Return-Path: Received: from mail-qc0-f169.google.com ([209.85.216.169] verified) by media-motion.tv (CommuniGate Pro SMTP 4.2.10) with ESMTP-TLS id 4924993 for AE-List@media-motion.tv; Fri, 14 Dec 2012 10:41:30 +0100 Received: by mail-qc0-f169.google.com with SMTP id t2so2380227qcq.28 for ; Fri, 14 Dec 2012 01:45:42 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-type:x-gm-message-state; bh=A2ZLANGP8wyr7eh3yJBRIdtAXTRaFDysye8h+xNOUe8=; b=HHtWNav73ybAJxR/8MEasP+fgkpAqANvrMdQzOobImZc8ljpVJ93a/MCVT7q1M5VAq tLL+syIxEqUt3tu8GZTpXhaBIGRwada7oVRt7vc/LEfRjvbqAPVA3GSwVwIPOHOjZfn7 4PDiXdiRdmhkoxKwGJfE82pGjzqVlgdQv6pPS9LCXn7o+KIfpihdzJENoMIDiZgwvK7C CXOG9TvElq77L0CxZwRpAm7mCJGbYHHT9W51aHAIGQPCHVAbkTauBw4AjXC325twz0Dh 4czThAYd+GVfzhf5h9CQMKcOPEGIPCOIf4woD/YtNHvNIoZ5hEW8UY86Y/t39ELff64w wI8A== Received: by 10.49.62.196 with SMTP id a4mr2498634qes.14.1355478341954; Fri, 14 Dec 2012 01:45:41 -0800 (PST) MIME-Version: 1.0 Received: by 10.49.60.161 with HTTP; Fri, 14 Dec 2012 01:45:20 -0800 (PST) In-Reply-To: References: From: Robert Kjettrup Date: Fri, 14 Dec 2012 10:45:20 +0100 Message-ID: Subject: Re: [AE] 32core benchmarks To: After Effects Mail List Content-Type: multipart/alternative; boundary=047d7bdc067a12acdf04d0cce55b X-Gm-Message-State: ALoCoQnxUY+jnSsEO88Qx5NLxklzzQQivNThHFeBpZtMiiwlCyjUKE0AKmBIDGG+6qAR5GzMiTDz --047d7bdc067a12acdf04d0cce55b Content-Type: text/plain; charset=ISO-8859-1 Hi Teddy So that i have filled all 16 slots on the motherboard is also one of the reasons for my lower memory score? There is allways a tradeoff, and i am quite happy with more memory vs lightning fast memory. AE eats it all up, and being able to rampreview several minutes 1080/25p timelines is a nice luxury :-D - Robert 2012/12/13 Teddy Gage > Yes over clocking the CPU definitely speeds up the ram bus, on top of > which I have quad channel memory in just the four quad channel slots (4x8gb > sticks 1600 MHz) on this mobo. If all ram slots are filled, the memory > bandwidth is decreased dramatically, so it's a balance between speed, over > clocking, price and capacity. > > As you can see by Danny's charts, Cuda performance is definitely gimped on > the 680 vs the 580. I have a 580 3gb card in my secondary render > box that blows away the 680. Sadly it has no 3 monitor support so the 680 > is in my main workstation. I'd highly recommend buying up a 580 or 590 on > eBay as they are currently the best price / performance cards for Cuda > right now. Not to mention easily overclocked out of the box with the > manufacturers software. > > > On Wednesday, December 12, 2012, Robert Kjettrup wrote: > >> Hi >> >> Just for fun i tried running my new 16core (32 threads) Z820 through some >> of these tests, and comparing it there also seems to be some odd numbers >> for your GPU in cineBench. I have the new Quadro K5000 that as far as i >> know is a lower clocked version of the GTX680 you have in your machine, but >> my OpenGL fps number are much higher. I only got the K5000 because i have >> had some bad experience with my former GTX card that just crashed the >> OpenGL driver when it was under load for longer periods. Would have gone >> with the GTX680 if i trusted the gamer cards more, but "known" stability >> was prioritised. >> >> CineBench scores: >> OpenGL 81.45 fps >> CPU 25.07 pts >> >> And running GeekBench (the free 32bit version) i can see that Teddy has >> some impressive memory numbers, higher than i get. Is the memory linked to >> the CPU clock frequency also? 4.5Ghz is a bit higher than my 3.1Ghz :-D >> >> Integer Processor integer performance 32414 >> Floating Point Processor floating point performance 42123 >> Memory Memory performance 7071 >> Stream Memory bandwidth performance 8585 >> Geekbench score: 28360 >> >> not to show off, but i thought some of you guys would like to see the >> numbers of the current top of the line intel xeon >> well, but i also guess that this machine has cost ALOT more than what you >> spent on yours. Intel/HP sure know how to price their current highend stuff >> :-) >> >> - Robert >> >> >> >> 2012/12/9 lsslss >> >> hi Teddy,**** >> >> ** ** >> >> yes you are right, it's at 1600MHz with latency 11 CLK at the moment, >> while it should be at 9.**** >> >> I'm going to figure out how to change that, will have set the voltage to >> manual probably in the BIOS.**** >> >> ** ** >> >> don't know if enabling the IOMMU will do anything for the speeds...**** >> >> ** ** >> >> kind regards,**** >> >> ** ** >> >> Marcel van Berk**** >> >> ** ** >> >> *Hoge Bomen/LSSLSS* >> >> info@lsslss.nl**** >> >> ** ** >> >> www.lsslss.nl**** >> >> ** ** >> >> *Van:* After Effects Mail List [mailto:AE-List@media-motion.tv] *Namens *Teddy >> Gage >> >> *Verzonden:* zondag 9 december 2012 21:11 >> >> >> *Aan:* After Effects Mail List >> *Onderwerp:* Re: [AE] 32core benchmarks**** >> >> ** ** >> >> not bad, but I may have spotted one of your bottlenecks: your ram memory >> performance and bandwidth are less than a third of the scores on my >> machine, you may want to check that your ram timings and voltage are >> correct in your bios. My RAM is 32 gb 1600 MHZ 7-7-7-20**** >> >> >> intel core i7 3930K hex core @ 4.5 ghz**** >> >> *Section* >> >> *Description* >> >> *Score* >> >> *Geekbench Score* >> >> Geekbench 2.4.0 Tryout for Windows x86 (32-bit)**** >> >> Integer**** >> >> Processor integer performance**** >> >> 17354**** >> >> 16973**** >> >> Floating Point**** >> >> Processo >> >> > > -- > Animator & Editor > www.teddygage.com > Brooklyn > > --047d7bdc067a12acdf04d0cce55b Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Hi Teddy

