![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
A (Larrabee like?) processor depending upon a non-existing (futuristic) ubercompiler to "autoparallelize your code".
Otherwise it's running on a standard CPU with an idle "parallel processing unit" !!!???
But the company can’t quite show any of that today — because Flow hasn’t built a chip and doesn’t necessarily intend to build one, its co-founders tell The Verge.
The company is a spinoff of the VTT Technical Research Centre of Finland, and it’s emerging from stealth today with roughly $4.3 million in funding. Like Arm, which builds reusable IP blocks that it sells to major chipmakers, Flow hopes to sign deals with the likes of AMD, Apple, Intel, Nvidia, and Qualcomm to take it from here.
At this point, Flow hasn’t validated its claims on genuine silicon but, rather, an FPGA board and some simulations while it works on a design for a PPU core and a compiler that it can license to other firms.
Right now, Flow is just a set of patented techniques, co-founders Martti Forsell and Timo Valtonen explain to me on a call.
https://www.theverge.com/2024/6/11/24176304/flow-computing-startup-parallel-processing-accelerator
Timo Valtonen, CEO and co-founder of Finland-based Flow Computing. and his collaborators laid out their case at the IEEE Hot Chips conference in August.
The PPU provides a speed-up in cases where the computing task is parallelizable, but a traditional CPU isn’t well equipped to take advantage of that parallelism, yet offloading to something like a GPU would be too costly.
To demonstrate the power of the PPU, Forsell and his collaborators built a proof-of-concept FPGA implementation of their design. The team says that the FPGA performed identically to their simulator, demonstrating that the PPU is functioning as expected. The team performed several comparison studies between their PPU design and existing CPUS. “Up to 100x [improvement] was reached in our preliminary performance comparisons assuming that there would be a silicon implementation of a Flow PPU running at the same speed as one of the compared commercial processors and using our microarchitecture,” Flow Computing CTO and co-founder Martti Forsell says.
Now, the team is working on a compiler for their PPU, as well as looking for partners in the CPU production space. They are hoping that a large CPU manufacturer will be interested in their product, so that they could work on a co-design. Their PPU can be implemented with any instruction set architecture, so any CPU can be potentially upgraded.
Otherwise it's running on a standard CPU with an idle "parallel processing unit" !!!???
Did startup Flow Computing just make CPUs 100x faster? Here’s the white paper and FAQs
The company claims its parallel processing ideas can double performance overnight.
But the company can’t quite show any of that today — because Flow hasn’t built a chip and doesn’t necessarily intend to build one, its co-founders tell The Verge.
The company is a spinoff of the VTT Technical Research Centre of Finland, and it’s emerging from stealth today with roughly $4.3 million in funding. Like Arm, which builds reusable IP blocks that it sells to major chipmakers, Flow hopes to sign deals with the likes of AMD, Apple, Intel, Nvidia, and Qualcomm to take it from here.
At this point, Flow hasn’t validated its claims on genuine silicon but, rather, an FPGA board and some simulations while it works on a design for a PPU core and a compiler that it can license to other firms.
Right now, Flow is just a set of patented techniques, co-founders Martti Forsell and Timo Valtonen explain to me on a call.
https://www.theverge.com/2024/6/11/24176304/flow-computing-startup-parallel-processing-accelerator
Flow Computing aims to boost central processing units with their ‘parallel processing units’
Instead of trying to speed up computation by putting 16 identical CPU cores into, say, a laptop, a manufacturer could put 4 standard CPU cores and 64 of Flow Computing’s so-called parallel processing unit (PPU) cores into the same footprint, and achieve up to 100 times better performance.Timo Valtonen, CEO and co-founder of Finland-based Flow Computing. and his collaborators laid out their case at the IEEE Hot Chips conference in August.
The PPU provides a speed-up in cases where the computing task is parallelizable, but a traditional CPU isn’t well equipped to take advantage of that parallelism, yet offloading to something like a GPU would be too costly.
Finnish Startup Wants to Build 100x Faster CPUs
Flow Computing has developed a new computing architecture that combines parallel processing units with CPUs. By reducing the overhead of parallelization, the company claims it can effectively make a CPU 100 times faster.To demonstrate the power of the PPU, Forsell and his collaborators built a proof-of-concept FPGA implementation of their design. The team says that the FPGA performed identically to their simulator, demonstrating that the PPU is functioning as expected. The team performed several comparison studies between their PPU design and existing CPUS. “Up to 100x [improvement] was reached in our preliminary performance comparisons assuming that there would be a silicon implementation of a Flow PPU running at the same speed as one of the compared commercial processors and using our microarchitecture,” Flow Computing CTO and co-founder Martti Forsell says.
Now, the team is working on a compiler for their PPU, as well as looking for partners in the CPU production space. They are hoping that a large CPU manufacturer will be interested in their product, so that they could work on a co-design. Their PPU can be implemented with any instruction set architecture, so any CPU can be potentially upgraded.