You are not registered yet. Please click here to register!
Super Talent Chrome DDR3-2000 / 2200 CAS8...? - Page 6 - i4memory.com - different look at memory
Tags Register Blogs FAQ Members List Calendar Mark Forums Read

Notices

Reply
 
Thread Tools
Old 30-05-2010, 10:29 AM   #41
spoolindsm127
Senior Member
spoolindsm127's PC Specs
 
Join Date: May 27 2010
Posts: 170
By GenAn View Post
It'll be cheaper to get the 2200 sticks then
Exactly, $248 after Bing Cashback . That's what I'll be settling for too.
spoolindsm127 is offline   Reply With Quote
Old 30-05-2010, 03:13 PM   #42
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,021
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Started testing 1st 2x2GB SuperTalent 2200C8 kit and definitely Elpida Hyper. With 2x uncore to mem ratio, easily got to 2000mhz 7-7-7-20 1T at 1.64v came easy for memtest86+ v4.10 at least. Capping CPU VTT at 1.45v and 1.66v vdimm managed 2076Mhz 8-8-8-24 in memtest86 on i7 980X with P6T6 WS Revolution. Probably need > 1.55v CPU VTT for 2200mhz or close that frequency on X58, probably easier to achieve on P55 platform.

Really all I am interested in is 7-7-7-20 and 6-6-6-x clocks, but lets see how all 3 kits of 2x2GB do
eva2000 is offline   Reply With Quote
Old 30-05-2010, 03:31 PM   #43
spoolindsm127
Senior Member
spoolindsm127's PC Specs
 
Join Date: May 27 2010
Posts: 170
By eva2000 View Post
Started testing 1st 2x2GB SuperTalent 2200C8 kit and definitely Elpida Hyper. With 2x uncore to mem ratio, easily got to 2000mhz 7-7-7-20 1T at 1.64v came easy for memtest86+ v4.10 at least. Capping CPU VTT at 1.45v and 1.66v vdimm managed 2076Mhz 8-8-8-24 in memtest86 on i7 980X with P6T6 WS Revolution. Probably need > 1.55v CPU VTT for 2200mhz or close that frequency on X58, probably easier to achieve on P55 platform.

Really all I am interested in is 7-7-7-20 and 6-6-6-x clocks, but lets see how all 3 kits of 2x2GB do
Great news Eva2000

I ordered three individual WS220UB2G8 2200Hz CL8 modules so lets hope they all behave identically.

You think you can compare and see if these behave the same as your Kingston 2000C8 at 2200MHz and 7-7-7-20 timings? I'm sure since they use the same IC they should but some concrete evidence would be great.
spoolindsm127 is offline   Reply With Quote
Old 30-05-2010, 10:30 PM   #44
kayl
Senior Member
kayl's PC Specs
 
Join Date: Jun 22 2005
Location: Port Hedland
Posts: 1,750
By eva2000 View Post
Started testing 1st 2x2GB SuperTalent 2200C8 kit and definitely Elpida Hyper. With 2x uncore to mem ratio, easily got to 2000mhz 7-7-7-20 1T at 1.64v came easy for memtest86+ v4.10 at least. Capping CPU VTT at 1.45v and 1.66v vdimm managed 2076Mhz 8-8-8-24 in memtest86 on i7 980X with P6T6 WS Revolution. Probably need > 1.55v CPU VTT for 2200mhz or close that frequency on X58, probably easier to achieve on P55 platform.

Really all I am interested in is 7-7-7-20 and 6-6-6-x clocks, but lets see how all 3 kits of 2x2GB do
that is some impressive results there eva2000.
Postie tried to deliver mine the other day, cant wait to pick them up from the post office
kayl is offline   Reply With Quote
Old 31-05-2010, 02:09 AM   #45
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,021
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
By spoolindsm127 View Post
Great news Eva2000

I ordered three individual WS220UB2G8 2200Hz CL8 modules so lets hope they all behave identically.