So that i have filled all 16 slots on the mothe= rboard is also one of the reasons for my lower memory score? There is allwa= ys a tradeoff, and i am quite happy with more memory vs lightning fast memo= ry. AE eats it all up, and being able to rampreview several minutes 1080/25= p timelines is a nice luxury :-D

- Robert



2012/12/13 Teddy Gage = <teddygage@gmai= l.com>
Yes over clocking the CPU definitely speeds up the ram bus, on top of which= I have quad channel memory in just the four=A0quad channel slots (4x8gb st= icks=A01600 MHz) on this mobo.=A0If all ram=A0slots are filled, the memory = bandwidth is decreased dramatically, so it's a balance between speed, o= ver clocking, price and capacity.=A0

As you can see by Danny's charts, Cuda performance is de= finitely gimped on the 680 vs the 580. I have a 580 3gb card in my=A0= secondary =A0 render box=A0that blows away the 680. Sadly it has no = 3 monitor support so the 680 is in my main workstation. I'd highly reco= mmend buying up a 580 or 590=A0on eBay as they are currently the best price= / performance cards for Cuda right now. Not to mention easily overclocked = out of the box with the manufacturers software.


On Wednesday, December 12, 2012, Robert Kjettrup wrote:
Hi

Jus= t for fun i tried running my new 16core (32 threads) Z820 through some of t= hese tests, and comparing it there also seems to be some odd numbers for yo= ur GPU in cineBench. I have the new Quadro K5000 that as far as i know is a= lower clocked version of the GTX680 you have in your machine, but my OpenG= L fps number are much higher. I only got the K5000 because i have had some = bad experience with my former GTX card that just crashed the OpenGL driver = when it was under load for longer periods. Would have gone with the GTX680 = if i trusted the gamer cards more, but "known" stability was prio= ritised.

CineBench scores:
OpenGL 81.45 fps
CPU 2= 5.07 pts

And running GeekBench (the free 32bit ver= sion) i can see that Teddy has some impressive memory numbers, higher than = i get. Is the memory linked to the CPU clock frequency also? 4.5Ghz is a bi= t higher than my 3.1Ghz :-D

Integer Processor integer performance 32414 =
Floating Point Processor f= loating point performance 42123=
Memory Memory performance<= span style=3D"white-space:pre-wrap"> 7071
Stream Memory bandwidth performance 8585
Geekbench score: 28360

not to show off, b= ut i thought some of you guys would like to see the numbers of the current = top of the line intel xeon
well, but i also guess that this m= achine has cost ALOT more than what you spent on yours. Intel/HP sure know = how to price their current highend stuff :-)

- Robert



2012/12/9 l= sslss <info@lsslss.nl>

hi Teddy,

=A0

yes you are right, it's at 1600MHz with laten= cy 11 CLK at the moment, while it should be at 9.

I'm going to figure out how to change that, w= ill have set the voltage to manual probably in the BIOS.

=A0

don't know if enabling the IOMMU will do anything for the speeds..= .

=A0

kind regards,

=A0

Marcel van Berk=

=A0

Hoge Bomen/LSSLSS

info@ls= slss.nl

=A0

www.lsslss.nl

=A0

Van: After = Effects Mail List [mailto:AE-List@media-motion.tv] Namens Ted= dy Gage


Verzonden: zondag 9 december 2012 21:11


Aan: After Effects Mail List
Onderwerp: Re= : [AE] 32core benchmarks

=A0

not bad, but I may have spotted one of your bott= lenecks: your ram memory performance and bandwidth are less than a third of= the scores on my machine, you may want to check that your ram timings and = voltage are correct in your bios. My RAM is 32 gb 1600 MHZ 7-7-7-20<= u>


intel core i7 3930K hex core @ 4.5= ghz

Floating Point

Section

Description

Score

Geekbench Score

Geekbench 2.4.0 Tryout for Windows x86 (32-bit)

Integer

Processor integer performance

17354

16973

Processo

=


--
<= span style>Animator & Editor
www.teddygage.com
Brooklyn


--047d7bdc067a12acdf04d0cce55b--