Custom Silicon
5 pieces on Custom Silicon.
News & Analysis
Architectural Specificity Now Outperforms GPU Scaling Alone
Three independent data points from one week show that workload-specific architecture delivers gains raw GPU procurement cannot match at any price.
Read more →
MTIA 300 Cuts Recommendation Model Comms Time 3.9x vs GPU
Meta's MTIA 300 puts 12 × 800 Gbps RDMA NICs on-package, hitting 940 GB/s intra-rack bandwidth and 3.9× lower communication time on a 150B-parameter model.
Read more →
OpenAI's Jalapeño ASIC Beats Nvidia GB200/GB300 on Latency and Throughput
OpenAI's Jalapeño chip delivers 1.5–1.9× more AI work per watt and 1.7–3.6× lower latency than Nvidia GB200/GB300 across three models.
Read more →
Jalapeño Beats GB200/GB300 by 1.9× Efficiency, 3.6× Latency
OpenAI's custom Jalapeño chip outperforms NVIDIA GB200 and GB300 on InferenceX across three open-weight models up to 1T parameters.
Read more →
OpenAI's Jalapeño Beats Blackwell on Inference Efficiency at Hot Chips
OpenAI's Jalapeño chip outperforms Nvidia Blackwell on tokens-per-user and throughput-per-kilowatt in SemiAnalysis' InferenceX benchmark.
Read more →