All Memory is Not Created Equal - HyperX 2133 CL12 versus Ripjaws 2133 CL11

[Klem 641]

J95

"Stolichnaya...? I prefer Polish fist half, clocks and timings fully programmable."

You mistake. You dont understand, i and Khenglish speak about Samsung modules! Of course Corsair`s SPD all programable, but Samsung after 2011 year not. And not only Samsung, but and Hynix, Nanya and some others.

[J95 959]

Stolichnaya...? I prefer Polish fist half, clocks and timings fully programmable.

You mistake. You dont understand, i and Khenglish speak about Samsung modules! Of course Corsair`s SPD all programable, but Samsung after 2011 year not. And not only Samsung, but and Hynix, Nanya and some others.

Yeah, I was talking about Corsair/Samsung chips...

Mistakes:

"Samsung K4B2G0846D HYK0. It is the best chips in the world!" wrong, K4B4G0446Q-HCMA K4B2G0846D-HCMA.

"K4B4G0846Q HYK0. In laptops max for this chips 2133 MHz with timings 10-11-11-27" wrong again.

"2133 10-10-10-27 (all other timings is stock) 32GB (4x8GB) CR1, but it not stable" wrong, you need to adjust CR1 and set 2 DIMMs per channel with Thaiphoon burner.

"Of course Corsair`s SPD all programable, but Samsung after 2011 year not. And not only Samsung, but and Hynix, Nanya and some others" 2008 Samsung RAM modules were already write protected.

[J95 959]

Quote

What are all of your BIOS timings for the second screen shot? My system will not function with 10-10-10-27 @ 2133. Please post them and I will attempt it again. Are these manual timing settings or an XMP profile, as the XMP profile J95 provided to me works, but downclocks to 2000 and the performance is not good enough at the slower clock speed.

It could be the Haswell memory controller is why it works for you... may explain why it does not work with Ivy Bridge.

- - - Updated - - -

How is this done? I don't see BIOS settings for it. Is it something you change in the SPD with Thaiphoon Burner?

Here is what Thaiphoon Burner is reporting from my SPD:

Spoiler

Manufacturing Description
Module Manufacturer:	Corsair
SDRAM Manufacturer:	Undefined
Module Part Number:	CMSX16GX3M2B2133C1
Module Series:	Vengeance
Module Manufacturing Date:	Week 2D, 200D
Module Manufacturing Location:	Fremont, USA
Module Serial Number:	00000000h
Module PCB Revision:	00h
SDRAM Die Revision:	00h
Legitimate Architectures
Fundamental Memory Type:	DDR3 SDRAM
Module Speed Grade:	DDR3-2133
Module Type:	SO-DIMM (67.60 mm)
Module Capacity:	8192 MB
Module Nominal Height:	29 < H <= 30 mm
Module Thickness Maximum, Front:	2 < T <= 3 mm
Module Thickness Maximum, Back:	2 < T <= 3 mm
Reference Raw Card:	B
Reference Raw Card Revision:	0
Number of DIMM Ranks:	2
SDRAM Device Package:	Standard Monolithic
SDRAM Device Die Count:	Not specified
Signal Loading:	Not specified
Number of Column Addresses:	10 bits
Number of Row Addresses:	16 bits
Number of Bank Addresses:	3 bits (8 banks)
SDRAM Device Width:	8 bits
Programmed SDRAM Capacity:	4 Gb
Calculated SDRAM Capacity:	4 Gb
SDRAM Page Size:	1 KB
Primary Memory Bus Width:	64 bits
Memory Bus Width Extension:	0 bits
Supported Voltage Levels:	1.50 V/1.35 V
DRAM Timing Parameters
Fine Timebase Dividend:	1
Fine Timebase Divisor:	1
Fine Timebase:	0.0010 ns
Medium Timebase Dividend:	1
Medium Timebase Divisor:	8
Medium Timebase:	0.125 ns
Clock Cycle Time (tCK min):	0.938 ns
CAS# Latencies Supported (tCL):	6T, 7T, 8T, 9T, 10T,
	11T
CAS# Latency Time (tAA min):	9.375 ns
RAS# to CAS# Delay Time (tRCD min):	10.250 ns
Row Active to Row Active Delay (tRRD min):	6.500 ns
Row Precharge Delay Time (tRP min):	10.250 ns
Active to Precharge Delay Time (tRAS min):	29.000 ns
Act to Act/Refresh Delay Time (tRC min):	39.325 ns
Refresh Recovery Delay Time (tRFC min):	260.750 ns
Write Recovery Time (tWR min):	15.000 ns
Write to Read Command Delay (tWTR min):	7.625 ns
Read to Precharge Command Delay (tRTP min):	8.375 ns
Four Active Windows Delay (tFAW min):	30.875 ns
RZQ / 6 Drive Strength:	Supported
RZQ / 7 Drive Strength:	Supported
DLL-Off Mode Support:	Supported
Thermal Parameters
Extended Temperature Range:	0-95 °C
Extended Temperature Refresh Rate:	2X (85-95 °C)
Auto Self Refresh (depending on temperature):	Not supported
Module Thermal Sensor:	Not Incorporated
On-die Thermal Sensor Readout:	Not supported
Partial Array Self Refresh:	Supported
SPD Protocol
SPD Revision:	1.1
SPD Bytes Total:	256
SPD Bytes Used:	176
SPD Checksum:	37E3h (OK)
CRC covers bytes:	0-116
Part number details
JEDEC DIMM Label:	8GB 2R x8 PC3L-17000S-10-11-B0

