G.SKILL F2-9600CL5D-4GBPI Sample testing

[RobinL]

I am not, i have changed my order to mushkins 4GB 8500 Black ascent.

[Roger_D25]

bigslappy - Did you get your kit from newegg after reading my post or did you find them available somewhere else? I just recently made the switch back to DDR2 from DDR3 so I'm glad to have been able to get this kit so soon after making the switch (especialy since it seems like this kit has been very hard to comeby)!

RobinL - Just curious, why did you change your order and get the Mushkins instead (personal choice, finacial reasons, etc..)? Granted this is an awesome 4GB kit of DDR2 (probably one of the best 4GB kits) but given there is much cheaper RAM with somewhat similar performance it can be hard to justify the cost. However I was happy to pay $299 for a kit and I'm looking forward to testing them soon!

[Jsunn]

At new egg now!

http://www.newegg.com/Product/Produc...82E16820231201

[Roger_D25]

Yeah I posted them last night but their sold out again now, hopefully a few of you guys got a kit while they were available!

[-n7-]

Sorry if this was already posted, Indonesian review with spreaders off, says it's Powerchips:http://www.ur-hardware.com/Memory/G.Skill-4GB-PI.html

Those have gotta be the worst pictures of ICs ever.

Anyway, in spite of the horrible focus & distance of the pictures, one can still see the layout of the markings on the IC, & they aren't PSC markings, they are ProMOS ones.

See my previous post: http://www.xtremesystems.org/forums/...&postcount=145

Top three pictures in that post are PSC.

Notice how the dot is perfectly centered to match the bottom line of text, & notice the font style/size.

Bottom three are ProMOS.

Notice how the dot is slightly lower than the text, & the font is smaller & different than the PSC's.

Then compare to the Indonesian Hardware reviews pics...
It'd be nice to have a decent pic, but i can already tell it's ProMOS, even from the blurry pics.

[WaterFlex]

$300...too much, but It is worth it.

[bigslappy]

[QUOTE=Roger_D25;3410724]bigslappy - Did you get your kit from newegg after reading my post or did you find them available somewhere else? I just recently made the switch back to DDR2 from DDR3 so I'm glad to have been able to get this kit so soon after making the switch (especialy since it seems like this kit has been very hard to comeby)!

Yeah I got an auto notifiy notice at tyhe same time u posted & i got a pair & they would not sell 2 pairs get them monday OR tuesday
G.Skill only NewEgg would get them 1St ....YAHOO FTW

[Roger_D25]

Thanks bigslappy, much appreciated! Unfortunately I haven't had enough time to actually push them or do much testing but they did boot right up at 1200Mhz @ 5-5-5-15 on 2.02v (Rampage Formula overvolts a bit on the vDIMM). I ran a few regular benchies just to test for basic stability (SuperPI 32M, Windows MemTest, Primed for 1.5hrs on blend-test, and gamed last night for 2.5hrs with zero problems), this stuff is amazing to say the least! I'm very happy to have made the investment and I assume it will have some clocking headroom with both clockspeeds and latencies! By the way I spent a few minutes right when I first opened the package trying to pull the heat spreader (to verify IC's) but wimped out after a few minutes!

UPDATE: I've been doing some various testing tonight with this new RAM trying to get decent Read/Write/Latancy numbers while still being somewhat stable and I just got the best I've been able to do so far. Keep in mind this isn't impressive compared to what most of you guys do but for me this is great! This RAM is very impressive I think and I've loving this new Rampage Formula of mine! Besides the high clockspeed run below I did alot of testing at lower clocks and tighter latencies but so far this yeilded the best numbers. Plus as I'm sure most of you guys know already the most important setting so far seems to be "Common Performance Level" or "Performance Level", it easily has the biggest impact on memory performance. If I was watercooling my NB I'd would have tried lowering the PL from 7 (which is what I used for below) to 6 but to get any stability I needed to use more NB voltage than I felt comfortable with.



Thumbnail below has higher resolution than image above.

