There are a number of JavaOne talks on concurrency with PDFs available (ok - that's definately not a blog, but the amount of content made available each year might take a year to absorb.)
Joe Duffy addresses concurrency and parallel programming a great deal in his archives.
He has a decidedly .Net spin, having been the PM for the Parallel extensions, but the principles he discusses are pretty universal. You'll just need to translate to the appropriate Java APIs. Only thing I can think of that might be fundamentally different is low level stuff that's dependent on acquire/release semantics of the underlying memory model.
Just a few recommendations on top of what's already been discussed, in approximate descending order of usefulness (according to me, anyway):
Alex Miller (as mentioned by spdenne) maintains a concurrency link blog on tumblr which aggregates a lot of cool Java concurrency stuff
Cliff Click's blog -- an incredibly smart guy who works at Azul Systems, where they pretty much take Java concurrency to the extreme (we're talking machines with, like, a bajillion cores and a squillion gig of RAM)
Jeremy Manson's blog -- pretty much the guy (well, one of the guys) on Java memory model/concurrency/performance related issues (he was one of the authors of the new memory model)
Java Performance Tuning -- combination of articles and links related to performance, often touches on concurrency stuff
Kirk Pepperdine, Java performance wiz -- more about general performance than concurrency specifically, but talks about concurrency sometimes (of course)
发布评论
评论（4）
Sutter's Mill（Herb Sutter 关于软件、硬件和并发性的文章）是一个博客许多关于并发的帖子。
Sutter’s Mill (Herb Sutter on software, hardware, and concurrency) is a blog with many posts about concurrency.
Joe Duffy 在他的档案中大量讨论了并发和并行编程。
他对 .Net 有明显的兴趣，曾担任并行扩展的 PM，但他讨论的原则非常普遍。您只需要转换为适当的 Java API。我能想到的唯一可能有根本不同的是依赖于底层内存模型的获取/释放语义的低级内容。
Joe Duffy addresses concurrency and parallel programming a great deal in his archives.
He has a decidedly .Net spin, having been the PM for the Parallel extensions, but the principles he discusses are pretty universal. You'll just need to translate to the appropriate Java APIs. Only thing I can think of that might be fundamentally different is low level stuff that's dependent on acquire/release semantics of the underlying memory model.
除了已经讨论过的内容之外，还有一些建议，按照实用性的大致降序排列（无论如何，根据我的说法）：
Just a few recommendations on top of what's already been discussed, in approximate descending order of usefulness (according to me, anyway):