Libre Domain-Specific Architectures
Every couple of years I run across a presentation that impacts my way of thinking about (network) technology.
read more...Networking:
Every couple of years I run across a presentation that impacts my way of thinking about (network) technology.
read more...I am using an OpenWRT based [Turris Omnia](https://www.turris.cz/en/omnia/) router with my ISP XS4ALL. Until recently, I used a simple VLAN setup, but that is not supported anymore. So I have moved to a routed setup.
read more...On 5 March Intel announced that is successfully integrated its 1.6 Tbps silicon photonics engine with a 12.8 Tbps Barefoot Tofino 2 ASIC. This switch uses optical fiber to the front panel ports instead of the usual copper tracks on the PCB (Printed Circuit Board) to QSFP28 front panel cages.
read more...Segment routing is a form of source routing in which a node that originates a packet (the headend node), adds an ordered list of segments (or instructions) to that packet. Such an ordered list is called a segment routing (SR) policy. These segments tell the routers along the path how that packet should be forwarded. Only the headend node to a segment routing domain maintains per flow state. There is no per flow state inside the network.
read more...For about half a century we have been living in a world where the speed of computers grew at an exponential rate. This is known as Moore's law, which is actually an observation of Gordon Moore that the number of transistors in an integrated circuit doubled approximately every two years. But today, that rate has almost levelled off. And Moore's law is not the only exponential that has come to its end. The same is true for Dennard scaling. In networking we are also hitting limits. We have reached the Shannon limit in optical communication. And in network ASIC design we reached the limit of serial bandwidth I/O. The next sections explore the current challenges and the final sections describe R&D that is exploring ways to overcome these challenges.
read more...Hash tables play an important role in routers and switches because lookups need to done in constant time. Minimal size packets (64 bytes on Ethernet) on a 100 Gb/s link arrive at a rate of once every 6.7 ns. All packet handling (lookup, rewrite of next hop, updating CRC, etc.) must be done within that time. Hash tables satisfy that requirement and they are typically used for exact match lookups, such as the MAC addressing forwarding table or host IP route lookups.
read more...The BPF framework can be used to load programs into the Linux kernel at runtime. It can be used for both tracing and for network programming. The BPF code runs in a virtual machine inside the kernel. BPF has a small (around 100 opcodes) RISC-like 64-bit Instruction Set Architecture (ISA). It uses 11 64-bit registers and a 512-byte stack. BPF programs can be written in various languages (C, Go, Rust, etc.) and compiled with clang/LLVM to BPF bytecode.
read more...