More cores is always a good choice for consumer chips - even if one program doesn't effectively use them, running multiple applications in parallel sure as hell does.
And these days, we're running everything under the virtual sun and expect it to work fast and smooth :-)
Any web server that serves at scale can readily use 1024 cores... Nginx, servers built in Go, Erlang and many more. It would reduce hosting costs and complexity substantially.
Re:Intel delays chip-making changes
by dragonwriter on Thursday July 16, 2015 @12:56PM
(#9898388)
Lets say, for the sake of argument, that's true; what percentage of computers are "web servers that serve at scale". What's the cost of setting up production lines for 1024-core chips for just that market -- and how many chips are you going to be able to amortize that cost over.
Even if it was possible, I think the costs would be exhorbitant, even if the marginal cost per chip wasn't.
Re:Intel delays chip-making changes
by JoachimSchipper on Thursday July 16, 2015 @01:39PM
(#9898660)
It's not as easy as you seem to think to e.g. (de-)multiplex a single TCP port to/from 1024 cores, and that's before you actually try to keep any application state in sync. Lots of stuff breaks at that level of concurrency.
First part makes sense, but the last part doesn't. Keeping application state in sync doesn't need to affect the throughput of events coming out of a multi-threaded application.
Those servers are also typically not CPU bottlenecked and while they would have 1024 threads (or even vastly more) they don't really benefit from 1024 cores.
Further, we're in an industry of VPSes where cores are virtualized and divided up. a 128 core CPU could produce 128 single core VPSes with probably better performance.
Of course this is consistent with your point, but your comment is negative, as is my parent comment.
Substantive discussion is not appreciated on hacker news.
Intel has 8-core i7s, up to 18-core Xeon processors, and up to 61 cores in Xeon Phi coprocessor packages.
> If you reduce the feature size and effectively thus double the surface area available, why not increase the cores?
Because, outside of narrow domains, software that can effectively use lots of cores doesn't exist, so you don't provide good bang for the customers buck by doing that. For most customers, there's no utility for that.
> Cache can improve performance but not like more cores.
And vice versa -- more cores can improve performance, but not like cache. Which is more useful depends on software and workloads.
> Under Erlang/Elixir, we are able to get close to linear speedup from additional cores.
On the right kind of workloads. And, the people with those kind of workloads, have many-core processors available.