Timing table
Frequency	tCL	tRCD	tRP	tRAS	tRC	tRFC	tRRD	tWR	tWTR	tRTP
1067 MHz	11	11	11	31	42	278	7	16	9	9
1067 MHz	10	11	11	31	42	278	7	16	9	9
933 MHz	9	10	10	28	37	244	7	15	8	8
800 MHz	8	9	9	24	32	209	6	12	7	7
667 MHz	7	7	7	20	27	174	5	10	6	6
533 MHz	6	6	6	16	21	140	4	8	5	5

Your scores are better than mine, still if you want to try XMP 2000MHz 9-10-10-27 (windows, Photoshop, Premeire Pro etc. better than 2133MHz 10-11-11-31) XMP Enhancer -> Templates -> XMP templates -> Import templates from file -> set XMP_2000MHz_9_10_10_27.emp - first copy XMP profile settings 'Custom profile' then set XMP 1 (1.5V) F10 You'll be able to switch from XMP to custom/default settings when benchmarking.

[Klem 641]

Mr. Fox

2133 10-10-10-27 (all other timings is stock), but it not stable.

And of course all memory controllers in CPU is different. Ability work with high frequency and agressive timings may be very different even for two identical CPUs.

[Khenglish 1154]

I do think that Haswell in general has a better IMC that IVB too. There's no way my CPU could run my memory at the rated 1866 10-10-10 at 1.35V, but I'm assuming that G.skill did adequately test the memory.

[Klem 641]

I do think that Haswell in general has a better IMC that IVB too. There's no way my CPU could run my memory at the rated 1866 10-10-10 at 1.35V, but I'm assuming that G.skill did adequately test the memory.

Whats chips use G.Skill in your memory moduls?

[Khenglish 1154]

Whats chips use G.Skill in your memory moduls?

I haven't checked yet. I assumed they were the same as the 2133 11-11-11 sticks, but maybe not.

[Klem 641]

I haven't checked yet. I assumed they were the same as the 2133 11-11-11 sticks, but maybe not.

Just check it. It is intrested, much things depended from mark of chips.

[Riri-Fifi 101]

Hi there,

Would you by any chance have the SPD of your Kingston 2133 ram ? I'm trying to make my Gskill RAM better flashing a Kingston SPD with Thaiphoon burner.

Thank you.

Marc

[J95 959]

My current XMP profiles, 100% stable.

