Tachyum builds the Prodigy® universal processor, a single-chip solution that combines general‑purpose, AI, and high‑performance computing workloads. The chip delivers up to 21× higher AI performance and 10× better performance‑per‑watt, cutting data‑center power use and total‑cost‑of‑ownership by up to 12× while running existing software stacks natively.
Funding
$220M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
ECFounders
Product
Problem
Data centers and high‑performance computing platforms face escalating power consumption, high total‑cost‑of‑ownership (TCO) and limited flexibility when scaling AI and HPC workloads. Existing processors often require multiple specialized chips, leading to fragmented architectures and inefficient resource utilization.
Solution
Tachyum’s Prodigy® is a universal processor that combines general‑purpose, AI and high‑performance computing capabilities in a single, compact silicon die. By delivering up to 21 × higher AI performance and 10 × higher performance per watt, Prodigy reduces data‑center power draw and cuts TCO by up to 12 ×. The chip runs existing software stacks natively, eliminating the need for code rewrites, and is offered with supporting IP such as TPU, TDIMM memory modules and FPGA‑based emulation for rapid integration. This enables operators to consolidate workloads onto one processor, achieve higher throughput with lower energy, and lower capital and operational expenses.
Target Audience
Primary customers are hyperscale data‑center operators and private‑cloud providers that run AI, machine‑learning and high‑performance computing workloads, as well as telecom and edge‑computing operators seeking high‑density, low‑power compute solutions.
Features
- Integrated universal ISA that executes general‑purpose, AI and HPC workloads without software modification
- 21 × AI training/inference performance and 10 × performance‑per‑watt improvement over current rack solutions
- Up to 12 × reduction in data‑center total‑cost‑of‑ownership
- Smallest, fastest and most energy‑efficient general‑purpose chip on the market
- Native support for existing software ecosystems and operating systems
- Complementary IP blocks: TPU accelerator, high‑density TDIMM memory, and Prodigy FPGA emulator for development and validation
- Scalable reference designs and software distribution packages for rapid deployment in hyperscale, private‑cloud and edge environments