- Behavioral Economics Guide 2016 : Document looking at behavioral economics, the key takeaway that can be applied almost anywhere: when facing a problem or situation where all outcomes and probabilities are known and where the past determine the future (or looks like it), then you should leverage fine-tuned solutions such as complex optimization techniques. However if you have ( a significant ) uncertainty, then using robust solutions, including simple heuristics deliver better results. Sadly, the human nature tend to go for complexe solution in uncertain situation..
- Tofu 2 : system interconnect (25Gbps) designed for the successor model of the FUJITSU Supercomputer PRIME HPC FX10. [slides]
- Low Latency Analytics and Edge Connected Vehicles with RapidIO Interconnect : Interesting to see that rapidio interconnect is still alive and kicking. Looks like they are positioning as a full network stack solution ranging from cpu - cpu to box - box and with emphasis on IoT / analytic
Wednesday, June 29, 2016
Tuesday, June 28, 2016
[Links of the day] 28/06/2016 : Heterogeneous Data-centers (FPGA - GPU) and User-mode Ethernet verbs
- Heterogeneous datacenters : Datacenter slowly evolve and we start to see the emergence of specialized dedicate hardware to squeeze the maximum efficiency per watts consumed. As general CPU start to hit their limit, user turn , like the HPC community, to FPGA or GPU to break away of the current power wall.
- A quantitative analysis on microarchitectures of modern CPU-FPGA platforms : Well if you want to venture in the heterogeneous datacenter you better read this paper on the different platform available out there.
- User Mode Ethernet Verbs : Probably the only serious contender to Intel DPDK, User mode verbs expose Verbs API allowing both user-mode applications/ULPs direct access to offload capabilities in the form of Raw Ethernet QPs. Basically you can send receive raw ETH packet using RDMA verbs API and leverage offload engine to accelerate the operations [video]
Monday, June 27, 2016
- CS 179 GPU Programming : Caltech GPU programming course
- Systems and application for Persistent Memory : Subramanya R. Dulloor PhD thesis on persistent memory programming. A must Read for anybody interested in the next generation storage class memory as the Author wrote PMFS as part of his thesis work.
- WrAP: Hardware and Software Support for Atomic Persistence in Storage Class, master thesis providing some insight on how complex and far reaching are the new storage class memory system and especially with their interaction with CPU cache.
Friday, June 24, 2016
[Links of the day] 24/06/2016: Velocity Conference, Cloud provider infrastructure investment, Connectx5
- Velocity 2016 : Slides and video of O'reilly Velocity 2016 conference.
- Infrastructure Investments by Cloud Service Providers : A fantastic work by Redmonk folks looking at the investment of the main Cloud service provider and what you can learn from this information. What is interesting is that it clearly show the lead that AWS and Microsoft has over all the other one. Also, it seems that SAP is not even on the radar which speak really loudly about their "cloud" strategy. It seems that they are desperately falling behind and might not be able to catch up..
- ConnectX5 : 100GBps interface are out there, but what is really interesting is the switch-less fabric capabilities. With the current race for higher number of port as well and NFV / SDN complexity explosion it seems interesting to see that we might see an emergence of switch-less infrastructure. Moving the switching and other network functionality in the NIC or server itself ( container / VM / unikernel ). Eliminating altogether the need for dedicated network hardware.
Thursday, June 16, 2016
[Links of the day] 16/06/2016 : Bayesan Betancourt's lecture, Physic breakthrough & Distributed systems, Linux Kernel Radix tree
- Betancourt Binge : Michael Betancourt’s video lectures in Tokyo, all about Bayesian model.
- Standing on Distributed Shoulders of Giants : as usual excellent ACM queue article drawing parallel between physics breakthrough and the world of distributed systems.
- Multi-order radix tree : The Linux kernel radix tree is a data-structure at the centerpiece of the memory management system. With the advent of new memory model ( persistent one). This data-structure needs to evolve. However, like with everything touching memory, it will take some time and many many re-submission to the mailing list. This article present a possible evolution path for this venerable data-structure.
Tuesday, June 14, 2016
[Links of the day] 14/06/2016: Freeze Frame File System, Strata+Hadoop16 Keynotes, Linky French Smart meter
- Freeze Frame File System : file system specifically designed to handle real time data stream processing applications. Basically it is able to handle streams of updates while supporting "temporal reads" on demand. To do so the system keep all update history in a memory-mapped log, cache recently retrieved data for repeat reads, and use a hybrid of a real-time and a logical clock to respond to read requests in a manner that is both temporally precise and causally consistent.
- Strata + Hadoop 2016 : All keynote like if you where there.
- Linky : This is a fantastic analysis (in french) of the electric LINKY smart meter that is being deployed across France. It dispel many of the FUD surrounding smart meter but also ask interesting question regarding upgradability , security and who really benefit from these devices. It seems that the customers won't be able to interact or benefit form the data generated by these device, which partially defeat the purpose of installing one. But it's not like they have a choice as the deployment of smart meters is written in french law.
Monday, June 13, 2016
[Links of the day] 13/06/2016 : Geometric Algorithms, Logical Clock are easy & distributed system testing
- Geometric Algorithm : Princeton algorithms and data structures lecture. Provide an excellent overview of the geometric algorithms out there and why they are important.
- Why Logical Clocks are Easy : well not really but still , AMC queue article on the ever recurrent issue of time in distributed systems.
- Distributed Systems Testing : When you cannot find good article on a subject, ask on twitter. The result is a good list of distributed system testing papers.
Wednesday, June 01, 2016
- Megaprocessor : Somebody decided to build a processor using actual individual transistor component. This impressive both in scale and dedication. [videos ]
- Artemis : The ARM Cortex A73 has been released. It is the 64-bit successor to the Cortex A17 . Yes,this is confusing, the naming and release cycle doesn't really indicate the actual generation / functionality. It is clearly aimed at the smartphone market with its little/big core architecture and is a marvelous piece of tech.
- AsAP : Asynchronous Array of Simple Processors , remind me of the connection machine architecture. Except that it use a single-chip processing system comprised of a large number of fine-grain asynchronously-operating programmable processors connected by a reconfigurable network.