๐ Medial Secures Investment on Shark Tank India - Fueling the Future of Professional Social Networking. ๐ฅ
โ
Login
Home
News
Messages
Startup Showcase
Trackers
Premium
Premium Content
Jobs
Notifications
Settings
Try our Valuation Calculator โ
Log In
News on Medial
Matrix multiplication breakthrough could lead to faster, more efficient AI models
Arstechnica
ยท
1y ago
Medial
Computer scientists have made a significant breakthrough in matrix multiplication efficiency, which could have a major impact on AI models. Researchers from Tsinghua University, the University of California, Berkeley, and MIT have discovered a new algorithm that reduces the complexity exponent of matrix multiplication, bringing it closer to the theoretical minimum number of operations needed. This improvement could lead to faster training and execution times for AI models, making them more efficient and capable. While further progress and understanding of the problem are needed, this development represents the biggest improvement in matrix multiplication in over a decade.
View Source
Related News
Researchers upend AI status quo by eliminating matrix multiplication in LLMs
Arstechnica
ยท
1y ago
Medial
Researchers have proposed a new method to run AI language models more efficiently by eliminating matrix multiplication, which is a core operation in neural networks. The study suggests that their approach could significantly reduce the environmental impact and operational costs of AI systems. The researchers developed a custom 2.7 billion parameter model without using matrix multiplication, achieving similar performance to conventional large language models. They also demonstrated running a 1.3 billion parameter model on a GPU accelerated by a custom-programmed FPGA chip, which consumed around 13 watts of power. The findings challenge the prevailing belief that matrix multiplication is essential for high-performing language models and could make large models more accessible and sustainable.
View Source
Google DeepMindโs new AI coding tool can solve complex math problems, design algorithms
Economic Times
ยท
2m ago
Medial
DeepMind's AI tool, AlphaEvolve, can solve complex math problems and design algorithms by utilizing large language models (LLMs) with automated evaluators for accuracy. It enhances Google's infrastructure, improving data center efficiency, chip design, and AI training. AlphaEvolve showed significant capability by improving solutions for mathematical problems in various fields, rediscovering or enhancing known solutions in 95% of cases. This advancement helps optimize operations such as matrix multiplication, reducing training time effectively.
View Source
Cognizant, Nvidia tap generative AI to boost drug discoveries
Economic Times
ยท
1y ago
Medial
Cognizant, a global IT services and consulting company, is utilizing generative AI (gen AI) technology in collaboration with Nvidia's BioNeMo platform to enhance drug discovery productivity and accelerate market reach in the health sciences sector. By leveraging gen AI, clinical researchers can analyze large datasets, predict drug compound interactions, and create new paths for drug development. This breakthrough technology has the potential to revolutionize the drug discovery process, making it faster and more accessible. Cognizant aims to provide clients with model-making services and pretrained models, simplifying the training and customization of enterprise models. They will establish an Nvidia AI Center of Excellence to further innovate with Nvidia technologies.
View Source
Google names insider Kavukcuoglu to lead AI-powered product development
Economic Times
ยท
1m ago
Medial
Google has appointed Koray Kavukcuoglu, chief technology officer of its DeepMind AI lab, as its new chief AI architect. In this expanded role, Kavukcuoglu will focus on integrating AI models into products for faster and more efficient development. This strategic move comes as Alphabet aims to showcase financial returns from AI investments amidst rising market pressures and competition. Kavukcuoglu will remain DeepMindโs CTO while reporting directly to CEO Sundar Pichai.
View Source
Today's GenAI models are cost and energy pigs. Quantum tech could fix that.
Business Insider
ยท
1y ago
Medial
Zapata, a startup spun out of Harvard, uses quantum physics math to train GenAI models with less data. It won over customer Andretti Racing to the point where Andretti is taking it public via a SPAC. Traditional AI models require large amounts of data and expensive computing resources, but Zapata's approach allows for more efficient and cost-effective training. While Zapata's technology is not a cheap DIY option, it offers a promising alternative to the current resource-intensive state of generative AI. Quantum tech could potentially revolutionize the field of AI by providing more efficient solutions.
