Synced or Unlinked? Results inside!

mythy · 12-07-07, 06:59 PM

So here you have it! I have benchmarked my Q6600 and DDR2-8500 ram at all available ratios/speeds and at 2 different core clocks! I have chosen 3 and 3.4GHz to test since they are the most common OC's around.

I have used the following tests:
*Sandra Lite Memory bandwidth
*Sandra Lite Memory latency
*MediaCoder Benchmark

First set of results are the most important since they compare the ram at a Synced state and a Unlinked! First you will see the bandwidth followed by latency and finally real world performance... Test are done in both 3 and 3.4GHz clock speeds.

Sandra Lite Memory bandwidth

Sandra Lite Memory latency

MediaCoder Benchmark

These next tests show you all the different speeds and strap available to you and there performance results.

Sandra Lite Memory bandwidth

Sandra Lite Memory latency

MediaCoder Benchmark

Sandra Lite Memory bandwidth

Sandra Lite Memory latency

MediaCoder Benchmark

So there you have it! Synced is slower and the lesser speed adds to its latency! While the performance gains arnt huge I should point out that every ns counts!

CaptNKILL · 12-07-07, 08:24 PM

Well I feel dumb... what is the "strap"?

I don't think I've ever heard that term before.

mythy · 12-07-07, 08:45 PM

Its a internal dram ratio inside the northbridge. Not all board allow you to change it. But you can see how it effects the clocks

$n][pErMan · 12-07-07, 08:47 PM

interesting... I wonder why the results are such... were the timings of the memory changed in the different strap modes or did you leave the timings alone? Generallly when running at a lower strap you can tighten the timings of the memory by quite a bit.

mythy · 12-07-07, 08:52 PM

I can run 4-4-4-12 to 533 but for the tests it was all at stock. and to be honest I noticed no performance increases at 4-4-4-12

The speed more then makes up from any gains you might see from the tighter timings.. I know that goes against the general hear say but hey what ever

$n][pErMan · 12-08-07, 01:36 PM

Did you try any game benchmarks by chance?

mythy · 12-08-07, 01:44 PM

do you have any ideas how many reboots I had to do?

I was at it for over a hour...

Never the less I am planing a retest at 2.4GHz stock at 1:1 and max dividers for my 1066 ram. At 1:1 I will use 4-4-4-12 and at 1066 I will use 5-5-5-15. I will run threw the crysis benchmark for results

Revs · 12-08-07, 03:24 PM

So you're saying i'm loosing bandwidth from having setup in 'Sync' mode. Hmm, I shall have to test this myself.

I'll be back!

**mullet** · 12-08-07, 04:03 PM

7x500 1:1 400 strap

Sync Mode
Int Buff'd iSSE2: 8517 MB/s
Float Buff'd iSSE2: 8517 MB/s

Memory latency
70ms

Revs · 12-08-07, 04:04 PM

My results with just the Sandra bandwidth test:

Sync Mode
Int Buff'd iSSE2: 7487 MB/s
Float Buff'd iSSE2: 7213 MB/s

Non-sync mode
Int Buff'd iSSE2: 6887 MB/s
Float Buff'd iSSE2: 6964 MB/s

So for me sync mode is faster.

Revs · 12-08-07, 04:08 PM

wow Mullet that's a hell of a FSB clock

mythy · 12-08-07, 07:10 PM

680i's memory control would be my guess as to why

and mullet... drooping the multi like that is cheating!

12-07-07, 06:59 PM	#1
mythy Single Male Join Date: Dec 2004 Posts: 2,115	Synced or Unlinked? Results inside! So here you have it! I have benchmarked my Q6600 and DDR2-8500 ram at all available ratios/speeds and at 2 different core clocks! I have chosen 3 and 3.4GHz to test since they are the most common OC's around. I have used the following tests: Sandra Lite Memory bandwidth Sandra Lite Memory latency MediaCoder Benchmark First set of results are the most important since they compare the ram at a Synced state and a Unlinked! First you will see the bandwidth followed by latency and finally real world performance... Test are done in both 3 and 3.4GHz clock speeds. Sandra Lite Memory bandwidth* Sandra Lite Memory latency MediaCoder Benchmark These next tests show you all the different speeds and strap available to you and there performance results. Sandra Lite Memory bandwidth Sandra Lite Memory latency MediaCoder Benchmark Sandra Lite Memory bandwidth Sandra Lite Memory latency MediaCoder Benchmark So there you have it! Synced is slower and the lesser speed adds to its latency! While the performance gains arnt huge I should point out that every ns counts! __________________ ---Q6600 @3.45 430FSB - P5K Deluxe--- ---1066 5-4-4-12 PL6! - 257 GTX OCed--- ---ASUS Xonar DX - Dell 2407 24 LCD---

