Chris wrote:
> Hypothetically, if you can guarantee that those threads get a specified
> amount of time, but may possibly get *more* cpu time and thus finish
> faster, what's the problem?
Each thread is already getting ~100% of one dedicated CPU. It can't
get any *more* than that ;). 256 active threads, 256 CPUs, *very*
tightly coupled, in this example, on say a 512 CPU system running
unrelated stuff on the other half.
If anything else interferes with the memory bandwidth of even one thread
to its node local memory, or takes any of that node local memory for
unrelated uses, it's a big problem. A few percent variation in job
runtime, for a two day job, means hours difference in the finish time.
People notice, when that's your 'big money' app.
This is not your fathers PC ;).
--
I won't rest till it's the best ...
Programmer, Linux Scalability
Paul Jackson <[email protected]> 1.925.600.0401
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
- References:
- [RFC][PATCH] UBC: user resource beancounters
- From: Kirill Korotaev <[email protected]>
- [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Kirill Korotaev <[email protected]>
- Re: [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Rohit Seth <[email protected]>
- Re: [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Alan Cox <[email protected]>
- Re: [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Rohit Seth <[email protected]>
- Re: [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Kirill Korotaev <[email protected]>
- Re: [RFC][PATCH 4/7] UBC: syscalls (user interface)
- From: Andrew Morton <[email protected]>
- Re: [PATCH 4/7] UBC: syscalls (user interface)
- From: Andrey Savochkin <[email protected]>
- Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- From: Dave Hansen <[email protected]>
- Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- From: Andrew Morton <[email protected]>
- Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- From: Chandra Seetharaman <[email protected]>
- Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- From: Paul Jackson <[email protected]>
- Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- From: "Chris Friesen" <[email protected]>
- [RFC][PATCH] UBC: user resource beancounters
- Prev by Date: Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- Next by Date: Re: [PATCH 3/7] [I/OAT] Don't offload copies for loopback traffic
- Previous by thread: Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- Next by thread: Re: [ckrm-tech] [PATCH 4/7] UBC: syscalls (user interface)
- Index(es):
 |