OCZ Platinum DDR3-1333 OCZ3P1333LVA6GK Animal Series @2000-2170Mhz!

[eva2000]

OCZ Platinum DDR3-1333Mhz 7-7-7-24 (OCZ3P1333LVA6GK 'Animal Series') memory modules are Micron DDR3-1866-2000Mhz CAS9 binned ICs sprinkled into the regular OCZ Platinum DDR3-1333Mhz series and are differentiated by the 'A' in the product code.

• OCZ3P1333LVA6GK = Animal Series with Micron DDR3-1866-2000Mhz CAS9 binned ICs
• OCZ3P1333LV6GK = Regular OCZ Platinum DDR3-1333Mhz.

Let's see how these modules do on Asus P6T6 WS Revolution.

System:

• i7 920 C0 3836A756 #2
• EK Supreme LT 1366 / FESER 480 / MCP655
• Asus P6T6 WS Revolution
• OCZ Platinum DDR3-1333Mhz 7-7-7-24 (OCZ3P1333LVA6GK 'Animal Series')
• 512MB Inno3D 8800GT OC Edition @742/1836/2000
• 750GB Samsung SATA
• Pioneer DVR-217 DVD-RW
• Corsair HX1000 PSU
• WinXP Pro SP3 (nLite fully updated) with Vista Theme

Back to Back CAS# Delay was left on AUTO which defaults to value of 0. The default Command Rate (DRAM Timing Mode) when set to AUTO seems to change to 2T at high memory frequency CAS9 memory clocks. So I decided to run DDR3-2000Mhz 9-9-9-24 2T first and then manually set to 1T DRAM Timing Mode for a 2nd set of benchmarks as follows:

• DDR3-2000Mhz 9-9-9-24 2T at 1.66v vdimm, 1.50v QPI/DRAM (CPU VTT uncore voltage)
• DDR3-2000Mhz 9-9-9-24 1T at 1.66v vdimm, 1.50-1.51875v QPI/DRAM (CPU VTT uncore voltage)

Super Pi 1M, 32M, Everest Ultimate bandwidth tests for 2T managed with QPI/DRAM of 1.50v only 1T config's Hyper Pi 8x16M Pi test required bumping QPI/DRAM from 1.50v to 1.51875v

click on image for full screenshots

DDR3-2000Mhz 9-9-9-24 2T at 1.66v vdimm











DDR3-2000Mhz 9-9-9-24 1T at 1.66v vdimm

[eva2000]

Continue testing of 6GB OCZ Platinum DDR3-1333Mhz 7-7-7-24 (OCZ3P1333LVA6GK 'Animal Series') kit but with voltage bumps for single Super Pi 32M stability.

For DDR3-2058Mhz 9-9-9-24 1T I tested different Back to Back CAS# Delay values, AUTO (0) vs 6 vs 10. B2B value of 8 for some reason resulted in memtest86+ v2.11 test #2 errors while the other values of AUTO (0), 6 and 10 didn't have such errors. 32M Pi difference between B2B AUTO (0) and 10 was only 3.063 seconds slower at 10 but a 17.6% and 9.96% drop in Everest Read and COPY bandwidth respectively. For DDR3-2108Mhz 9-9-9-24 1T at 1.74v I only tested B2B value of AUTO (0).

click on image for full screenshots

DDR3-2058Mhz 9-9-9-24 1T at 1.68v vdimm

B2B = AUTO (0)







B2B = 10



Everest Ultimate - B2B AUTO(0) vs 6 vs 10
17.6% and 9.96% drop in Everest Read and COPY bandwidth respectively.

[eva2000]

DDR3-2108Mhz 9-9-9-24 1T at 1.74v vdimm










DDR3-2170Mhz 9-9-9-24 1T at 1.74v vdimm
Max CPUZ Validation was done with same settings as for above DDR3-2108Mhz benchmarks. Yes that's right 4343Mhz validation at 1.408v vcore HT Enabled not bad for old C0 stepping

[eva2000]

They're also capable of DDR3-1600-1760mhz 7-7-7-24 at 1.60-1.66v on P6T6 WS Revolution http://i4memory.com/117786-post4.html

[clon22]

Oh man, these are nice sticks!
Thanks for sharing George.
Too bad they can't be found in my local market.

[eva2000]

Yeah not available in Australia either, luckily www.priceguidenetwork.com in USA ships internationally and they stock these animals for US$125. I ordered alot of crucial ballistix d9gmh 8500 from PGN in the past so know the folks there very well

[clon22]

Thanks for link.

[KILLER_K]

Good review and thanks for the test and links. Always enjoy reading you reviews, keep up the good work.

[eva2000]

thanks glad you like

direct link at PGN http://www.priceguidenetwork.com/ocz...333lva6gk.html US$125.

[nut311]

nice eva did you try cas 9-9-8-20 TRFC 68-74

[over@locker886]

Nice review George!

I'm sure this chips Micron D9KPT These have best performance/value ratio.

I will back with my Kingmax^^

[xpatar]

Nice review And for the price they really kick

[saaya]

have you tried higher vdimm?

[eva2000]

nice eva did you try cas 9-9-8-20 TRFC 68-74

not yet still very early tests as the P6T6 WS Revolution was only just setup

Nice review George!

