We are witnessing a major shift from traditional enterprise data centers to much larger warehouse-scale cloud data centers. This is driven by the economics of scale and the benefits of cloud computing, and is happening for both for public and private clouds.
These large data centers need a much higher performance networks that bears little resemblance with traditional enterprise networks. A cloud data center network needs to interconnect many thousands of servers with predictable bandwidth and low-latency.
Our original goal was a switch that could connect 10,000 servers with a simple, 2-stage network, that would deliver predictable Gigabit performance for each server, and do this at a price point that is compatible with web and cloud business models. Just to be clear, such a network requires 10 Terabits/second throughput (10,000 x 1 Gbps), active-active load-sharing redundancy to avoid any single point of failure, and the ability to run 24×7 since there are no maintenance windows in the cloud world.
I am very pleased with the product that resulted from this development, the Arista 7500 data center switch. It turned out really great, even better than we originally anticipated.
The Arista 7500 switch is the highest throughput 10G Ethernet switch in Continue reading
These days there is much discussion whether switches or routers should be built with proprietary custom ASICs or standard “merchant silicon” chips. At one level, the question is “Why does it matter?” After all, networking vendors have been building custom silicon chips since the invention of the LAN switch in the early ’90s.
In my own career, I have led the development of several generations of very high volume network switch silicon. However, even I could not design better silicon switch chips than what is available now on the merchant market. To me this is an inflection point for the industry that is not unlike what happened in the computer industry with the adoption of industry standard architectures.
While CPU and switch silicon architectures differ in many ways, the underlying economics are very similar. In the 1980s and 1990s, CPU architectures flourished: there were MIPS, PowerPC, SPARC, ARM, and X86. Each architecture staked out their position in the market. The RISC architectures led the server market with 64-bit addressing and multi-processing capability, and also focussed on embedded applications. X86 was the standard for desktop computers. However as the years passed, most of the volume growth in the market has Continue reading