12-07-07, 08:24 PM	#2
CaptNKILL CUBE Join Date: Jan 2003 Location: PA, USA Posts: 18,844	Re: Synced or Unlinked? Results inside! Well I feel dumb... what is the "strap"? I don't think I've ever heard that term before. __________________ ---- Primary Rig ---- CoolerMaster 690 II Advance - Gigabyte GA-EP45-UD3P - Intel Core 2 Quad Q9550 @ 4.0Ghz + Thermalright Ultra 120 Extreme 6GB DDR2 @ 942Mhz 5-5-5-20 1.9v (2x1Gb Wintec AMPX PC2-8500 & 2x2Gb G.Skill PC2-6400) - EVGA Geforce GTX 470 @ 750/1500/1850 (1.050v) Sparkle Geforce GTS 250 1Gb Low-Profile (Physx) - Crucial RealSSD C300 64Gb SSD - Seagate 7200.12 500Gb SATA - Seagate 7200.10 320Gb SATA ASUS VW266H 25.5" LCD - OCZ GameXStream 700W PSU - ASUS Xonar DX - Logitech Z-5500 5.1 Surround - Windows 7 Professional x64 ---- HTPC ---- Asus M3A78-EM 780G - AMD Athlon X2 5050e 45W @ 2.6Ghz - 2x2GB Kingston PC2-6400 DDR2 - Sparkle 350W PSU Seagate 7200.10 320Gb SATA - Seagate 7200.10 250Gb SATA - Athenatech A100BB.350 MicroATX Desktop - Creative X-Fi XtremeMusic

12-07-07, 08:45 PM	#3
mythy Single Male Join Date: Dec 2004 Posts: 2,115	Re: Synced or Unlinked? Results inside! Its a internal dram ratio inside the northbridge. Not all board allow you to change it. But you can see how it effects the clocks __________________ ---Q6600 @3.45 430FSB - P5K Deluxe--- ---1066 5-4-4-12 PL6! - 257 GTX OCed--- ---ASUS Xonar DX - Dell 2407 24 LCD---

12-07-07, 08:47 PM	#4
$n][pErMan GTX 285 SSC Goodness! Join Date: Nov 2004 Location: Ohio Posts: 3,719	Re: Synced or Unlinked? Results inside! interesting... I wonder why the results are such... were the timings of the memory changed in the different strap modes or did you leave the timings alone? Generallly when running at a lower strap you can tighten the timings of the memory by quite a bit. __________________ Intel® Core 2 Quad Q6600 G0 @ 3.51GHz (390 x 9 / 1.39v) \| ASUS P5B Deluxe/WiFi-AP LGA 775 Intel P965 (1236 BIOS) \| OCZ GOLD 4GB (2GB x 2) DDR2 PC2-8500 Dual Channel (4-4-4-12 780Mhz/1.95V 1:1 Ratio) \| EVGA GeForce GTX 285 SSC Edition 1GB 512-bit GDDR3 (Currently @ 720MHz Core / 2750MHz Memory) \| Sound Blaster X-Fi Fatal1ty FPS \| Zalman CNPS9700 LED Ultra Quiet CPU Cooler \| 5 Western Digital 500GB SATA II 7200RPM w/32MB Buffer \| Acer X241W 24in Widescreen LCD 1920 x 1200 32bit \| Samsung DVD-RW-DL 18x Burner w/ lightscribe\| SILVERSTONE OP650 ATX12V / EPS12V 650W Power Supply \| AOpen 5.1 Surround sound system \| Windows Vista Business 64-Bit 3dMarkVantage: P16103 3dMark06: 17606 SuperPi 1M: 14s 100% Orthos Stable!

