$ 99 Pocket Cluster
The Adapteva company (about which you most likely hear for the first time) plans to make a supercomputer that will be accessible to everyone . Since 2008, they have been developing energy-efficient RISC processors on the orders of manufacturers of smartphones and other mobile devices.
“We're going down the food chain,” says CEO and founder Andreas Olofsson. But Adapteva wants to give its technology directly to people through a project on Kickstarter, if they collect at least $ 750K with an ultimate goal of $ 3M.
')
Adapteva calls its project “Parallella: A Supercomputer For Everyone”, which is a 16-core processor with a total frequency of all 16GHz cores, 26 GFLOPS performance and a price of $ 99. If they reach the target of $ 3M, they will make a motherboard with a 64-core processor (45GHz and 90 GFLOPS) for $ 199. (Adapteva considers frequency and performance as a sum over all processor cores). Both cards also include a dual-core ARM A9 SoC with a 16- or 64-core RISC processor that acts as a co-processor to the mainstream. Adapteva claims that they achieved energy efficiency of 70 GFLOPS per watt and 25 GHz per watt, respectively.
The promised delivery time (subject to collection of $ 750K) - May 2013 for the 16-nuclear version.
The device will be a fully working computer with Ubuntu 11.10 ARM, 1GB RAM, two USB 2.0 ports, 16GB MicroSD card, HDMI and Gigabit Ethernet. Open source SDK will support C, C ++ and OpenCL. The size of the device will be approximately 8.5 x 5.5 cm (the size of the Raspberry Pi).
There are great similarities between Parallella and projects like the Raspberry Pi (a computer with Linux for $ 35, and an Arduino for $ 30), but Olofsson claims that Parallella will be 10 to 50 times faster than Raspberry Pi (compared to the 64-core version) and while only 3 times more expensive. He notes that with a price of $ 99, this is much cheaper than most current parallel computing platforms. The boards containing Adapteva processors, which their partners sell now cost from several thousand dollars.
As a scope, Olofsson calls the development of mobile and embedded systems, the creation of new programming languages, the study of parallel algorithms, and so on. “What people can do with Parallella is unlimited,” he says. "We hope that they will be used by open source projects by enthusiasts who today lack a platform of this type to meet their needs."
In addition to enthusiasts and developers, Olofsson hopes that parallel computing will be widely distributed among conventional firms. “Today there is a gap between researchers and business as usual,” he says. “I know that in the end parallel computing will become ubiquitous, it’s just a matter of time, but I would prefer it to happen now rather than in three years.”
The definition of a “supercomputer” seems somewhat vague. The slowest supercomputer in top-500 gives 61 TFLOPS. A cluster of hundreds of 16-core Parallellas will cost $ 10,000 and provide 10 TFLOPS, he said. Even if it is not a supercomputer, then it can be useful for so many people.
Adapteva says they need to raise $ 3M through Kickstarter to begin production of the 64-core version of Parallella. While a 16-core chip is produced using a 65-nanometer process, a 64-core chip will be produced using a more complex and expensive 28-nanometer process.
Kickstarter has recently begun to deny iron developers, saying, "Kickstarter is not a store." But Adapteva managed to convince them that they are not ordinary retailers.
But why did Adapteva go to Kickstarter? The company raised $ 2.5 million in venture capital, but Olofsson says it is “very small for an electronics developer. ... Our research budget is probably 1/1000 of that at Intel. To implement this project, we need millions of dollars. We talked with venture capitalists, but for them the “iron” model of a startup no longer works. They don't get a good return on investment. ”
Despite the fact that Parallella is based on existing processors, collecting money is the only way to get a price of $ 99, and they come from the production costs at the Global Foundries factory. And such a small price is the key to the success of the project, says Olofsson.
Original
Talk on slashdot
Now a few words from myself:
If they do what they want, then this will really be a breakthrough in entering the wide market of multiprocessor systems. Unlike GPGPU (NVidia CUDA, etc.), which are in the majority of SIMDs, this thing is really MIMD and each core can execute completely independent code from other cores. I think that with the emergence of Project Glass from google there will be a strong need for such pieces of hardware, for example, for image processing in the implementation of augmented reality.
Yes, while they lose on the performance of video cards, but at the same time their chip eats 2 (!) Watts. Characteristics of a single processor node (collected from different parts of the documentation):
But their main difference is that their architecture, tools, libraries, and APIs are fully OpenSource (hello NVIdia with your binary drivers and architecture under NDA).
Well, even if you abstract from this chip, then IMHO itself development board with ARM A9 for $ 99 is more than normal.
UPD:
Kickstarter survey results regarding possible applications
To the question that it is much weaker than a video card:
In the single-chip version, yes, but the authors added the MINI-CLOUD ($ 495) and CLUSTER ($ 975) variants. For this price, they promise a solution of four / eight boards ( 72/134 processors (just the processor, not the cores, as I understood it)), a gigabit switch and a power supply.
