This machine is EOL. Check the revised and upgraded version here.
One of the most important problems with utilizing the computational power of modern processors is their overall energy inefficiency. Most of the electrical energy used by a GPU or a CPU eventually turns into heat, which then is basically wasted. Moreover, we have to spend even more energy to protect the system from overheating using external cooling.
Results for image classification models from tensorflow.org is 500 (image\sec), allowing to accelerate your AI and Deep Learning tasks to the level, previously available just to the supercorps.
Never get distracted from your work by the screaming supercomputing equipment, thanks to the A1’s liquid cooling.
Your office is not a sauna, so why to allow your equipment to heat it? Thanks to liquid cooling, your equipment stay cool as well as your working environment.
Plug it in and it just works. Enjoy the convenience of the system, which is fine-tuned to your computational needs.
Liquid cooling is the core of our know-how. It allows to work in the normal temperature, so the system will serve you well for a long time to come. Most Comino components are designed to be used 24/7. We use only high quality and cost effective components.
Heat recovery. Our devices do not heat the space in vain. The heat generated can be reused for other purposes, such as remote areas and spaces or water heating.
Is used to achieve the maximum efficiency of heat recovery, cooperating directly with the inventor of this technology.
Typically, heavily loaded computing systems emit large amounts of noise, due to the use of high-speed air fans. The fanless system is not only way more effective, but almost completely silent. Our devices are probably quieter than your fridge!
Of risers, cases, plastic and copper components. The Comino’s production facility is located in Germany.
We offer round-the-clock online customer support and provide a warranty for our devices for a period of 12 months.
Comino liquid cooling system enables overall increase in performance of GPU servers
Reduce cooling energy consumption by 40%
Reduce overall annual power consumption by 10M kWh
Operate with a Power Utilization Effectiveness (PUE) of 1.05*
Reduce data center build costs by up to 60%
High Quality. Plastic and copper components are produced in Germany.
* PUE is stated for the systems with the external liquid cooling infrastructure