It's going well! As Norm said in the talk, Ironwood features AI-based ALU circuits and floor plan optimization. Here are some slides on the topic from a talk I gave at the NSF AI4EDA workshop at NeurIPS in Dec last year: https://
drive.google.com/file/d/1Gq8baK
6aUBwNSNXG5U7V8zJligr3gvHU/view?usp=drivesdk
… Last slide points to many papers.
AI HARDWARE
-
Ironwood AI-Based ALU Circuits and Floor Plan Optimization
By
–
-
Google Unveils TPUv7 Ironwood: 9216 Chips Reaching Zettaflop Scale
By
–
My Google colleagues Norm Jouppi & Sridhar Lakshmanamurthy gave a talk today at Hot Chips on TPUv7 ("Ironwood"). The TPUv7 system offers 9216 chips / pod (42.5 exaflops of fp8), but we can scale across many of these pods to provide multiple zettaflops.
-
Trillion Labs Cuts LLM Training Time Using NVIDIA NeMo Curator
By
–
Trillion Labs used NVIDIA NeMo Curator to cut data prep from days to hours to train highly accurate LLMs in Korea. Explore the key results in advancing sovereign AI development: Reduced training time Improved language accuracy Increased cost and energy savings
-
GPU Cooling Modification: Installing Hardline Waterblock on Graphics Card
By
–
might remove the shroud and put a hardline waterblock like I did with my 3080 Ti. should be short enough without it
-
Computer-2 Case: Plastic Quality Assessment
By
–
computer-2 case, very nice. how's the plastic quality?
-
Industry Giants Adopt NVIDIA RTX PRO Servers for AI Workloads
By
–
Ready for the age of AI? Industry giants like @Disney
, @HonHai_Foxconn
, @Hitachi
, @HMGnewsroom
, @EliLillyandCo
, @SAP
, TSMC, and more are among the first to upgrade with NVIDIA RTX PRO Servers—supercharging AI, physical AI, and business workloads. Built on the new RTX PRO 6000 -
Optimizing Python Network Performance on Single Board Computers
By
–
Struggling with slow #Python network performance on SBCs? An Axelera community member recently tackled the issue with effective diagnostic strategies. Curious? See the full breakdown right here: https://
eu1.hubs.ly/H0mnW530 -
American Money Migrates From Offices to AI Data Centers
By
–
In the US, money is leaving offices for data centers. In 2025, $40 billion invested in AI infrastructure versus $46 billion in office real estate.
-
Learning Short Range Vision Applied Long Range Lidar
By
–
You learn to see at short range and apply it at long range. Stanley did that with lidar: https://
robots.stanford.edu/papers/thrun.s
tanley05.pdf
… -
Intel’s AI Struggles Amid Emerging Chip Makers Competition
By
–
Wall Street’s Best Ideas: Intel may be doomed Intel has struggled for a long time to break into AI. Government taking a stake won’t help. Also, newly public AI chip maker Ambiq gets lackluster reviews, and Figma is still too expensive, underwriters say. https://
thetechnologyletter.com/the-posts/wall
-streets-best-ideas-intel-may-be-doomed
…