Tachyum just announced the Prodigy T16128 Universal Processor, a CPU that comes with 128 cores and tons of processing power.
Although you can’t install it on your desktop PC, it can be a serious competitor to Intel in the field of high-performance computing (HPC).
The specifications of Tachyum’s Prodigy CPU are truly mind-boggling. The flagship variant comes with 128 64-bit CPU cores that can reach speeds of up to 5.7 GHz. It also comes with 16 DDR5 memory controllers and 64 PCIe 5.0 lanes. As can be seen from the specs, this is a processor you might find in places where high-performance computing is required.
According to Tachyum, Prodigy is the world’s first ‘universal processor’. It can handle AI and machine learning workloads, among several other computing tasks, and we really mean the most taxing tasks imaginable – after all, it’s made monstrously powerful for a reason.
While it sounds impressive on paper, Intel and Nvidia both have their own HPC offerings that can do the job. But Tachyum seems convinced that it has made something much better. It stands out in that it is an all-in-one kind of chip that can perform all kinds of tasks, but it is much more energy efficient than its rivals and appears to have been produced at a much lower cost.
As reported by Tom’s Hardware, Tachyum promises to deliver performance up to four times better than the Intel Xeon 8470 and triple the performance of Nvidia’s H100 in HPC workloads. This kind of performance is made possible by the chip’s innovative architecture.
Most of the server hardware from Tachyum’s well-known competitors, such as Nvidia, Intel, and AMD, focuses on performing individual tasks with little overlap. Simply put, each block has its own job and data has to be transferred from one block to another for everything to function. Tachyum took a different approach with the Prodigy and it will be able to run all kinds of resource-heavy workloads within a single chip on individual cores.
The ability to multitask like this without relying on other hardware is nothing short of impressive, and given the server chip’s supposedly low price tag, it could be truly groundbreaking and rival Intel processors in a significant way. Reducing power consumption compared to Intel also means lower cost to use Tachyum, not just buy it. After all, this means that companies may be able to package more of these chips, increasing their capacities.
The top chip in the lineup could run 12 AI PetaFLOPs and 90 TeraFLOPs of HPC workloads. Tom’s Hardware compares this to Nvidia’s A100, and the results are unmistakable – the A100 maxes out at 5 A.1. PetaFLOPs.
Tachyum Prodigy is built from scratch with matrix and vector processing capabilities. As a result, it can support an impressive array of different data types such as FP64, FP32, BF16, FP8, and TF32. It also comes with 128MB of combined L2 and L3 cache. We know it has a 5nm process node, but we don’t know its source.
This beastly CPU range will also be available in a 64-core and 32-core variant, but launch is not underway yet. Production will start in 2023.