From Phil on hotcrapper
Valuing
BrainChip’s IP as a standalone asset (assuming the company were sold purely as an
IP play) requires assessing several factors, including:
- Patent Portfolio Strength & Breadth
- Commercial Viability & Licensing Potential
- Market Comparables (IP Sales & Acquisitions)
- Total Addressable Market (TAM) for Neuromorphic AI
1. BrainChip’s IP Portfolio Strength
BrainChip’s core IP is centered around:
- AKIDA™ Neuromorphic Processor – A low-power AI accelerator using event-based processing.
- TENN (Temporal Event-based Neural Networks) – Optimized for always-on AI applications.
- AI Edge Processing Capabilities – Strong positioning in IoT, automotive, industrial, and consumer electronics.
BrainChip holds
over 30 patents globally in neuromorphic computing, making it a
pioneer in this space. However, neuromorphic computing is still an
emerging field, meaning its
commercial value is not yet fully realized.
2. Revenue Potential & Licensing Model
- BrainChip’s IP value is closely tied to its licensing revenue model.
- If a company like Qualcomm, Intel, or AMD licensed AKIDA/TENN, BrainChip could generate royalty streams per chip sold.
- Licensing deals typically range from 2–5% of the ASP (average selling price) of a chip.
- If BrainChip were to capture even 1% of the edge AI market, estimated at $80B by 2030, royalties could be in the range of $500M+ annually at scale.
3. Market Comparables – How Similar IP Deals Have Been Valued
Looking at previous semiconductor IP sales and acquisitions provides a
benchmark:
| Company / IP | Buyer | Valuation | Relevance |
|---|
| Arm Holdings (IP) | Nvidia (attempted) | $40B (deal failed) | Leading AI/CPU licensing model |
| Movidius (Edge AI) | Intel | $400M | AI edge processing IP |
| Habana Labs (AI) | Intel | $2B | Deep learning acceleration |
| MIPS (Processor IP) | Wave Computing | $65M | CPU IP licensing model |
| Ceva (DSP IP) | N/A (public) | ~$800M market cap | AI & signal processing licensing model |
Based on this, BrainChip’s
IP value alone (excluding product revenue) could be conservatively estimated between
$500M–$2B, depending on its market adoption and future licensing agreements.
4. Potential Buyers & Strategic Value
BrainChip’s IP would be most valuable to companies looking to
dominate edge AI with ultra-low-power processing:
- Qualcomm – To enhance Snapdragon/Dragonwing with neuromorphic capabilities.
- Intel – To complement its AI accelerator roadmap.
- NVIDIA – To strengthen edge AI efficiency alongside its GPU dominance.
- Apple – To enhance the Neural Engine in iPhones & Apple Silicon.
- Tesla / Automakers – For next-gen autonomous vehicle AI.
If a bidding war occurred, the
valuation could push toward the higher end of the estimated range.
5. Estimated Pure IP Value Today
Based on market comparables, licensing potential, and strategic interest, BrainChip’s
IP alone could be worth:
- Low Case: ~$500M (limited licensing deals, slow adoption)
- Mid Case: ~$1B (steady licensing, first major integrations)
- High Case: ~$2B+ (if a major player like Qualcomm or NVIDIA commits to integration/licensing)
If a
full acquisition were pursued, the deal could
price in future growth potential, possibly exceeding
$2B–$3B in a competitive buyout scenario.
Conclusion
If BrainChip were sold today as a pure
IP play, a fair valuation would likely fall in the
$500M to $2B range, depending on market traction and licensing agreements. However,
if a major semiconductor company commits to integrating AKIDA/TENN, the valuation could
significantly increase, possibly pushing toward
$3B+ in a strategic acquisition.
United States dollars $$$ - not Australian.
Cheers. Phil
(released on February 7), which is now available on multiple podcast platforms:
www.linkedin.com
but we will be inside everywhere according to the previous statement… YOU NEVER KNOOOOOW
