Supercomputer-maker Cray has launched the ClusterStor E1000, a storage offering designed to serve the entire triumvirate of HPC workloads: simulations, artificial intelligence, and analytics. …
Cray Revamps ClusterStor For The Exascale Era was written by Michael Feldman at The Next Platform.
SD-branch is the next logical step from SD-WAN, creating a software-defined branch that's easily...
The operator also unveiled a partnership with Siemens to deliver corporate services to German...
NetApp executives say the company’s all-flash and hybrid storage arrays have been a burden on...
A few years ago, about two dozen Oracle employees began work in downtown office space in Seattle to start mapping out how to make the enterprise software giant and occasional system maker a competitive player in a crowded public cloud space that is dominated by the likes of Amazon Web Services and Microsoft Azure, and also includes a host of other high-profile companies, from Google to Salesforce to IBM. …
Oracle Chases A Huge HPC Opportunity In The Cloud was written by Jeffrey Burt at The Next Platform.
Observability in Data Center Networks: In this session, you’ll learn how the sFlow protocol provides broad visibility in modern data center environments as they migrate to highly meshed topologies. Our data center workloads are shifting to take advantage of higher speeds and bandwidth, so visibility to east-west traffic within the data center is becoming more important. Join Peter Phaal—one of the inventors of sFlow—and Joe Reves from SolarWinds product management as they discuss how sFlow differs from other flow instrumentation to deliver visibility in the switching fabric.THWACKcamp is SolarWinds’ free, annual, worldwide virtual IT learning event connecting thousands of skilled IT professionals with industry experts and SolarWinds technical staff. This video was one of the sessions.
“We have the largest developer community in the [IoT] industry. Almost 200,000 folks build their...
How does technical implementation and user feedback shape a cloud-based solution? When is it time to make a significant change in your design? And how do you know you’re headed in the right direction? This Day Two Cloud podcast episode tackles these questions with guest Michael Fraser, co-founder and CEO of Refactr.
The post Day Two Cloud 021: Nice Design; We Need To Change It – The Reality Of Building A Cloud Service appeared first on Packet Pushers.
In June, we announced a wide-scale post-quantum experiment with Google. We implemented two post-quantum (i.e., not yet known to be broken by quantum computers) key exchanges, integrated them into our TLS stack and deployed the implementation on our edge servers and in Chrome Canary clients. The goal of the experiment was to evaluate the performance and feasibility of deployment in TLS of two post-quantum key agreement ciphers.
In our previous blog post on post-quantum cryptography, we described differences between those two ciphers in detail. In case you didn’t have a chance to read it, we include a quick recap here. One characteristic of post-quantum key exchange algorithms is that the public keys are much larger than those used by "classical" algorithms. This will have an impact on the duration of the TLS handshake. For our experiment, we chose two algorithms: isogeny-based SIKE and lattice-based HRSS. The former has short key sizes (~330 bytes) but has a high computational cost; the latter has larger key sizes (~1100 bytes), but is a few orders of magnitude faster.
During NIST’s Second PQC Standardization Conference, Nick Sullivan presented our approach to this experiment and some initial results. Quite accurately, Continue reading
I don’t remember who pointed me to the excellent How Complex Systems Fail document. It’s almost like RFC1925 – I could quote it all day long, and anyone dealing with large mission-critical distributed systems (hint: networks) should read it once a day ;))
Enjoy!
Learning certifiably optimal rule lists for categorical data Angelino et al., JMLR 2018
Today we’re taking a closer look at CORELS, the Certifiably Optimal RulE ListS algorithm that we encountered in Rudin’s arguments for interpretable models earlier this week. We’ve been able to create rule lists (decision trees) for a long time, e.g. using CART, C4.5, or ID3 so why do we need CORELS?
…despite the apparent accuracy of the rule lists generated by these algorithms, there is no way to determine either if the generated rule list is optimal or how close it is to optimal, where optimality is defined with respect to minimization of a regularized loss function. Optimality is important, because there are societal implications for lack of optimality.
Rudin proposed a public policy that for high-stakes decisions no black-box model should be deployed when there exists a competitive interpretable model. For the class of logic problems addressable by CORELS, CORELS’ guarantees provide a technical foundation for such a policy:
…we would like to find both a transparent model that is optimal within a particular pre-determined class of models and produce a certificate of its optimality, with respect Continue reading
Networks are growing, and growing fast. As enterprises adopt IoT and mobile clients, VPN technologies, virtual machines (VMs), and massively distributed compute and storage, the number of devices—as well as the amount of data being transported over their networks—is rising at an explosive rate. It’s becoming apparent that traditional, manual ways of provisioning don’t scale. Something new needs to be used, and for that, we look toward hyperscalers; companies like Google, Amazon and Microsoft, who’ve been dealing with huge networks almost since the very beginning.
The traditional approach to IT operations has been focused on one server or container at a time. Any attempt at management at scale frequently comes with being locked into a single vendor’s infrastructure and technologies. Unfortunately, today’s enterprises are finding that even the expensive, proprietary management solutions provided by the vendors who have long supported traditional IT practices simply cannot scale, especially when you consider the rapid growth of containerization and VMs that enterprises are now dealing with.
In this blog post, I’ll take a look at how an organization can use open, scalable network technologies—those first created or adopted by the aforementioned hyperscalers—to reduce growing pains. These issues are increasingly relevant as new Continue reading
It’s par for the course for AI chip startups to focus on peak performance on outdated benchmarks to appeal to the hardware folks who might give their gear a go for deep learning training or inference. …
Deep Divides Between AI Chip Startups, Developers was written by Nicole Hemsoth at The Next Platform.