LAS VEGAS--(BUSINESS WIRE)--With volume production of Prodigy®, the world’s first Universal Processor, scheduled for 2024, Tachyum® officials expect to build off the momentum it gained through a series of successes last year while capitalizing on the current semiconductor upcycle to revolutionize the data center.
Anticipated company highlights of the coming year include:
- Advancements in Product Development – Reference design is expected to soon be revealed once the Tachyum team finishes final debugging and cleaning of the chip.
- Increased Business Development – With the final reference chip available, Tachyum will target early adopter markets, including high performance computing and artificial intelligence applications.
Tachyum diligently worked throughout 2023 to ensure that it is positioned at the precipice of success for the coming year. Achievements in hardware and software, advancements in IP component technology and a major purchase order to build a large-scale system were key contributing factors that are expected to benefit the company’s upcoming year.
In 2023, Tachyum received a major purchase order from a U.S. company to build a large-scale system based on its 5nm Prodigy Universal Processor chip. With performance of 50 EF of FP64 and 8 ZF of AI training for large language models, the announced system has gained a lot of attention from other interested parties looking to build a similar scale system for their AI applications and workloads.
The company made available its Tachyum Processing Unit (TPU®) IP as a licensable core, allowing developers to take full advantage of intelligent, datacenter-trained AI when making IoT and Edge devices.
Details for a new approach to designing HPC/AI supercomputer data centers have been released as part of a series of whitepapers in order to prepare developers for what they can expect when Prodigy is commercially available. Topics included how to use 4-bit Tachyum AI (TAI) and 2-bit effective per weight (TAI2) formats in Large Language Models (LMMs); how Prodigy overcomes hardware inefficiencies that make Generative AI cost prohibitive and energy excessive; and how Prodigy enables quantization using 8-bit floating point (FP8) to enable enormous model sizes.
To put Prodigy’s capabilities in perspective, a 1.6 trillion parameters Switch Transformer would require 52 NVIDIA H200 80GB GPUs at a cost of $41,789 each and seven Supermicro GPU servers at $25,000 each for a total cost of $2,348,028. In contrast, a single Prodigy socket system with 2TB DDR5 DRAM could fit and run such big models and bring them into the mainstream for generative AI applications for only $23,000 – 1/100th the cost of the first example.
Software achievements this past year included:
- Development completion of the software stack for Prodigy, allowing for alpha testing to begin
- Addition of LLVM for AI and Linux Rust support
- Running of Xen hypervisor on QEMU software emulator
- Running of Kubernetes natively on Prodigy ISA
Hardware achievements completed in 2023 include:
- Completion of vector based High-Performance LINPACK (HPL) testing using 1kb vectors on the Prodigy FPGA
- Running of FreeBSD OS on Prodigy FPGA
- Seamless execution of non-native (x86_64) applications under Linux on FPGA
- Integration of Baseboard Management Controller (BMC) and Unified Extensible Firmware Interface (UEFI) platforms into the Prodigy FPGA prototype system
- Porting of other applications to test a wide range of workloads on the FPGA prototype
Additionally, Tachyum Prodigy 2 was selected by the Important Project of Common European Interests (IPCEI) program representing Slovakia to deliver exa-scale HPC and zetta-scale AI for Europe. The company hosted Slovak President Zuzana Čaputová at its Bratislava offices to demonstrate the progress of the Prodigy Universal Processor.
As a Universal Processor offering utility for all workloads, Prodigy-powered data center servers can seamlessly and dynamically switch between computational domains (such as AI/ML, HPC, and cloud) on a single architecture. By eliminating the need for expensive dedicated AI hardware and dramatically increasing server utilization, Prodigy reduces CAPEX and OPEX significantly while delivering unprecedented data center performance, power, and economics. Prodigy integrates 192 high-performance custom-designed 64-bit compute cores, to deliver up to 4,5x the performance of the highest-performing x86 processors for cloud workloads, up to 3x that of the highest performing GPU for HPC, and 6x for AI applications.
“Tachyum showed its resilience during a challenging 2023 when we experienced the worst semiconductor downturn in years and weathered the challenges faced with failed memory and interface IP blocks from a key vendor,” said Dr. Radoslav Danilak, founder and CEO of Tachyum. “The successes we were able to achieve while tweaking our product roadmap have brought us to a 2024 full of anticipation as we move towards volume production of Prodigy and the fulfillment of a multibillion-dollar sales pipeline. We look forward to fulfilling our commitment to transforming ordinary data centers into Universal Computing Centers in the near future.”
Follow Tachyum
https://twitter.com/tachyum
https://www.linkedin.com/company/tachyum
https://www.facebook.com/Tachyum/
About Tachyum
Tachyum is transforming the economics of AI, HPC, public and private cloud workloads with Prodigy, the world’s first Universal Processor. Prodigy unifies the functionality of a CPU, a GPU, and a TPU in a single processor to deliver industry-leading performance, cost and power efficiency for both specialty and general-purpose computing. As global data center emissions continue to contribute to a changing climate, with projections of their consuming 10 percent of the world’s electricity by 2030, the ultra-low power Prodigy is positioned to help balance the world’s appetite for computing at a lower environmental cost. Tachyum recently received a major purchase order from a US company to build a large-scale system that can deliver more than 50 exaflops performance, which will exponentially exceed the computational capabilities of the fastest inference or generative AI supercomputers available anywhere in the world today. When complete in 2025, the Prodigy-powered system will deliver a 25x multiplier vs. the world’s fastest conventional supercomputer – built just this year – and will achieve AI capabilities 25,000x larger than models for ChatGPT4. Tachyum has offices in the United States and Slovakia. For more information, visit https://www.tachyum.com/.