/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Toronto-based chip startup Taalas, which hardwires AI models into custom silicon to achieve faster inference, raised $169M, bringing its total funding to $219M

Toronto-based chip startup Taalas said on Thursday it had raised $169 million and has developed a chip capable …

Reuters Max A. Cherney

Context & Ripple Effects

Taalas first emerged publicly with $50 million across two rounds and a plan to unveil an LLM chip; this $169 million round takes its disclosed funding to $219 million and gives that specialized inference approach substantially more backing. Its earlier stealth exit and initial funding framed the company as a focused AI-chip entrant rather than a general-purpose processor vendor.

The company was launched by Tenstorrent founder Ljubisa Bajic, linking its progress to a Toronto AI-chip cluster that has also attracted major financing for Tenstorrent. The new round matters because it funds a distinct hardware strategy: embedding models directly into custom silicon to prioritize inference speed.

First-order effects

  • Taalas gains $169 million to advance and commercialize chips built around hardwired AI models, while existing backers see the company funded at a much larger scale than at its launch.
  • The round raises Taalas's visibility among inference-chip buyers and partners, placing its model-specific design approach in more direct contention with broader AI-chip architectures.

Second-order effects

  • Inference-chip rivals will face sharper pressure to demonstrate where flexible processors versus model-specialized silicon offer the better deployment trade-off; d-Matrix had already raised funding for chips optimized specifically for inference.
  • Tenstorrent gains an adjacent point of validation for Toronto's AI-chip ecosystem, though Taalas's founder connection also makes its progress a closer competitive benchmark for the larger chip designer, which later raised a $700 million Series D.

Third-order effects

  • If customers adopt hardwired-model designs, AI hardware could segment further between adaptable general-purpose accelerators and purpose-built inference systems, with software-model compatibility becoming as important as raw chip performance.
  • That segmentation would make capital, chip-design talent, and routes to deployment more consequential for smaller vendors; the funding signals investor willingness to finance differentiated alternatives, not proof that the approach will achieve broad adoption.

The trend: This is part of an AI hardware strategy split in which startups seek inference advantages through increasingly specialized silicon rather than a one-size-fits-all accelerator approach.

Discussion

  • @kvamme Alex on x
    @sallywf @taalas_inc And in exchange for those 30 (incremental) tape-outs (that takes only 8 weeks), you're running your frontier model 60x faster and 2x+ cheaper. And that's just the 1st generation! The ROI becomes positive very quickly when running inference at scale.
  • @davidbennett__ David Bennett on x
    This is an exciting effort. @cerebras has proven that speed matters. Ljubisa and the team have taken the same approach: develop new engineering to make a model go hyper fast. From a manufacturing perspective, @taalas_inc seems like it's tradeoffs are more digestible & scalable
  • @kvamme Alex on x
    Hard to describe until you try it. I would tag the Taalas team but they are so locked in they don't have X accounts. [video]
  • @kvamme Alex on x
    Breaking the latency barrier
  • @kvamme Alex on x
    We've been waiting 2 years to share this with the world. It's a monumental achievement.
  • @michaelxbloch Michael Bloch on x
    The biggest bottleneck in AI just got obliterated by @taalas_inc. 17,000 tokens/sec. 50x faster than Nvidia's best GPU, at a fraction of the cost and power. Cheap, instant, intelligent AI for everyone is no longer theoretical. This is Taalas' generation one. Their opening act.
  • @sallywf Sally Ward-Foxton on x
    AI chip startup Taalas @taalas_inc is showing off a chip that can do 16,000 tps/user on Llama3.1-8B, many multiples of its nearest competitor. The catch? The chip ONLY runs Llama3.1-8B, and a model like DeepSeekR1-671B would need 30 separate tapeouts: https://www.eetimes.com/...
  • @taalas_inc @taalas_inc on x
    24 dedicated people. $30M spent on development. Extreme specialization, speed, and power efficiency. Today we launch Taalas' first product. Check it out: Details: https://taalas.com/... Demo chatbot: https://chatjimmy.ai/ API: https://taalas.com/...
  • @tbpn @tbpn on x
    Happy Thursday. On today's show: - @zoink (Figma) - @esaagar (Breaking Points) - @0xSigil (Web 4.0) - @PeterMoralesX (Code Metal) - Erik Palitsch (Freeform) - Ljubisa Bajic (Taalas) See you on the stream.
  • @quietcapital @quietcapital on x
    Today, @taalas_inc is unveiling breakthrough inference chips to make AI cheap, fast, and ubiquitous. Read more from our partner, @kvamme. https://quiet.com/...