テクノロジー
Mac mini販売停止の裏側：AppleがIntel・Samsungへ打診したチップ多源化の真相
AIチップ需要の急増によりTSMCの先端プロセスが飽和し、AppleはMac miniの販売停止とMac Studioの供給遅延に直面している。この事態を受け、AppleはTSMCへの依存を解消するため、IntelとSamsung Electronicsとの交渉を進め、マルチファウンドリ戦略への転換を図っている。
Term
Sri Santhanam
Overview
最終更新: 2026年7月9日Appleに18年以上在籍するベテランエンジニア。組織再編後の「Silicon」部門を担い、異なるファウンドリの製造ルールに適合させる設計フローの構築を担当する。
Mentioned Articles
1 件Research Papers
5 件- DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines
The ML community is rapidly exploring techniques for prompting language models (LMs) and for stacking them into pipelines that solve complex tasks. Unfortunately, existing LM pipelines are typically implemented using hard-coded"prompt templates", i.e. lengthy strings discovered via trial and error. Toward a more systematic approach for developing and optimizing LM pipelines, we introduce DSPy, a programming model that abstracts LM pipelines as text transformation graphs, i.e. imperative computational graphs where LMs are invoked through declarative modules. DSPy modules are parameterized, meaning they can learn (by creating and collecting demonstrations) how to apply compositions of prompting, finetuning, augmentation, and reasoning techniques. We design a compiler that will optimize any DSPy pipeline to maximize a given metric. We conduct two case studies, showing that succinct DSPy programs can express and optimize sophisticated LM pipelines that reason about math word problems, tackle multi-hop retrieval, answer complex questions, and control agent loops. Within minutes of compiling, a few lines of DSPy allow GPT-3.5 and llama2-13b-chat to self-bootstrap pipelines that outperform standard few-shot prompting (generally by over 25% and 65%, respectively) and pipelines with expert-created demonstrations (by up to 5-46% and 16-40%, respectively). On top of that, DSPy programs compiled to open and relatively small LMs like 770M-parameter T5 and llama2-13b-chat are competitive with approaches that rely on expert-written prompt chains for proprietary GPT-3.5. DSPy is available at https://github.com/stanfordnlp/dspy
- CLINICAL PRACTICE GUIDELINES FOR PERITONEAL ACCESS
- DSPy: Compiling Declarative Language Model Calls into State-of-the-Art Pipelines
- Curative effect of arjunolic acid from Terminalia arjuna in non-alcoholic fatty liver disease models.
The prevalence of Non Alcoholic Fatty Liver Disease (NAFLD) is increasing globally. Terminalia arjuna W. & Arn. (Combretaceae) is an endemic tree found in India and Sri Lanka and used traditionally for its cardioprotective and hepatoprotective effects. Arjunolic acid (AA) is an oleanane triterpenoid found mainly in the heartwood of T. arjuna. This study was aimed to evaluate the hepatoprotective effect of AA using cellular and rodent models of NAFLD. AA was isolated from the ethyl acetate extract of the heartwood of T. arjuna. The structure of AA was confirmed by physical and spectroscopic data. Steatosis was induced in HepG2 cells using palmitate-oleate mixture and the effects of AA on triglyceride accumulation and lipotoxicity were assessed. In vivo effect of AA on NAFLD was assessed using HFD fed rats. The treatment with AA did not affect the cell viability upto 100 μM and showed GI25 value of 379.9 μM in HepG2 cells. The treatment with AA significantly lowered the ORO concentration by 35.98% and triglyceride accumulation by 66.36% at 50 μM concentration (P < 0.005) compared to the vehicle treated group. The treatment with AA also reduced the leakage of ALT and AST by 61.11 and 48.29% in a significant manner (P < 0.005). The in vivo findings clearly demonstrated that the animals treated with AA at 25 and 50 mg/kg concentrations showed a significant decrease in the levels of transaminases, phosphatase and GGT (P < 0.005). In the liver, the expression of PPARα and FXRα expressions were upregulated, while PPARγ expression was downregulated by the treatment with AA. The liver histology of the animals showed reduction in steatosis and MNC infiltration. These preliminary evidences suggested that AA might be a promising lead to treat NAFLD. Future robust scientific studies on AA will lead to tailoring it for the treatment of NAFLD.
- a 12-years in