“We're going down the food chain,” says CEO and founder Andreas Olofsson. But Adapteva wants to give its technology directly to people through a project on Kickstarter, if they collect at least $ 750K with an ultimate goal of $ 3M.
')
Adapteva calls its project “Parallella: A Supercomputer For Everyone”, which is a 16-core processor with a total frequency of all 16GHz cores, 26 GFLOPS performance and a price of $ 99. If they reach the target of $ 3M, they will make a motherboard with a 64-core processor (45GHz and 90 GFLOPS) for $ 199. (Adapteva considers frequency and performance as a sum over all processor cores). Both cards also include a dual-core ARM A9 SoC with a 16- or 64-core RISC processor that acts as a co-processor to the mainstream. Adapteva claims that they achieved energy efficiency of 70 GFLOPS per watt and 25 GHz per watt, respectively.
The promised delivery time (subject to collection of $ 750K) - May 2013 for the 16-nuclear version.
The device will be a fully working computer with Ubuntu 11.10 ARM, 1GB RAM, two USB 2.0 ports, 16GB MicroSD card, HDMI and Gigabit Ethernet. Open source SDK will support C, C ++ and OpenCL. The size of the device will be approximately 8.5 x 5.5 cm (the size of the Raspberry Pi).
There are great similarities between Parallella and projects like the Raspberry Pi (a computer with Linux for $ 35, and an Arduino for $ 30), but Olofsson claims that Parallella will be 10 to 50 times faster than Raspberry Pi (compared to the 64-core version) and while only 3 times more expensive. He notes that with a price of $ 99, this is much cheaper than most current parallel computing platforms. The boards containing Adapteva processors, which their partners sell now cost from several thousand dollars.
As a scope, Olofsson calls the development of mobile and embedded systems, the creation of new programming languages, the study of parallel algorithms, and so on. “What people can do with Parallella is unlimited,” he says. "We hope that they will be used by open source projects by enthusiasts who today lack a platform of this type to meet their needs."
In addition to enthusiasts and developers, Olofsson hopes that parallel computing will be widely distributed among conventional firms. “Today there is a gap between researchers and business as usual,” he says. “I know that in the end parallel computing will become ubiquitous, it’s just a matter of time, but I would prefer it to happen now rather than in three years.”
The definition of a “supercomputer” seems somewhat vague. The slowest supercomputer in top-500 gives 61 TFLOPS. A cluster of hundreds of 16-core Parallellas will cost $ 10,000 and provide 10 TFLOPS, he said. Even if it is not a supercomputer, then it can be useful for so many people.
Adapteva says they need to raise $ 3M through Kickstarter to begin production of the 64-core version of Parallella. While a 16-core chip is produced using a 65-nanometer process, a 64-core chip will be produced using a more complex and expensive 28-nanometer process.
Kickstarter has recently begun to deny iron developers, saying, "Kickstarter is not a store." But Adapteva managed to convince them that they are not ordinary retailers.
But why did Adapteva go to Kickstarter? The company raised $ 2.5 million in venture capital, but Olofsson says it is “very small for an electronics developer. ... Our research budget is probably 1/1000 of that at Intel. To implement this project, we need millions of dollars. We talked with venture capitalists, but for them the “iron” model of a startup no longer works. They don't get a good return on investment. ”
Despite the fact that Parallella is based on existing processors, collecting money is the only way to get a price of $ 99, and they come from the production costs at the Global Foundries factory. And such a small price is the key to the success of the project, says Olofsson.
Original
Talk on slashdot
Now a few words from myself:
If they do what they want, then this will really be a breakthrough in entering the wide market of multiprocessor systems. Unlike GPGPU (NVidia CUDA, etc.), which are in the majority of SIMDs, this thing is really MIMD and each core can execute completely independent code from other cores. I think that with the emergence of Project Glass from google there will be a strong need for such pieces of hardware, for example, for image processing in the implementation of augmented reality.
Yes, while they lose on the performance of video cards, but at the same time their chip eats 2 (!) Watts. Characteristics of a single processor node (collected from different parts of the documentation):
- 32-bit single-precision floating point only (no double-precision)
- Core local store global-addressable on 32-bit flat address space; can address
- 32KB local store on each core, 32 GB / s @ 1 GHz
- External interface 2 GB / s * 4 (four directions) = max 8 GB / s
- Global address space
- Fast inter-core write, slow read requests (1 for each 8 cycles)
But their main difference is that their architecture, tools, libraries, and APIs are fully OpenSource (hello NVIdia with your binary drivers and architecture under NDA).
Well, even if you abstract from this chip, then IMHO itself development board with ARM A9 for $ 99 is more than normal.
UPD:
Kickstarter survey results regarding possible applications
To the question that it is much weaker than a video card:
In the single-chip version, yes, but the authors added the MINI-CLOUD ($ 495) and CLUSTER ($ 975) variants. For this price, they promise a solution of four / eight boards ( 72/134 processors (just the processor, not the cores, as I understood it)), a gigabit switch and a power supply.
Source: https://habr.com/ru/post/154267/
All Articles