Positron Secures $230M AI Funding for $1B Valuation

Nvidia has another rival to think about.

Positron AI has raised $230 million in an oversubscribed Series B funding round at a post-money valuation exceeding $1 billion, marking one of the largest recent financings in the AI inference hardware space.

The round was co-led by ARENA Private Wealth, Jump Trading, and Unless, with participation from Qatar Investment Authority (QIA), Arm, and Helena. The announcement was made at Web Summit Qatar, underscoring the company’s growing international profile.

The financing comes as the AI industry grapples with a new set of constraints. While compute performance has dominated AI hardware discussions for years, energy availability and memory limitations are increasingly shaping deployment decisions. Positron is positioning itself as a company built specifically to address those bottlenecks.

Energy and memory

As AI models grow in size and complexity, particularly in inference workloads, power consumption and memory capacity have emerged as critical challenges. Data centers face rising energy costs, grid constraints, and regulatory pressure, while advanced models increasingly require massive context windows that strain existing hardware architectures.

“We’re grateful for this investor enthusiasm, which itself is a reflection of what the market is demanding,” said Mitesh Agrawal, CEO of Positron AI. “Energy availability has emerged as a key bottleneck for AI deployment. And our next-generation chip will deliver 5x more tokens per watt in our core workloads versus Nvidia’s upcoming Rubin GPU.”

Agrawal emphasized that memory, not raw compute, is becoming the dominant limitation in many inference scenarios. “Memory is the other giant bottleneck in inference, and our next generation Asimov custom silicon will ship with over 2304 GB of RAM per device next year, versus just 384 GB for Rubin,” he said. “This will be a critical differentiator in workloads including video, trading, multi-trillion parameter models, and anything requiring an enormous context window.”

And yes, that is a reference to the famous SF author Isaac Asimov. On Positron’s site its vision is stated, “Isaac Asimov gave us the dream. We’re building the reality.”

From shipping systems to custom silicon

Positron has a decent foundation as it is already shipping product. Its current system, Atlas, is designed for AI inference workloads and emphasizes fast deployment, power efficiency, and predictable supply. The company highlights that Atlas is fully fabricated and manufactured in the US, an increasingly important consideration for customers concerned about supply chain resilience and geopolitical risk.

The new funding will accelerate Positron’s transition from Atlas to its next-generation custom silicon platform, Asimov. The company is targeting tape-out of the Asimov chip in late 2026, with production planned for early 2027. According to Positron, Asimov is designed to support roughly 2 terabytes of memory per accelerator and up to 8 terabytes per system in its Titan platform, expanding the feasible size of inference workloads at rack scale.

Dual role

A notable feature of the Series B is Jump Trading’s decision to co-lead the round after first deploying Positron’s Atlas systems internally. Jump, a player in high-performance and latency-sensitive computing, evaluated Atlas against existing GPU-based solutions.

Alex Davies, Chief Technology Officer of Jump Trading, said, “In our testing, Positron Atlas delivered roughly 3x lower end-to-end latency than a comparable H100-based system on the inference workloads we evaluated, in an air-cooled, production-ready footprint with a supply chain we can plan around.”

Timeline

Positron expects strong revenue growth in 2026 and says it is on track to achieve large-scale commercial traction roughly 2.5 years after its launch, a good timeline for a silicon company. The firm is currently working with customers across cloud infrastructure, advanced computing, and performance-sensitive verticals.

As competition intensifies around AI inference, Positron is betting that a memory-first, energy-efficient architecture can carve out a durable position in a market long dominated by general-purpose GPUs.

Whether that bet pays off will depend not just on silicon performance, but on how quickly the industry’s needs continue to shift toward efficiency over brute force compute.

Anthropic dropped a legal AI plugin causing some of Europe’s biggest data companies to lose billions in market value.

The post Positron Secures $230M AI Funding for $1B Valuation appeared first on eWEEK.