I'm sure this chips Micron D9KPT These have best performance/value ratio.

I will back with my Kingmax^^

yeah probably same ICs - can probably squeeze out DDR3-1600Mhz 6-6-6-22 1T if i'm lucky

Nice review And for the price they really kick

yeah price is cheap!

have you tried higher vdimm?

not yet still getting a feel for the new motherboard - flashed to a new bios for testing

[over@locker886]

yeah probably same ICs - can probably squeeze out DDR3-1600Mhz 6-6-6-22 1T if i'm lucky

Two of them hit 2200Mhz CL8 with 1,9v for validation (in Rampage Extreme). Third stick worse, max 2030Mhz Need more find..

[eva2000]

Two of them hit 2200Mhz CL8 with 1,9v for validation (in Rampage Extreme). Third stick worse, max 2030Mhz Need more find..

Darn, maybe musical dimm slots for each module to see if the poorer module does better in different dimm slot ?

[Shiranui Gen-An]

Two of them hit 2200Mhz CL8 with 1,9v for validation (in Rampage Extreme). Third stick worse, max 2030Mhz Need more find..

Good deal, I wondered how those sticks would do on a 775 board that can throw VDIMM at them without having to worry about the CPU.

[denice25]

thanks for the very nice share!! keep it up!!

[zsamz_]

i seen many value kits with D9KPT even some crucial pc3-8500

had a chance to get a set for 30$ maybe i shoulda lol

[eva2000]

Might try these sticks in REX X48 board later on

thanks for the very nice share!! keep it up!!

You're welcome thanks

i seen many value kits with D9KPT even some crucial pc3-8500

had a chance to get a set for 30$ maybe i shoulda lol

cheap but unknown performance i guess hehe.

Quick 32M Pi bench pushing these OCZ Platinum 1333 Animals to DDR3-1702Mhz 7-7-7-24 1T at:

• 1.66v vdimm, 1.40v CPU VTT (QPI/DRAM) for default 3404mhz Uncore frequency
• 1.74v vdimm, 1.60v CPU VTT (QPI/DRAM) for 20x multi Uncore frequency of 4255mhz

As most folks know as you overlock the Uncore (NB) frequency you'll need more vdimm voltage to get stable. But as with DFI UT X58 uncore frequency tests, Asus P6T6 WS Revolution needs more CPU VTT uncore (QPI/DRAM) voltage as well. You can see more Uncore frequency comparisons here. Asus P6T6 WS Revolution certainly needs alot more CPU VTT uncore volts to reach similar bclk/uncore/mem frequencies than on DFI UT X58-T3EH8 though!

click image for full screenshot











System:

• i7 920 C0 3836A756 #2
• EK Supreme LT 1366 / FESER 480 / MCP655
• Asus P6T6 WS Revolution
• OCZ Platinum DDR3-1333Mhz 7-7-7-24 (OCZ3P1333LVA6GK 'Animal Series')
• 512MB Inno3D 8800GT OC Edition @742/1836/2000
• 750GB Samsung SATA
• Pioneer DVR-217 DVD-RW
• Corsair HX1000 PSU
• WinXP Pro SP3 (nLite fully updated) with Vista Theme

[saaya]

Asus P6T6 WS Revolution certainly needs alot more CPU VTT uncore volts to reach similar bclk/uncore/mem frequencies than on DFI UT X58-T3EH8 though!

where and how do you meassure vtt and vdimm?
you do know that the dfi board meassures vdimm in the cpu socket while the asus and most other boards meassure it close to the vtt pwm right?

[over@locker886]

Darn, maybe musical dimm slots for each module to see if the poorer module does better in different dimm slot ?

Best dimm slot is 4th on R.E. , so tested all in this.

1980Mhz CL9 everest result in DFI DK X58-T3eH6:

[radaja]

thanks for the link eva2000,just ordered a kit.
now ive got a good kit for when i go i7,and will try them
out in my REX for now.

[eva2000]

where and how do you meassure vtt and vdimm?
you do know that the dfi board meassures vdimm in the cpu socket while the asus and most other boards meassure it close to the vtt pwm right?

you learn something new everyday if true on DFI UT X58-T3H8 has measure pts for DMM measurements but Asus just going by what I set in bios - whether real is undervolted or overvolted it's still seems like a lot to me hehe

Best dimm slot is 4th on R.E. , so tested all in this.

1980Mhz CL9 everest result in DFI DK X58-T3eH6:

guess you need more kits to play with

thanks for the link eva2000,just ordered a kit.
now ive got a good kit for when i go i7,and will try them
out in my REX for now.

nice, let us know how they fair on REX

[saaya]

you learn something new everyday if true on DFI UT X58-T3H8 has measure pts for DMM measurements but Asus just going by what I set in bios - whether real is undervolted or overvolted it's still seems like a lot to me hehe

that was a typo, i meant vtt, not vdimm in the first sentence
and im sure the dfi board uses less vtt and vdimm for the same results, after all oskar spends quite some time finetuning memory and hes pretty damn good at it

but i dont think the diference is that big... the asus probably just reads voltages in a bad way while the dfi is very accurate at voltage reading, thats another thing oskar focusses on afaik.