PowerCore Technology

ics sr7 ultra box

Adaptive Multi‑Core Architecture for Professional and Workstation Workloads

PowerCore is ICS’s next‑generation high‑efficiency compute engine designed for professional creators, engineers, AI developers, and workstation‑class users. Unlike HyperCore — which focuses on extreme single‑core performance for gaming — PowerCore is built to maximize throughput, scalability, and intelligent multi‑thread distribution across large core counts.


PowerCore debuts in the upcoming ICS Regent SR‑47 Ultra series, and is engineered to deliver unmatched performance in rendering, simulation, AI training, video production, and scientific workloads.

What Makes PowerCore Different

HyperCore compresses cores. PowerCore expands them.

Where HyperCore fuses multiple physical cores into fewer, ultra‑fast HyperCores, PowerCore does the opposite:


• A PowerCore can logically split itself into multiple execution units


• It can handle two medium‑weight threads more efficiently than a single traditional core


• It dynamically adapts to workload type, thread pressure, and memory behavior



This makes PowerCore ideal for:


• multi‑threaded rendering


• AI inference and training


• video encoding and color grading


• large‑scale compilation


• scientific and engineering simulations

PowerCore Split Execution

PowerCore introduces a new execution model called Split Execution, allowing each PowerCore to:

• divide itself into two “micro‑cores” for parallel medium‑intensity tasks


• merge back into a single high‑power core for heavy single‑threaded bursts


• dynamically switch modes in microseconds



This ensures:


• higher utilization


• less idle time


• smoother scaling across 32, 48, 64, or even 96 cores

PowerCore AI Scheduler

PowerCore uses the same AI foundation as HyperCore — but with a completely different goal.

PowerCore AI Scheduler:


• analyzes thread behavior, memory access patterns, and workload type


• predicts when to split or merge PowerCores


• prioritizes tasks based on throughput, not latency


• balances thermal and power budgets across large core clusters


• ensures maximum sustained performance under heavy loads


This allows PowerCore CPUs to maintain high performance even during long, demanding sessions.

PowerCore Compute Islands

To support extreme scalability, PowerCore introduces Compute Islands:

• clusters of 8–12 PowerCores


• each island has its own L3 cache region and memory controller


• islands communicate through a high‑bandwidth RA5 interconnect


• designed for CPUs with up to 96C / 192T


Compute Islands allow PowerCore CPUs to scale far beyond traditional monolithic designs.

PowerCore Performance Highlights

On the flagship ICS Regent GX 9‑9995RX4D, HyperCore delivers:

On the upcoming ICS Regent SR‑47 Ultra, PowerCore delivers:


Up to 96C / 192T


Adaptive Split Execution for dual‑thread efficiency


PowerCore AI Scheduler for intelligent workload distribution


Compute Island Architecture for massive scalability


Superior performance in rendering, AI, and scientific workloads


Significantly higher sustained throughput than competing workstation CPUs


PowerCore is built for one thing:

maximum multi‑core efficiency and professional‑grade performance.