home.social

#zettascale — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #zettascale, aggregated by home.social.

  1. Chatting with a pal reminded me of a fun pre-SC activity: looking back at old conference takes that aged like milk. Remember this one?

    #HPC #zettascale #hedoesntworkthereanymore

  2. Chatting with a pal reminded me of a fun pre-SC activity: looking back at old conference takes that aged like milk. Remember this one?

    #HPC #zettascale #hedoesntworkthereanymore

  3. Chatting with a pal reminded me of a fun pre-SC activity: looking back at old conference takes that aged like milk. Remember this one?

    #HPC #zettascale #hedoesntworkthereanymore

  4. Chatting with a pal reminded me of a fun pre-SC activity: looking back at old conference takes that aged like milk. Remember this one?

    #HPC #zettascale #hedoesntworkthereanymore

  5. Chatting with a pal reminded me of a fun pre-SC activity: looking back at old conference takes that aged like milk. Remember this one?

    #HPC #zettascale #hedoesntworkthereanymore

  6. On today's episode of the #ITOps Query #podcast, I sit down with Oracle senior vice president Sudha Raghavan, who heads up #AIinfrastructure for OCI. This means she oversees all expansions and new buildouts for some of the world's largest infrastructure platforms, which support #GPU clusters, including network design.

    As the company prepares to build new #gigawatt #datacenters worldwide, she explains how its #database heritage helps in operating increasingly demanding networks with high performance and efficiency. We also discuss the company's recent updates at #AIWorld, including a new #Acceleron multi-planar network, converged #NIC and #zettascale clusters.

    youtube.com/watch?v=2j-awG9IYJ

  7. On today's episode of the #ITOps Query #podcast, I sit down with Oracle senior vice president Sudha Raghavan, who heads up #AIinfrastructure for OCI. This means she oversees all expansions and new buildouts for some of the world's largest infrastructure platforms, which support #GPU clusters, including network design.

    As the company prepares to build new #gigawatt #datacenters worldwide, she explains how its #database heritage helps in operating increasingly demanding networks with high performance and efficiency. We also discuss the company's recent updates at #AIWorld, including a new #Acceleron multi-planar network, converged #NIC and #zettascale clusters.

    youtube.com/watch?v=2j-awG9IYJ

  8. On today's episode of the Query , I sit down with Oracle senior vice president Sudha Raghavan, who heads up for OCI. This means she oversees all expansions and new buildouts for some of the world's largest infrastructure platforms, which support clusters, including network design.

    As the company prepares to build new worldwide, she explains how its heritage helps in operating increasingly demanding networks with high performance and efficiency. We also discuss the company's recent updates at , including a new multi-planar network, converged and clusters.

    youtube.com/watch?v=2j-awG9IYJ

  9. On today's episode of the #ITOps Query #podcast, I sit down with Oracle senior vice president Sudha Raghavan, who heads up #AIinfrastructure for OCI. This means she oversees all expansions and new buildouts for some of the world's largest infrastructure platforms, which support #GPU clusters, including network design.

    As the company prepares to build new #gigawatt #datacenters worldwide, she explains how its #database heritage helps in operating increasingly demanding networks with high performance and efficiency. We also discuss the company's recent updates at #AIWorld, including a new #Acceleron multi-planar network, converged #NIC and #zettascale clusters.

    youtube.com/watch?v=2j-awG9IYJ

  10. On today's episode of the #ITOps Query #podcast, I sit down with Oracle senior vice president Sudha Raghavan, who heads up #AIinfrastructure for OCI. This means she oversees all expansions and new buildouts for some of the world's largest infrastructure platforms, which support #GPU clusters, including network design.

    As the company prepares to build new #gigawatt #datacenters worldwide, she explains how its #database heritage helps in operating increasingly demanding networks with high performance and efficiency. We also discuss the company's recent updates at #AIWorld, including a new #Acceleron multi-planar network, converged #NIC and #zettascale clusters.

    youtube.com/watch?v=2j-awG9IYJ

  11. RIKEN Future Computing Plan: Fugaku Towards #AI "Zettascale" FugakuNEXT (@ 40MW) Prof. Matsuoka’s presentation at DoE ASCAC FugakuNEXT will be #Zettascale in the same way as #Fugaku was #Exascale science.osti.gov/-/media/ascr... #HPC

  12. RIKEN Future Computing Plan: Fugaku Towards #AI "Zettascale" FugakuNEXT (@ 40MW) Prof. Matsuoka’s presentation at DoE ASCAC FugakuNEXT will be #Zettascale in the same way as #Fugaku was #Exascale science.osti.gov/-/media/ascr... #HPC

  13. RIKEN Future Computing Plan: Fugaku Towards #AI "Zettascale" FugakuNEXT (@ 40MW) Prof. Matsuoka’s presentation at DoE ASCAC FugakuNEXT will be #Zettascale in the same way as #Fugaku was #Exascale science.osti.gov/-/media/ascr... #HPC

  14. RIKEN Future Computing Plan: Fugaku Towards #AI "Zettascale" FugakuNEXT (@ 40MW) Prof. Matsuoka’s presentation at DoE ASCAC FugakuNEXT will be #Zettascale in the same way as #Fugaku was #Exascale science.osti.gov/-/media/ascr... #HPC

  15. RIKEN Future Computing Plan: Fugaku Towards #AI "Zettascale" FugakuNEXT (@ 40MW) Prof. Matsuoka’s presentation at DoE ASCAC FugakuNEXT will be #Zettascale in the same way as #Fugaku was #Exascale science.osti.gov/-/media/ascr... #HPC

  16. #OpenAI latest model (assuming 100 epochs) requires 3 Octillion FLOPS (3 x 10^27). That will take 6 million years on an #Intel i9 CPU. #NVIDIA has achieved in a couple of years.

    ➡️ So, #NVIDIA has truly conquered #exascale & within reach of #zettascale compute

  17. #OpenAI latest model (assuming 100 epochs) requires 3 Octillion FLOPS (3 x 10^27). That will take 6 million years on an #Intel i9 CPU. #NVIDIA has achieved in a couple of years.

    ➡️ So, #NVIDIA has truly conquered #exascale & within reach of #zettascale compute

  18. #OpenAI latest model (assuming 100 epochs) requires 3 Octillion FLOPS (3 x 10^27). That will take 6 million years on an #Intel i9 CPU. #NVIDIA has achieved in a couple of years.

    ➡️ So, #NVIDIA has truly conquered #exascale & within reach of #zettascale compute

  19. #OpenAI latest model (assuming 100 epochs) requires 3 Octillion FLOPS (3 x 10^27). That will take 6 million years on an #Intel i9 CPU. #NVIDIA has achieved in a couple of years.

    ➡️ So, #NVIDIA has truly conquered #exascale & within reach of #zettascale compute

  20. #OpenAI latest model (assuming 100 epochs) requires 3 Octillion FLOPS (3 x 10^27). That will take 6 million years on an #Intel i9 CPU. #NVIDIA has achieved in a couple of years.

    ➡️ So, #NVIDIA has truly conquered #exascale & within reach of #zettascale compute