Accelerate Your Discoveries

AMD Instinct™ MI100 accelerator is the world’s fastest HPC GPU, engineered from the ground up for the new era of computing. Powered by the AMD CDNA architecture, the MI100 accelerators deliver a giant leap in compute and interconnect performance, offering a nearly 3.5x the boost for HPC (FP32 matrix) and a nearly 7x boost for AI (FP16) throughput compared to prior generation AMD accelerators.

Technology

ROCm Open Software Platform for HPC

ROCm is the first open-source HPC/Hyperscale-class platform for GPU computing that’s also programming-language independent. ROCm incorporates the UNIX philosophy of choice, minimalism and modular software development to GPU computing. The new ROCm foundation lets you choose or even develop tools and a language run time for your application.

• Scalable, fully open source AMD ROCm software platform.
• Open source compiler, tools and libraries from the metal forward.

• Easier, more flexible programming model through

Additional ROCm documentation:


» ROCm Homepage
» ROCm Software Platform Page for Manual Installs
» ROCm Docker Hub for Containers

Partnered with Leaders in Innovation

Filters
Rack Height
1U (1)
2U (4)
4U (5)
See more
See less
Processor Manufacturer
AMD (5)
Intel (5)
See more
See less
Number of Processors Supported
1 (2)
2 (8)
See more
See less
Processor Socket
Socket SP3 (5)
Socket P LGA-3647 (3)
Socket P+ LGA-4189 (2)
See more
See less
Processor Supported
7002-Series (4)
7003-Series (3)
Bronze 31XX (3)
Bronze 32XX (3)
Gold 51XX (3)
Gold 52XX (3)
Gold 61XX (3)
Gold 62XX (3)
Platinum 81XX (3)
Platinum 82XX (3)
Silver 41XX (3)
Silver 42XX (3)
7001-Series (2)
Gold 53XX (2)
Gold 63XX (2)
Platinum 83XX (2)
Silver 43XX (2)
See more
See less
Maximum Memory
2 TB (3)
1 TB (2)
4 TB (2)
3 TB (1)
768 GB (1)
8 TB (1)
See more
See less
Number of Double-Wide Card Slots
4 (5)
10 (3)
5 (1)
8 (1)
See more
See less
Storage Drive Form Factor
2.5" (8)
3.5" (6)
M.2 (6)
U.2 (5)
See more
See less
Number of Drives Supported
2 (1)
4 (1)
8 (3)
9 (1)
22 (1)
24 (3)
See more
See less
Network Technology
1000BASE-T (9)
100BASE-TX (9)
10BASE-T (9)
10GBASE-T (4)
See more
See less
Number of Network (RJ-45) Ports
4 (1)
2 (8)
See more
See less
Management Port
Yes (10)
See more
See less
Maximum Power Supply Wattage
1000 W (2)
1200 W (1)
1600 W (4)
1800 W (2)
1968 W (1)
1980 W (2)
2000 W (3)
2200 W (3)
See more
See less
Green Compliance Certificate/Authority
80 Plus Platinum (6)
80 Plus Titanium (3)
See more
See less
Done