9 months ago
Supermicro Expands AI Portfolio with Air-Cooled AMD Instinct MI355X GPUs
Supermicro, a company that makes computers for big tasks like AI, has added a new server to its collection.
This server is special because it uses powerful AMD Instinct MI355X GPUs and doesn't need water to cool down, making it easier to use in regular data centers.
It's really fast at doing AI tasks and can handle a lot of data at once.
You can buy it now and see it at a big computer conference in Missouri.
Supermicro introduces a new 10U air-cooled server featuring AMD Instinct MI355X GPUs
The server offers up to 4x generation-on-generation AI compute improvement and up to 35x inferencing performance leap
The new server is part of Supermicro's Data Center Building Block® (DCBBS) architecture
It delivers up to double-digit more performance compared to the air-cooled 8U MI350X system
The server is currently shipping and showcased at SC25 in St. Louis, MO
- Who
- Super Micro Computer, Inc. (SMCI)
- What
- Introduced a new 10U air-cooled server with AMD Instinct MI355X GPUs
- Where
- Supercomputing Conference, St. Louis, MO
- When
- November 20, 2023
- Why
- To expand its portfolio of AI solutions and offer customers more choices
Key facts
- GPUs
- AMD Instinct MI355X
- Cooling
- Air-cooled
- Performance Improvement
- Up to 4x AI compute, up to 35x inferencing
- Server Type
- 10U
- Availability
- Currently shipping
Quotes
Vik Malyala, SVP Technology and AI at Supermicro
Vik Malyala is the Senior Vice President of Technology and AI at Super Micro Computer, Inc.
“Supermicro leads the industry with the most experience in delivering high-performance solutions in AI and HPC for our customers.”
theprint.in
Travis Karr, corporate vice president of business development, Data Center GPU Business, AMD
Travis Karr is the corporate vice president of business development for the Data Center GPU Business at AMD.
“AMD is proud to collaborate with Supermicro to bring the air-cooled AMD Instinct MI355X GPU to market, making it easier for customers to deploy advanced AI performance within existing infrastructure.”
theprint.in
