NVIDIA Introduces EGX Edge Supercomputer


A couple of weeks ago at the Mobile World Congress event in Los Angeles, NVIDIA unveiled its EGX Edge Supercomputing platform. The product is designed for all edge applications including, but not limited to, artificial intelligence and internet of things. This was the first time NVIDIA has been part of a Mobile World Congress.

EGX is a reference design that other server manufacturers can use to roll out a validated design for edge computing solutions using NVIDIA GPUs. Cisco, Fujitsu, Lenovo, HPE and others have been confirmed as partners. Powered by NVIDIA’s CUDA Tensor Core GPU, it offers cryptographic acceleration, networking for Mellanox, NVMe over TCP and RDMA for storage, an industrial-strength cloud native stack, and hardened AI software.

Computing is moving to where the data resides, and this will take power. 5G is one of the key enablers of edge computing. 5G brings multi-gigabit speeds to wireless and enables more data to be processed in more places. One view is that all data can be processed in the cloud, but that doesn’t work in for everyone.

Another factor behind edge computing is the rise of the IoT. We live in a world were literally everything is connected and creating massive amounts of data to be analyzed. Most traditional computing platforms don’t have the horsepower for AI with large data sets.

The NVIDIA Edge Stack software is optimized for real-time AI uses cases, such as video analytics, video, speech and audio as well as Red Hat OpenShift for Kubernetes container orchestration. EGX also has support for NVIDA Metropolis, which is an IoT-optimized ecosystem for smart cities, smart factories, traffic engineering, retail and other places where large amounts of sensor data are being pulled in and analyzed.

The edge is where the emerging action is today and where the mainstream action will be soon. The cloud isn’t going away, but more and more businesses will deploy edge computing solutions to analyze data in more places. The GPU based EGX is ideally suited for many of the advanced edge use cases.

Source: Zeus Kerravala, eWeek, November 01, 2019.

Comments

Post a Comment

Popular posts from this blog

EOQ Calculations in Excel

Image
Last week in Operations Management we used Excel to calculate the Economic Order Quantity and graph Carrying Costs, Ordering Costs, and Total Costs. The Economic Order Quantity or "EOQ" is the order size that "minimizes" Total Costs. Any more or less and you are spending too much on ordering or too much on keeping inventory. For example, in the Excel spreadsheet below, if you had an Annual Demand of 12000 units, Ordering Costs of $10 per order, and Holding costs of $4 per unit per year, the EOQ would be 245 units and Total Costs would be $980.00: We used the following formula in Excel to calculate EOQ: =SQRT((2*B2*B3)/B4) And the following formula to calculate Total Costs at this point: =$B$6/2*$B$4+$B$2/B6*$B$3 To create the graph, we used the following formulas and simply copied them over a range of 100 to 500 units. Ordering Costs: =$B$2/D2*$B$3 Holding Costs: =D2/2*$B$4 Total Costs: =F2+G2 Looking at this chart, we can clearly see that our order size of 245 is ...
Read more

Reliability Calculations in Excel

Image
In Operations Management today we used Excel to calculate product reliability. Reliability is defined as "the ability of a product, service, part, or system to perform its intended function under a prescribed set of circumstances" (Stevenson, 2009). In order to calulate reliability over a given length of time we need to use the following formula: P(no failure) = e -t/MTBF where: e = 2.7183 t = Length of service before failure MTBF = Mean Time Between Failures One of the cases we did in class involved an auto detailing company that wanted to calculate the reliability of its vacuum cleaners. Excel provides an "EXP" funtion that we can easily apply here. After entering all data for the problem as shown on the spreadsheet below, we type the formula =EXP(-A5/$B$2) in cell B5 and copy down. Next we plot the distribution for t=0 to the maximum life expectancy of 20 years using a line graph. What we end up with looks like a classic case of "ex...
Read more

“How Are We Doing?” Efficiency, Utilization, and Productivity

Image
In business, knowing how we are performing in relation to our competition is vital. Internally we want to know if we are using resources effectively and getting the most out of what we have. Efficiency and Productivity are two calculations that are fundamental to the success of any organization. While the calculations are simple, choosing what measures to use and collecting the necessary data requires a thorough understanding of your industry. To calculate Efficiency you will have to first decide what to use as a base for comparison. Does your company already have a “target” of some kind? Like total units produced? If so then you can compare everything to that. Once you have that, there are a couple of basic formulas that can be used: Efficiency = Actual Output/Effective Capacity X 100% Utilization = Actual Output/Design Capacity X 100% Design Capacity is the maximum output achieved under ideal conditions. Effective Capacity is always less than Design Capacity because of factors...
Read more