Today’s computing power of the whole world is roughly equal to only one human brain. In 2018 the Computing Capacity Worldwide was roughly accounted to be around 2*10^20 – 1.5*10^21 FLOPS. A human brain performs an equivalent of between 3*10^13 to the 3*10^25 FLOPS (Floating Point Operations Per Second) calculations in terms of the computing …
We have been working on the project for quite some time. Megamind gradually becomes more and more what we plan it to be: a computing network with easy access of hardware owners, developer-friendly interface and a robust core which parallelizes all the tasks among the machines which are most economical in this particular case. A …
The system consists of the following elements:– Central server– Docker registry compatible image registry managed by Megamind– Computing nodes– Developers– End users Central server. Megamind Central server consists of the following parts: Server REST API. Provides a top-level API for task management and connecting computing nodes. The task in this context is a docker image …
Russian programmers have developed a platform for distributed training of large neural networks. It is adapted for a network of many computers of different power, any of which can exit the process at any time. As with distributed scientific computing projects such as Folding@home, this approach, with the help of many volunteers, can create a …
Supercomputers for NNs to be replaced by a distributed network Read More »
Professionals in Data Science have a variety of options when it comes to computing. As far as the algorithms can be paralleled (it’s exactly the case with most machine learning models — with a lot of matrix multiplication), model training can be done using CPUs, GPUs and also TPUs. What does it mean from the …
Everybody can agree that the lockdown slowed down all kinds of businesses. Well, it hardly affected Megamind: we continue service development as planned. According to the Roadmap we plan to release Beta in Q4 2020 and become operational in 2021. At the moment we are working on the platform core: in particular load balancer, API/SDK, …
In the previous post we discussed how to measure performance of a computer. Let me remind you what approaches we have tackled: Theoretical “on paper”: by multiplying clock by cores and thus getting FLOPS; Theoretical by benchmarking: by using Linpack utility. The point is that we use CPU to multiply N matrices where we know …
As you know , FLOPS is a unit for measuring the computing power of computers in floating point operations, which is often used to measure whether one computer has more power than the other. It is especially important to measure FLOPS in the world of Top500 supercomputers in order to find out who is the …
Mining has never been a low-risk business. Due to various factors it could either be a success or it could suck all the money out of a miner without giving him any substantial profit. Of course there were some moments in mining history when mining rigs had a payback period shorter than 7 days (e.g. …
We designed Megamind with a big picture in mind: a marketplace of distributed computing resources orchestrated by the central authority. The authority divides computing tasks into batches and sends them to the most appropriate pieces of hardware which are connected to the network. At the moment we are developing API/SDK to enable a variety of …
Architecture and technological solutions of MEGAMINÐ Read More »