You think you can compare and see if these behave the same as your Kingston 2000C8 at 2200MHz and 7-7-7-20 timings? I'm sure since they use the same IC they should but some concrete evidence would be great.
Just a tad behind Kingston 2000C8 kit for 1st 2x2GB 2200C8 set testing i.e. max 6-6-6-20 at same timings/subtimings at 1.7v
  • super talent = 894mhz memtest86+ v4.10 can't go higher even at 1.72-1.74v but haven't tried different subtimings only using same subtimings to get comparison with kingston 2000c8
  • kingston = 910mhz memtest86+ v4.10 at 1.7v and 940mhz at 1.72-1.74v memtest and 925mhz at 1.72v 32M Pi pass.

but super talent is scaling more easily at 6-7-6-x than kingston i.e. 900mhz 6-7-6-20 at 1.66v memtest okay, still early tests for memtest to get a comparative feel. But remember this is 3x2GB Kingston 2000C8 vs 2x2GB Super Talent 2200C8 memtesting.

By kayl View Post
that is some impressive results there eva2000.
Postie tried to deliver mine the other day, cant wait to pick them up from the post office
don't ya hate that heh
eva2000 is offline   Reply With Quote
Old 31-05-2010, 04:05 AM   #46
spoolindsm127
Senior Member
spoolindsm127's PC Specs
 
Join Date: May 27 2010
Posts: 170
Hey Eva2000,

Thanks for the information that's very useful data. So far I am very impressed and relieved to have bought the Super Talent 2200MHz.

Ahh thanks for reminding me about the triple vs dual difference. So if the Super Talents were run in same triple config as Kingston 2000C8 can we expect basically identical results? I don't know if there is a difference in modules when they are in a "dual" kit or " triple" kit, or even when sold individually.

I guess I will have to wait and see. Which brings me to my next point; motherboard. I am looking at buying the AsRock X58 Extreme3. But it is a "Advanced V8 Power Phase Design" so I'm assuming that is 8+2 Power Phase. How important do you think this is when overclocking? I plan to OC a i7 920 and try to stay within specified voltage ie 1.35V
spoolindsm127 is offline   Reply With Quote
Old 31-05-2010, 04:21 AM   #47
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,021
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Triple channel more stressful than dual channel meaning it's easier to run dual channel than triple channel. Triple channel most likely need a bit more volts for same config whether it be more CPU VTT and/or VDIMM.

Just started testing 2nd Super Talent 2x2GB 2200C8 kit to save time starting at 6-6-6-20 clocks and pretty similar to first set bit better with 900mhz 6-6-6-20 at 1.7v versus 1st kit's 894mhz 6-6-6-20 at 1.7v. Like first set once it hits a wall for mem frequency at 894-900mhz 6-6-6-20, loosening tRCD to 7 for 6-7-6-x allows it to scale bit more. 2nd set is bit better in that regard too 925mhz 6-7-6-20 at 1.7v vs 1.72v

Phases help for oc'ing especially under full 100% cpu loads i.e. prime, wprime etc for normal benchmarks that don't stress all cores probably not as important. But Asrock is an unknown as i haven't used one before, if you need help etc from folks best to stick with commonly used brand/models from Asus, DFI or Gigabyte
eva2000 is offline   Reply With Quote
Old 31-05-2010, 07:47 AM   #48
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,021
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Okay done preliminary memtest86+ v4.10 comparisons only with one full test #1-8 pass + 10 loops of test #5 looping. As stated here, memtest86+ v4.10 doesn't mean same windows stability, for that you need to test in windows.

3x 4GB (2x2GB) Super Talent 2200 8-8-8-24 Dual channel kits labeled each module
  • Kit #1 = module #1 + #2
  • Kit #2 = module #3 + #4
  • Kit #3 = module #5 + #6

Kit #1
  • DDR3-1600Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.375v CPU VTT and 1.66v
  • DDR3-1758Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.3875v CPU VTT and 1.66v
  • DDR3-1788Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.4375v CPU VTT and 1.70v
  • DDR3-1800Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.45v CPU VTT and 1.66v
  • DDR3-1820Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.45v CPU VTT and 1.66v
  • DDR3-1828Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.68v
  • DDR3-1850Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.72v (1.70v had 2 errors)
  • DDR3-1870Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.72v
  • DDR3-1880Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.74v
  • DDR3-2040Mhz 7-7-7-20 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.64v
  • DDR3-2076Mhz 8-8-8-24 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.66v
* Didn't try any higher for 7-7-7-20/8-8-8-24
* DDR3-1800 6-6-6-20 couldn't pass even with 1.475V CPU VTT



Kit #2
  • DDR3-1600Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.375v CPU VTT and 1.66v
  • DDR3-1800Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.70v
  • DDR3-1800Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.45v CPU VTT and 1.66v
  • DDR3-1850Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.70v
  • DDR3-1880Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.72v
  • DDR3-1900Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.74v
  • DDR3-2040Mhz 7-7-7-20 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.64v
  • DDR3-2086Mhz 8-8-8-24 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.66v
  • DDR3-2110Mhz 8-8-8-24 tRFC 110 tFAW 31 B2B 4 with 1.475v CPU VTT and 1.66v

