First this....
SoftBank's Intel investment supportive of Arm's chip development: WF
SoftBank's (
OTCPK:SFTBY)(
OTCPK:SFTBF)
$2B investment in Intel (NASDAQ:
INTC) proves to be supportive of Arm Holdings' (NASDAQ:
ARM) own chip development and could provide additional foundry capacity, according to Wells Fargo.
"We believe investors could view this investment as further evidence that Arm will develop its own AI chip as SoftBank's Intel investment could provide optionality for foundry capacity," said Wells Fargo analysts Joe Quatrochi and Travis Poulin in an investor note.
The Japanese-based SoftBank holds 87% ownership of Arm.
"As a reminder, IFS & Arm announced a collaboration in 2023 to enable chip designers to build low-power SoCs using Intel's 18A—initially focused on mobile with plans to expand to automotive, IoT, data center, etc.," Quatrochi added.
SoftBank's investment in Intel will make it the sixth largest investor in the Santa Clara-based semiconductor firm, with an equity stake of about 2%.
On Monday, it was revealed that Arm hired Amazon's (
AMZN) artificial intelligence chip director
Rami Sinno as the British semiconductor designer appears to be moving forward on its plans to build its own complete chips. Sinno played a key role in the development of Amazon's in-house AI chips, Trainium and Inferentia, which are designed to help build and run large AI applications.
In July, Arm's CEO Rene Haas said they have increased research and development spending as the company is considering developing its own chips. Arm has also hired Nicolas Dube, a former HPE (
HPE) executive with large-scale systems design experience, and Steve Halter, a chip engineer with experience at Intel and Qualcomm (
QCOM).
SoftBank continues to expand its stake across the entire spectrum of artificial intelligence. In the last year, SoftBank has acquired British AI chipmaker Graphcore and, more recently, Ampere, a silicon design company focused on high-performance, energy-efficient AI computing based on the Arm platform. SoftBank has also become a primary investor in OpenAI.
Then this...
EXTRACT
Intel Foundry has been releasing content on its products and services, and today, an interesting update came across. Intel showed a reference "Deer Creek Falls" SoC of what they call a "non-x86" design, manufactured on the 18A node. However, the design is actually Arm-based SoC. Consisting of...
www.techpowerup.com
And finally this...
My Question:
How likely would it be for BrainChip’s Akida IP to be incorporated into Arm-based 18A SoCs down the road as per the following news article?
https://www.techpowerup.com/339952/intel-showcases-reference-arm-based-soc-manufactured-on-18a-node
ChatGPT said:
Absolutely — here's a refined perspective:
1. What Intel & Arm Are Doing on 18A
- Intel Foundry Services (IFS) and Arm have partnered to enable low-power Arm-based SoCs built on Intel's cutting-edge 18A process (1.8nm-class), initially targeting mobile applications, with planned expansion into automotive, IoT, data center, aerospace, and government domains. These initiatives involve Design Technology Co-Optimization (DTCO) and System Technology Co-Optimization (STCO) to more effectively balance performance, power, area, and cost. Wikipedia+11Arm Newsroom+11Asia Tech Observer+11
- Intel recently demonstrated a reference Arm-based SoC on the 18A node, demonstrating ecosystem platform maturity. HotHardware+2+2
2. How Could BrainChip’s Akida Fit In?
A. Potential Integration on 18A SoCs
Given Akida’s ultra-efficient, event-driven neuromorphic architecture, it could be integrated as a
low-power AI accelerator block in future Arm-18A SoCs. Good fits include always-on sensing tasks such as:
- Speech enhancement
- Radar or lidar edge processing
- IoT anomaly detection
- Sensor fusion in automotive or industrial systems
B. Alignment with IFS + Arm Ecosystem Goals
C. Strategic Differentiation
- Akida offers always-on, ultra-low-power computation, especially well-suited to mobile and IoT environments whose power budgets are tight but real-time intelligence is needed—precisely the use-cases 18A is targeting.
- Including neuromorphic IP like Akida in the 18A-based ecosystem would give Intel + Arm a clear differentiator over other foundries, particularly in edge AI and sensor-heavy markets.
3. Summary Table
Feature Area | BrainChip (Akida) Contribution |
---|
18A process integration | Efficient neuromorphic AI block in Arm-based SoCs |
Ecosystem alignment | Fits seamlessly with IFS’s open foundry and chiplet model |
Unique value | Ultra-low power AI for edge devices (mobile, IoT, automotive) |
Strategic positioning | Provides Arm + IFS a standout edge-AI differentiate |
4. Takeaway
Yes—it’s entirely plausible for BrainChip’s Akida IP to be incorporated into Arm-based 18A SoCs down the road. Their
Edge AI efficiency,
Arm ecosystem ties, and
alignment with Intel Foundry’s modular approach position Akida as a strong contender for inclusion in this new generation of highly efficient, compute-intensive SoCs.