[Klem 641]

Hi there,

Would you by any chance have the SPD of your Kingston 2133 ram ? I'm trying to make my Gskill RAM better flashing a Kingston SPD with Thaiphoon burner.

Thank you.

Marc

Whats chips use G.Skill in your memory moduls?

[Riri-Fifi 101]

I would like to test spd Kingston on Gskill modules

[Khenglish 1154]

@RiRi_Fifi So are you having troubling getting 2133 to run as well? This memory absolutely hates my channel 1 memory slots. If I use just channel 1 slots I can run 2133. Even just 1 stick in a channel 1 slot is no good, so it's not dual channel, but just channel 1.

[Klem 641]

I would like to test spd Kingston on Gskill modules

[ATTACH=CONFIG]11076[/ATTACH]

In this photo we can see only memory moduls, no memory chips. Lift one corner of the stickers and make a photo again. To be able to see at least one memory chip.

[Riri-Fifi 101]

Run 3dmark 13 Memory 1600mhz

NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-4930K,Notebook P570WM

[unityole 42]

Your scores are better than mine, still if you want to try XMP 2000MHz 9-10-10-27 (windows, Photoshop, Premeire Pro etc. better than 2133MHz 10-11-11-31) XMP Enhancer -> Templates -> XMP templates -> Import templates from file -> set XMP_2000MHz_9_10_10_27.emp - first copy XMP profile settings 'Custom profile' then set XMP 1 (1.5V) F10

You'll be able to switch from XMP to custom/default settings when benchmarking.

why is 2000mhz 9-10-10 better for windows photoshop etc than 2133mhz 10-11-11? is it on CAS alone?

[J95 959]

why is 2000mhz 9-10-10 better for windows photoshop etc than 2133mhz 10-11-11? is it on CAS alone?

Actually 9-10-9-27 and 9-10-9-28 lower latency 133MHz difference, for benchmarking set 2133MHz 10-11-11-31 (30) as default (Mr. Fox) http://forum.techinferno.com/general-notebook-discussions/6063-all-memory-not-created-equal-hyperx-2133-cl12-versus-ripjaws-2133-cl11-5.html#post85089 4x8GB XMP profiles/Thaiphoon Burner you need to adjust CR1 (1 ns) to 1.250 ns / 1.500 ns (CR2 2.000 ns) and set 2 DIMMs per channel.

[Riri-Fifi 101]

hello

test winsat mem

[Khenglish 1154]

Whats chips use G.Skill in your memory moduls?

The G.Skill 1866 10-10-10 8GB 1.35V sticks use these chips:

G.Skill

D3 512M8GE9

1347 G7280E

However 1 stick just completed its 5th pass in TestMem5 at 10-11-10-27 2133...

All that matters is the DIMM channel. Channel 0 can run this in either slot. The same stick I'm using now cannot boot windows at 2133 10-11-10-27 using a channel 1 slot. Gonna look through my mobo datasheet to see any reason why the channel could be inferior. The 2GB samsung sticks I have can bench at 9-10-10 2133 in dual channel, so this is weird.

[tmash 30]

Try to use XMP Profile 2...

[Riri-Fifi 101]

hello

[Klem 641]

Has anyone tried to flash SPD chip with Taiphoon Burner on the Haswell platforms with 8 series of chipsets?

[Casowen 13]

Where you able to determine why the g skills where faster? Im thinking of getting them, or the kingston hyperx sodimms. I dont over clock, and neither does my dell latitude, so im really wondering you input.

[mw86 1362]

Hi there,

Would you by any chance have the SPD of your Kingston 2133 ram ? I'm trying to make my Gskill RAM better flashing a Kingston SPD with Thaiphoon burner.

Thank you.

Marc

Hello sorry for late reply, which ram stick did you need spd from. If you still are tinkering on your ram I would be glad to help.

[mw86 1362]

Has anyone tried to flash SPD chip with Taiphoon Burner on the Haswell platforms with 8 series of chipsets?

I was flashing on Alien 18 at one point, not sure which chipset that was.