Kit #3
  • DDR3-1600Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.375v CPU VTT and 1.66v
  • DDR3-1788Mhz 6-6-6-20 tRFC 88 tFAW AUTO with 1.4375v CPU VTT and 1.70v
  • DDR3-1800Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.45v CPU VTT and 1.66v
  • DDR3-1850Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.72v
  • DDR3-1870Mhz 6-7-6-20 tRFC 88 tFAW AUTO with 1.475v CPU VTT and 1.74v
  • DDR3-2040Mhz 7-7-7-20 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.64v
  • DDR3-2076Mhz 8-8-8-24 tRFC 110 tFAW 31 B2B 4 with 1.45v CPU VTT and 1.66v
  • DDR3-2100Mhz 8-8-8-24 tRFC 110 tFAW 31 B2B 4 with 1.475v CPU VTT and 1.66v

How to compare results, basically came down to max 6-6-6-20, 6-7-6-20. and 8-8-8-24 timing memory frequencies at same voltages and subtimings to differentiate between the 3 kits.
  • Only Kit #2 managed DDR3-1800Mhz 6-6-6-20 at 1.7v, while Kit #1 & #3 fell just short at DDR3-1788Mhz 6-6-6-20 at 1.7v.
  • Max 6-7-6-20 timings frequency also saw Kit #2 lead at DDR3-1900Mhz 6-7-6-20 at 1.74v, then Kit #1 at DDR3-1880mhz 6-7-6-20 at 1.74v then last Kit #3 at DDR3-1870Mhz 6-7-6-20 at 1.74v
  • Kit #2 was also stronger on max 8-8-8-24 timings with DDR3-2086Mhz 8-8-8-24 at 1.66v capped at CPU VTT 1.45v (as you see bump CPU VTT volts theres more headroom at DDR3-2110Mhz).
  • Kit #1 + #3 tied at capped 1.45v CPU VTT at DDR3-2076Mhz 8-8-8-24 but push CPU VTT to 1.475v and Kit #2 beat out #3 kit DDR3-2110Mhz vs DDR3-2100Mhz 8-8-8-24 at 1.66v.

Final part of memtest is to find a module out of Kit #1 and #3, which best matches Kit #2 performance for 3x2GB custom binned kit

How's it compare to Kingston 2000C8 kit at http://i4memory.com/f83/khx2000c8d3t...i7-980x-23731/ ? I'd say at each timings Kingston 2000C8 kit has around 20-25mhz (DDR3 equivalent of 40-50mhz i.e. DDR3-1850Mhz vs DDR3-1800Mhz) extra headroom over the Super Talent 2200C8 so far.

Last edited by eva2000; 31-05-2010 at 01:39 PM.
eva2000 is offline   Reply With Quote
Reply

Bookmarks

Tags
ddr3, elpida 50nm hyper, elpida mgh-e hyper, elpida mnh-e hyper, p55 dual channel, super talent, super talent wb200ux6g8, super talent ws220ub2g8, super talent ws220ux4g8, x58 triple channel


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools


Similar Threads for: Super Talent Chrome DDR3-2000 / 2200 CAS8...?
Thread Thread Starter Forum Replies Last Post
[Techpowerup.com] (PR) Super Talent Announces 2000 MHz Dual and Triple Channel DDR3 Kits News Reviews & News Online 0 08-07-2010 08:22 PM
[Techpowerup.com] (PR) Super Talent Rolls Out Dual Channel DDR3 Kits for P55 Core i5 and Core i7 Systems News Reviews & News Online 0 17-10-2009 03:41 AM
[Techpowerup.com] Super Talent Releases New DDR3 Apple Certified Memory Modules News Reviews & News Online 0 05-05-2009 11:30 PM
[Techpowerup.com] (PR) Super Talent Launches DDR3 RDIMMs and ECC UDIMMs for Next Generation Servers News Reviews & News Online 0 27-02-2009 03:40 AM
[Techpowerup.com] (PR) Super Talent Launches DDR3 SO-DIMM Modules for Next Generation News Reviews & News Online 0 28-05-2008 10:01 AM


All times are GMT +11. The time now is 03:37 AM.

no new posts