In a recent presentation at the LADIS 2009 workshop, Jeff Dean, Google Fellow in Google's Systems Infrastructure Group, gave an interesting talk on the large scale distributed systems that are in use inside google. Although the presentation does not reveal trade secrets, it provides interesting hints about some of the issues that google is addressing and the size of their infrastructure.

One of the slides also reminds about some numbers that any developer of distributed application should know if he wants to optimise his application :

• L1 cache reference 0.5 ns
• Branch mispredict 5 ns
• L2 cache reference 7 ns
• Mutex lock/unlock 25 ns
• Main memory reference 100 ns
• Compress 1K bytes with Zippy 3,000 ns
• Send 2K bytes over 1 Gbps network 20,000 ns
• Read 1 MB sequentially from memory 250,000 ns
• Round trip within same datacenter 500,000 ns
• Disk seek 10,000,000 ns
• Read 1 MB sequentially from disk 20,000,000 ns
• Send packet CA->Netherlands->CA 150,000,000 ns