View Source
How BharatGPTโs Indic-First LLM Is Bridging Language Barriers, Empowering Enterprises
Inc42
ยท
1y ago
Medial
OpenAI's ChatGPT, powered by generative AI (GenAI), is a major breakthrough in natural language understanding and processing. The intelligent chatbot utilizes large language models to generate content in various formats and languages. McKinsey & Company estimates that GenAI and other technologies could automate the majority of work activities, leading to increased productivity. However, there are concerns about GenAI falling short of community expectations, as demonstrated by the limited accuracy in generating a portrait of Lord Rama. Access to information is a crucial aspect of GenAI's future, as there is a significant gap in availability of non-English content for the majority of the Indian population. OpenAI is continually working to improve the conversational abilities of its GenAI models, with claims that the upcoming GPT-4o will be faster and more natural in its responses.
View Source
Microsoft works to add non-OpenAI models into 365 Copilot products, sources say
Economic Times
ยท
7m ago
Medial
Microsoft is working on incorporating internal and third-party AI models into its flagship AI product, Microsoft 365 Copilot, to reduce costs and lessen its reliance on OpenAI. The move aims to make 365 Copilot faster, more efficient, and cheaper for Microsoft to run. While Microsoft continues to partner with OpenAI on frontier models, they are also working on training their own smaller models and customizing other open-weight models. The goal is to provide cost savings to customers. Microsoft 365 Copilot is still working to prove its ROI to enterprises, but adoption and sales are expected to increase.
View Source
How Nvidia adapted its chips to stay ahead of an AI industry shift
Livemint
ยท
5m ago
Medial
Nvidia faced a challenge as AI shifted from model training to inference, inviting competition from companies like Advanced Micro Devices. To maintain its lead, Nvidia introduced its Blackwell AI chips, optimized for inference with more memory and efficient computations. Despite competition from startups and established firms developing specialized AI chips, Nvidia remains focused on enhancing inference capabilities, seeing inference as integral to evolving AI models. Nvidia's success is reflected in exceeding sales forecasts and optimistic future prospects.
View Source
AI could consume more power than Bitcoin by the end of 2025
The Verge
ยท
2m ago
Medial
AI could soon surpass Bitcoin mining in energy consumption, potentially using nearly half of all electricity consumed by global data centers by 2025. Despite efficiency gains, rising energy demands from expanding AI models and new data centers are contributing to this increase. The lack of transparency in reporting AI energy use and environmental impact by tech companies makes tracking and managing these demands challenging. Researchers urge for more efficient models and greater transparency in sustainability reporting.
View Source
AI chip startup d-Matrix raises $110 million in funding from Temasek, Microsoft, others
Economic Times
ยท
1y ago
Medial
Silicon Valley-based AI chip startup d-Matrix secured $110 million in funding, with Microsoft among the investors. The Series B round, led by Temasek, supports d-Matrix's chip technology, designed for efficient generative AI applications. Differentiating from Nvidia, d-Matrix focuses on AI inference processing, aiming to deliver lower latency and reduced energy consumption. Microsoft plans to evaluate the chips for its use. d-Matrix anticipates generating $70-75 million in annual revenue within two years.
View Source
Trackers
Active Indian VCโs
OG Capital
Email
With a hands-on approach, OG Capital aims to invest in over 20 promising...
Accel Partners
Email
Early and growth-stage investments in disruptive technology companies with...
Blume
Email
Early-stage venture capital firm investing in technology startups in India. Focus on...
Access All Trackers
Startup Showcase Winners
June 2025
Buddy
Helping your parents when you are miles away
BiteStop
The Pit Stop Your Cravings Deserve
Bloomer
The next generation E-commerce platform
Enter Ongoing Startup Showcase
Top Users
Trending News on Medial
Download the medial app to read full posts, comements and news.
Go to Medial App
Not Now
Know everything thatโs happening in the startup ecosystem, first.
Enable Notifications?
No, thanks
Count me in