We are planning a podcast with d-Matrix about inference accelerators. What would you like to know?

Sep 21, 2026 · 4:52 AM UTC

2
18
13,111
Sort replies: Relevant Recent Liked
Replying to @semidoped
1) Who are the likely winners of the inference accelerator market? 2) What workloads benefit the most from inference acceleration? 3) What percentage of overall workloads will likely be accelerated versus slower? By when?
165
Replying to @semidoped
How does the speed up in bandwidth compete with SRAM based solutions like CBRS or Groq? Assuming it cannabalizes some of those workloads
34