infrastructure
-
Scale AI valuation surge and Alexandr Wang’s expansion
Scale AI reached a $29 billion valuation following a $14.3 billion investment from Meta Platforms. The company leverages a workforce of 100,000…
-
Lightmatter’s photonic interconnect and Passage silicon photonics chiplet
Lightmatter addresses AI data center bandwidth bottlenecks using the Passage platform, which provides 114 Tbps of bidirectional bandwidth via a 3D vertical…
-
Memory Exhaustion and Communication Bottlenecks in Llama 3.1 405B
Llama 3.1 405B requires 810 GB of memory for FP16 precision, necessitating distribution across multiple SambaNova SN40L RDUs. Improper sharding in feedforward…
-
Wasted Infinity Fabric bandwidth in Mixtral 8x22B clusters
Engineers often stall Mixtral 8x22B performance by ignoring ROCm 7 kernel optimizations and incorrect NUMA modes. Proper configuration of AITER kernels and…
-
D-Wave revenue growth and the hardware gap
D-Wave reported a 42% increase in June quarter revenue to $3.1 million while its Advantage2 system with over 4,400 qubits outperformed classical…
-
CoreWeave closes $3.1 billion debt facility and $1.5 billion notes
CoreWeave secured a $3.1 billion delayed draw term loan and a $1.5 billion senior notes offering to fund AI infrastructure. The company…
-
The $50B Grok-3 rollout faces massive power and regulatory hurdles
xAI faces significant infrastructure challenges as Memphis utility MLGW warns that expanding the Colossus facility to 1 million GPUs could double power…
-
Sakana AI’s evolutionary approach to Japanese AI sovereignty
Sakana AI utilizes evolutionary model merging to build efficient AI ecosystems, achieving a $2.65 billion valuation. This strategy leverages existing open models…
-
The history of Graphcore and its IPU architecture
Nigel Toon and Simon Knowles founded Graphcore in 2016 to challenge NVIDIA with Intelligence Processing Units. Despite the Colossus MK2 GC200 chip…
-
NVIDIA Blackwell manufacturing failures and CoWoS-L supply gaps
Thermal expansion mismatches and TSMC CoWoS-L capacity shortages are causing three-month delays for NVIDIA Blackwell shipments. Massive orders from Google for 400,000…
-
HBM fragmentation and bandwidth bottlenecks in TPU v5p clusters
HBM fragmentation in TPU v5p clusters can exceed 50%, causing allocation failures during Gemini 2.5 Pro fine-tuning. Performance also depends on managing…
-
The climb from a San Francisco dinner to a global power
OpenAI evolved from a 2015 nonprofit founding into a massive commercial force following the 2022 ChatGPT launch. This growth triggered a $200B…