On Thu, 12 Mar 2015 17:50:20 -0500
kazutomo wrote:
> On 03/12/2015 05:05 PM, Jacob Pan wrote:
> > On Thu, 12 Mar 2015 16:59:43 -0500
> > kazutomo wrote:
> >
> >>> according to the document, future DRAM energy unit will always be
> >>> hardcoded to 15.3uJ, no enumeration since there is no domain
On 03/12/2015 05:05 PM, Jacob Pan wrote:
> On Thu, 12 Mar 2015 16:59:43 -0500
> kazutomo wrote:
>
>>> according to the document, future DRAM energy unit will always be
>>> hardcoded to 15.3uJ, no enumeration since there is no domain
>>> specific MSR for energy unit.
>> Which document are you ref
On Thu, 12 Mar 2015 16:59:43 -0500
kazutomo wrote:
> > according to the document, future DRAM energy unit will always be
> > hardcoded to 15.3uJ, no enumeration since there is no domain
> > specific MSR for energy unit.
> Which document are you referring to?
Intel DocID:330784-001 Sept 2014
Int
On Thursday, March 12, 2015 01:17:39 PM Jacob Pan wrote:
> On Wed, 11 Mar 2015 23:01:30 +0100
> "Rafael J. Wysocki" wrote:
>
> > On Wednesday, March 11, 2015 05:55:09 AM Jacob Pan wrote:
> > > The current driver assumes all RAPL domains within a CPU package
> > > have the same energy unit. This i
On 03/12/2015 04:24 PM, Jacob Pan wrote:
> On Wed, 11 Mar 2015 16:25:52 -0500
> kazutomo wrote:
>
>> Hi Jacob,
>>
>> Wow, this is a really pitfall for people who are writing their own
>> RAPL tool.
>> Anyway, I've tested your patch on a Haswell system (2699v3), running a
>> dgemm
>> benchmark. NOT
On Wed, 11 Mar 2015 16:25:52 -0500
kazutomo wrote:
> Hi Jacob,
>
> Wow, this is a really pitfall for people who are writing their own
> RAPL tool.
> Anyway, I've tested your patch on a Haswell system (2699v3), running a
> dgemm
> benchmark. NOTE: userspace governor is selected. All core are set
On Wed, 11 Mar 2015 23:01:30 +0100
"Rafael J. Wysocki" wrote:
> On Wednesday, March 11, 2015 05:55:09 AM Jacob Pan wrote:
> > The current driver assumes all RAPL domains within a CPU package
> > have the same energy unit. This is no longer true for HSW server
> > CPUs since DRAM domain has is own
On Wednesday, March 11, 2015 05:55:09 AM Jacob Pan wrote:
> The current driver assumes all RAPL domains within a CPU package
> have the same energy unit. This is no longer true for HSW server
> CPUs since DRAM domain has is own fixed energy unit which can be
> different than the package energy unit
Hi Jacob,
Wow, this is a really pitfall for people who are writing their own RAPL
tool.
Anyway, I've tested your patch on a Haswell system (2699v3), running a
dgemm
benchmark. NOTE: userspace governor is selected. All core are set to 2.3
GHz.
No power cap is set.
# before the patch is applied
$
The current driver assumes all RAPL domains within a CPU package
have the same energy unit. This is no longer true for HSW server
CPUs since DRAM domain has is own fixed energy unit which can be
different than the package energy unit enumerated by package
power MSR. In fact, the default HSW EP pack
10 matches
Mail list logo