Tachyum Prodigy: New CPU works with x86, ARM and RISC-V software

For the launch of the "universal processors" Prodigy next year, Tachyum has now reached a milestone: the compatibility with large software ecosystems, especially x86 and ARM, but also the rising RISC-V. This legacy support should complete the start with native new software for the Cpus.

Prodigy code, it also runs legacy x86, ARM and RISC-V binaries
Tachyum Shows Prodigy Running Existing x86 ARM

But still the efforts are all just simulations and emulations based on smaller samples with a similar setup, because the final processor has not yet had its tape-out behind it. However, Tachyum already explains that the final product is faster than anything ARM and RISC-V have on the market today.

Despite software emulation on the Prodigy chip, the ARM and RISC-V binaries will run much faster on Tachyum Prodigy than on ARM or RISC-V available today. This is a testament to the raw brute force performance of the Prodigy processor.

This is Tachyum Prodigy


Tachyum Prodigy is the start-up’s first project. Behind this is a small processor with up to 64 cores, which has a DDR4/DDR5 memory interface with eight channels, offers up to 72 Pcie 5.0 lanes and also natively provides up to Dual 400 Gigabit Ethernet. Various smaller variants should be able to hold water for almost every current Intel product on the market, because each of the CPU cores should be faster than that of a current Xeon, but even smaller than an ARM core, claims Tachyum based on simulated Specint and Specfp 2006 benchmarks.

Tachyum Prodigy


Tachyum again speaks of the smallest and fastest general-purpose, 64-bit processor, requiring 10x lower power and 3x lower sell price than competing products with equivalent performance. It is still unclear whether this extremely ambitious project will ultimately prove to be true. The launch originally planned for this year has been postponed to 2021, as a fully functional FPGA prototype will be rolled off the assembly line at TSMC only later this year, which is the last major milestone before the tape-out of the product.

Prodigy outperforms the fastest Xeon processors at 10x lower power on data center workloads, as well as outperforming NVIDIA’s fastest GPU on HPC, AI training and inference. The 125 HPC Prodigy racks can deliver a 32 tensor EXAFLOPS. Prodigy’s 3X lower cost per MIPS and 10X lower power translates to a 4X lower data center Total Cost of Ownership (TCO)

Next Post Previous Post