Researchers submitted a paper on 24 Sep 2026 presenting Loom, a tuning-free symbolic compiler framework for tile-based SPMD programs on spatial dataflow architectures. Loom formulates one CP-SAT probl…
A developer working on tt-metal, Tenstorrent's ML framework, identified a subtle numerical bug in the SFPU math unit that causes intermediate overflow in floating-point computation chains such as exp(…
Tenstorrent's Blackhole and Wormhole AI accelerators return infinity or NaN for large negative inputs to the softplus activation function due to an unclamped range-reduction step in the SFPU (Scalar F…
A September 9, 2026 analysis of twelve AI accelerators from NVIDIA, AMD, Google, AWS, Cerebras, and Tenstorrent finds that memory per dollar, not FLOPS, is the primary purchasing criterion for AI comp…
Nvidia announced at Hot Chips 2026 that CUDA will officially support RISC-V CPUs, making RISC-V the third supported host architecture alongside x86 and ARM, with strict requirements including the RVA2…
Veso AI open-sourced tenstorrent-wormhole-kimi-linear, kernels and correctness gates that run Kimi-Linear-48B on Tenstorrent Wormhole hardware, taking the model from not starting to serving 16,384 tok…
The High Bandwidth Flash consortium, led by Sandisk and SK Hynix with members including Google and Tenstorrent, released its first technical specification for a new NAND-based memory standard targetin…
SanDisk and SK hynix introduced High Bandwidth Flash (HBF), a new memory standard combining NAND flash with wafer bonding to deliver HBM-level read bandwidth at 8 to 16 times the capacity, targeting A…
Tenstorrent, an AI chip company, is hiring a Staff CPU Microarchitect for Vector and Matrix Compute in the United States, offering a salary range of $100k–$500k per year. The role involves driving the…
AMD has acquired Canadian AI chip startup Taalas to integrate its technology into AMD Instinct GPUs for more efficient inference workloads. Financial terms were not disclosed. Taalas, founded in Augus…
AMD has acquired Toronto-based chipmaker Taalas, the latest in a series of Canadian semiconductor startups snapped up by U.S. firms, prompting calls for a national strategy to retain homegrown talent.…
AMD agreed to acquire Taalas, a 25-person Toronto-based chip startup that builds model-specific silicon, for an undisclosed amount. Taalas's first chip, the HC1, runs only Llama 3.1 8B, with weights h…
Advanced Micro Devices (AMD) announced a definitive agreement to acquire Toronto-based Taalas, a startup that hardwires AI models onto custom chips, with financial terms undisclosed. Taalas, founded i…
AMD has acquired Taalas, the AI inference chip startup co-founded by former Tenstorrent CEO Ljubisa Bajic, who will rejoin AMD under Vamsi Boppana's AI organization. Taalas's first chip achieves over …
Sandisk and SK hynix released the first High Bandwidth Flash (HBF) technical specification through the Open Compute Project, defining NAND-based memory stacks up to 512GB with performance up to 3.0TB/…
Sandisk and SK hynix released the first High Bandwidth Flash (HBF) technical specification through the Open Compute Project on August 3-4, defining an open framework for NAND-based near-compute memory…
Sandisk and SK hynix have published the first open technical specification for High Bandwidth Flash (HBF) through the Open Compute Project, creating a new memory tier between High Bandwidth Memory and…
SK Hynix and Sandisk unveiled the first standard specifications for high bandwidth flash (HBF), a next-generation memory technology designed to address supply chain constraints of high bandwidth memor…
SK hynix and Sandisk released the first industry standard for high bandwidth flash (HBF), defining products with capacities up to 512 gigabytes and bandwidth from 0.4 to 3 terabytes per second, aiming…
Sandisk Corporation and SK hynix Inc. released the first High Bandwidth Flash (HBF) technical specification through the Open Compute Project (OCP), six months after the consortium began work in Februa…