Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
TensorwaveA leader in AI infrastructure solutions supplied with AMD, announced today the deployment of GPU MI355X AMD Instinct MI355X in its high performance cloud platform.
As one of the first cloud suppliers to put on the MI355X of the AMD Instinct, Tensorwave allows customers to unlock the following level performance for the most demanding IA workloads – all with the integration and unmatched white glove support.
The new MI355X AMD Instinct GPU is built on 4th generation DNC architecture and has 288 GB of HBM3E memory and 8 TB / S memory bandwidth, optimized for generative AI training, inference and high performance calculation (HPC).
The early adoption of Tensorwave allows its customers to benefit from the compact and compact scalable design of the MI355X and the advanced architecture, offering a high density calculation with a large -scale advanced cooling infrastructure.
“Tensorwave’s deep specialization in AMD technology makes us a highly optimized environment for the workloads of the new generation of the AI,” said Piotr Tomasik, president of Tensorwave, in a press release. “With the instinct MI325X now deployed on our cloud and our instinct MI355X soon, we allow startups and companies to reach up to 25% of efficiency gains and 40% cost reductions, the results that we have already seen with customers using our infrastructure fueled by AMD.”
The exclusive use of AMD GPUs by Tensorwave provides customers with an open and optimized AI software battery fed by AMD ROCM, avoiding the locking of suppliers and by reducing the total cost of possession. Its emphasis on scalability, developer-integration and business quality ALS make it the essential partner for hierarchizing performance and choice organizations.
“GPUs of the AMD Instinct MI350 series offer revolutionary performance for the most demanding workloads of AI and HPC,” said Travis Karr, Vice-President of Business Development, Data Center GPU Business, AMD, in a press release. “The AMD instinct portfolio, as well as our ROCM Open Software ecosystem, allows customers to develop cutting-edge platforms that feed generator, the scientific discovery led by AI and high-performance computer applications.”
Tensorwave also currently builds the largest IA training group specific to AMD in North America, making its mission progress to democratize access to high performance calculation. By providing end -to -end support for AI workloads based on AMD, Tensorwave allows customers to transmit, optimize and fail transparently in an open and rapidly evolving ecosystem.
For more information, please visit:
