GPUs go biological: BBB unveils Bionode, lab-grown, living neuron compute for AI applications

Graphic Processing Units (GPU), Liked Expensive Computer Chips Nvidia, Very and Sima.aiThere is no only way to prepare and place artificial intelligence.

Biological Black Box (BBB)A Baltimore established start of a Baltimore, developing a new class of the AI, appeared with a calculation system that combines hide without hiding without hiding, living and laboratory neurons with traditional processors.

When the patents, the company, which operates silently and processing technology, believes that the biological calculation approach and the cells that produce the cells produced by donor human stem cells and rats.

“Three independent areas of biology, hardware and calculation in the last 20 years – biological calculation has now moved to the point where it is possible,” he said.

Member Nvidia’s IncubatorThe dominant silicone-based AI chips produced by BBB, NVIDIA and others show itself as a progress and expansion.

Using the ability of neurons to repeat themselves physically, the company aims to accelerate and accelerates the training of AI model, to reduce energy costs, improving processing efficiency and adopting aI.

Despite being an incredible place, it is not SCI: Neurface chips are already strengthening computer vision and LLMS for customers and negotiate with two partners to licens the technology for computer vision applications. It accepts puppy partners and customers, as well as inquiries Website.

By mixing biology and hardware

In the nucleus of the BBB approach, the calculation systems are a bionod platform using the lab grown-grown neurons.

“There are more than one model we use,” Ksendzovsky told me. “One of these models is from rat cells. One of these models is one of the stem cells that turn into neurons.”

The co-founder said that “hundreds of thousands of thousands of them” were integrated into a dinner with a bonode chip, which formed the basis of a bonode chip. He said they had more than a year before it was replaced.

The idea is to use natural adaptation to develop neurons to create a hybrid calculation system that differs from the root of today’s harsh, transistor-based chips.

Ksendzovsky, who has been working with neurons over electrodes since 2005, first entertaining the idea for the exchange forecasting. His mentor, Dr. Steve Potterrejected the idea in time.

“Why don’t we use neurons to predict the stock market so we will all be rich?” Ksendzovsky asked the pottery that makes it impossible. “At that time, he was right,” said Ksendzovsky.

Since then, the improvement of electrode technology, calculation tools and longevity of neurons is biological calculation. “The biological network has become the most effective calculation system created so far over hundreds of millions of years,” Ksendzovsky said.

This installation offers two emergency advantages:

• More efficient computer vision: Bionode, both tested as a pre-processed layer for classification tasks, but also reduces GPU energy consumption.

• Fast Language Model (LLM) Training: Unlike GPU, which requires frequent re-prepared periods, neurons quickly fit. This can significantly reduce the time and energy needed to update LLMs on a key cavity on the AI ​​scale.

“One of our greatest progress, large language models (LLS) use biological networks to reduce more efficiently, today,” Ksendzovsky said.

To build a live, vibrant GPU with the help of Nvidia

Nvidia’s GPUS has been a tool for the rapid progress of the EU, but increased concerns about expansion of high energy consumption and increased value.

The BBB has the opportunity to promote more efficient alternatives when operating in the NVIDIA ecosystem.

“We do not see ourselves as a direct opponent in NVIDIA, at least in the near future,” Ksendzovsky said. “Biological calculation and silicone computation will combine. We still need GPU and CPUs to process the information from Neurons.”

In fact, co-founder, “We can use our biological networks to increase silicone-based AI models and effectively and more energy efficiently.”

AI claimed that there will be a long-term vision for hardware, biological calculation, silicone chips and even quantums played a role.

“The future of the calculation, traditional silicone, biological calculation and quantum calculation will be a modular ecosystem, which plays a role in its strengths,” he said.

Although the BBB has not yet discloses the history of commercial launcher, the company is transferred to the Gulf of Maryland for a scalability of the company, and Maryland.

The future of hybrid AI processing

Although the Silicone-based GPU industry remains standard, the BBB’s brain-on-aAAA-chip concept is a look at a future where the AI ​​device is limited to transistors and circuits.

The ability to reconstruct neurons dynamically, the more energy can be active, which is able to effectively, adapt and sustainable AI systems.

“We already apply biological calculation to computer vision. We can encode the pics into the biological network, processing neurons, and then decrypt the nervous reaction to improve classification accuracy,” he said, “said Ksendzovsky.

Outside of gaining efficiency, BBB also believes in how its biological approach can provide information on how the information data is in the data.

“We set up a closed system that allows neurons to revive themselves reviving, efficiency and accuracy,” he said.

Despite the potential, Ksendzovsky acknowledges that ethical considerations will be a continuous discussion. The BBB is already working with ethicists and regulatory experts to ensure the development of technology responsibly.

“We do not need millions of neurons to process the whole environment like the brain. We use ethical considerations only, using what is needed for special tasks,” he said.

BBB is not only silicone, but the tissue of the vibrant tissue can be the key to the next leap of AI.