12-07-07, 08:52 PM	#5
mythy Single Male Join Date: Dec 2004 Posts: 2,115	Re: Synced or Unlinked? Results inside! I can run 4-4-4-12 to 533 but for the tests it was all at stock. and to be honest I noticed no performance increases at 4-4-4-12 The speed more then makes up from any gains you might see from the tighter timings.. I know that goes against the general hear say but hey what ever __________________ ---Q6600 @3.45 430FSB - P5K Deluxe--- ---1066 5-4-4-12 PL6! - 257 GTX OCed--- ---ASUS Xonar DX - Dell 2407 24 LCD---

12-08-07, 01:36 PM	#6
$n][pErMan GTX 285 SSC Goodness! Join Date: Nov 2004 Location: Ohio Posts: 3,719	Re: Synced or Unlinked? Results inside! Did you try any game benchmarks by chance? __________________ Intel® Core 2 Quad Q6600 G0 @ 3.51GHz (390 x 9 / 1.39v) \| ASUS P5B Deluxe/WiFi-AP LGA 775 Intel P965 (1236 BIOS) \| OCZ GOLD 4GB (2GB x 2) DDR2 PC2-8500 Dual Channel (4-4-4-12 780Mhz/1.95V 1:1 Ratio) \| EVGA GeForce GTX 285 SSC Edition 1GB 512-bit GDDR3 (Currently @ 720MHz Core / 2750MHz Memory) \| Sound Blaster X-Fi Fatal1ty FPS \| Zalman CNPS9700 LED Ultra Quiet CPU Cooler \| 5 Western Digital 500GB SATA II 7200RPM w/32MB Buffer \| Acer X241W 24in Widescreen LCD 1920 x 1200 32bit \| Samsung DVD-RW-DL 18x Burner w/ lightscribe\| SILVERSTONE OP650 ATX12V / EPS12V 650W Power Supply \| AOpen 5.1 Surround sound system \| Windows Vista Business 64-Bit 3dMarkVantage: P16103 3dMark06: 17606 SuperPi 1M: 14s 100% Orthos Stable!

12-08-07, 01:44 PM	#7
mythy Single Male Join Date: Dec 2004 Posts: 2,115	Re: Synced or Unlinked? Results inside! do you have any ideas how many reboots I had to do? I was at it for over a hour... Never the less I am planing a retest at 2.4GHz stock at 1:1 and max dividers for my 1066 ram. At 1:1 I will use 4-4-4-12 and at 1066 I will use 5-5-5-15. I will run threw the crysis benchmark for results __________________ ---Q6600 @3.45 430FSB - P5K Deluxe--- ---1066 5-4-4-12 PL6! - 257 GTX OCed--- ---ASUS Xonar DX - Dell 2407 24 LCD---

12-08-07, 03:24 PM	#8
Revs Registered User Join Date: May 2007 Location: UK Posts: 6,362	Re: Synced or Unlinked? Results inside! So you're saying i'm loosing bandwidth from having setup in 'Sync' mode. Hmm, I shall have to test this myself. I'll be back!

12-08-07, 04:03 PM	#9
mullet Join Date: Oct 2005 Posts: 7,998	Re: Synced or Unlinked? Results inside! 7x500 1:1 400 strap Sync Mode Int Buff'd iSSE2: 8517 MB/s Float Buff'd iSSE2: 8517 MB/s Memory latency 70ms __________________ • LIAN-LI PC-A70B • ASUS MAXIMUS V FORMULA (0701) • i7-3770k + NOCTUA\|NH-D14 • Intel 520 240GB SSD • G.SKILL Ripjaws X Series 16GB DDR3 1600 • • EVGA GeForce GTX 680 • PCP&C 750 Quad • ASUS 12x BD-ROM • DELL U2412M E-IPS • Windows 7 Pro x64 SP1 • Logitech Z5500 5.1 •

12-08-07, 04:04 PM	#10
Revs Registered User Join Date: May 2007 Location: UK Posts: 6,362	Re: Synced or Unlinked? Results inside! My results with just the Sandra bandwidth test: Sync Mode Int Buff'd iSSE2: 7487 MB/s Float Buff'd iSSE2: 7213 MB/s Non-sync mode Int Buff'd iSSE2: 6887 MB/s Float Buff'd iSSE2: 6964 MB/s So for me sync mode is faster.

12-08-07, 04:08 PM	#11
Revs Registered User Join Date: May 2007 Location: UK Posts: 6,362	Re: Synced or Unlinked? Results inside! wow Mullet that's a hell of a FSB clock

12-08-07, 07:10 PM	#12
mythy Single Male Join Date: Dec 2004 Posts: 2,115	Re: Synced or Unlinked? Results inside! 680i's memory control would be my guess as to why and mullet... drooping the multi like that is cheating! __________________ ---Q6600 @3.45 430FSB - P5K Deluxe--- ---1066 5-4-4-12 PL6! - 257 GTX OCed--- ---ASUS Xonar DX - Dell 2407 24 LCD---