[LuckySevn]

http://www.newegg.com/Product/Produc...82E16820231194

they where a little higher a few days ago.

they're cheaper then my GEIL evo one pair & a bit better... :/ (screwed again >_<)

EVO ONE 4x2GB DDR2-1060 @ 2.00v

PowerChips. No matter how high they're binned they only take as high as 2.1v, any higher they refuse to boot. The highest binned PowerChips only need 1.9v to run DDR2-1066 CL5. I've tried 5 different batches of PowerChips none of them take higher than 2.1v...unless PSC (Li-Jing) secretly released a new chip...

Anyhow, those PI's are crazy, the high clocks can allow me to push the Wolfdales further! but $299.99 is too steep...

[Jor3lBR]

Anyhow, those PI's are crazy, the high clocks can allow me to push the Wolfdales further! but $299.99 is too steep...

Yeap you're right you have o decide, CoreI7 or G.Skills9600 Pi

[LuckySevn]

Thanks bigslappy, much appreciated! Unfortunately I haven't had enough time to actually push them or do much testing but they did boot right up at 1200Mhz @ 5-5-5-15 on 2.02v (Rampage Formula overvolts a bit on the vDIMM)....

My EVO ONE's boot DDR2-1200 CL5 @ 2.00v too but can't get in Vista no matter what. I was told by some folks on Coolaler that I should use XP 'cuz Vista addresses the memory way different than XP. Haven't tried XP with these so not sure. I'd like to see if those PI's can run Vista x64 100% stable @ DDR2-1200!!! Anyone here can share that particular experience? Thank you thank you~~

Yeap you're right you have o decide, CoreI7 or G.Skills9600 Pi...

I know man...I wanna get an i7 and a Deneb when it launches....but not having crazy arse memory too push them is a pain....I can't even sleep on it... T_T

[LuckySevn]

And here's my EVO ONE's:


I'd like to see clearly what's under the hood of these Pi's~~!!

[Roger_D25]

LuckySevn - Those GEIL Evo One modules are very decent and I almost got them instead of these GSkill 4GBI's (plus their much cheaper)! $300 is indeed very expensive for a 4GB kit of DDR2 but after having them for a week or so I'm very happy I bought them and will definately keep them for a while (at least until I make the switch to i7).

In regards to the IC's used on these modules I'm not 100% sure because the heatspreader is difficult to remove (much harder than my last kit of GSKill RAM that had the clips instead of the adhesive thermal pad used now). However I was able to carefully pull-up one side of one of the modules and they look just like this I think (but its hard to say without fully removing the heatspreader)

If I am able to safely remove the heatspreader I will take a few high-resolution photos and post them ASAP! I can also say that this kit scales pretty well with increased voltage (just like Denny's results show at the beginning of this thead). I didn't go past 2.10v in BIOS (which overvolts to about 2.18v-2.21v) but I feel confident that it would have continued scalling for a bit.

You had also asked about Vista and if these modules run on it, well my kit is running very strong on both WinXP SP2 and Vista 64-bit SP1. I haven't had a single issues with either OS and I've done lots of benchmaking and gaming on both OS's. I'm surprised your not having any luck with Vista using those Evo One modules but its clear from your picture above that the 4GBPI's are using different IC's (at least from what I saw and what I think they are). I'm even more interested in pulling this darn heatspreader off but like everyone else I also don't want to kill my very expensive RAM! Let me know if you want any specific Vista/XP results or tests and I'll do the best I can.

[-n7-]

And here's my EVO ONE's:


I'd like to see clearly what's under the hood of these Pi's~~!!

Thanx for the pic confirming your kit is PSC; i'll add it to my collection

We figured out the ICs of these G.Skills a few pages back
See Miravo confirm they are ProMOS here: http://www.xtremesystems.org/forums/...&postcount=146
(This was the post containing the review link: http://www.xtremesystems.org/forums/...&postcount=140)

Another review linked here: http://www.xtremesystems.org/forums/...&postcount=180
In spite of the bad pics, also shows they are ProMOS.

[LuckySevn]

LuckySevn - Those GEIL Evo One modules are very decent and I almost got them instead of these GSkill 4GBI's (plus their much cheaper)! $300 is indeed very expensive for a 4GB kit of DDR2 but after having them for a week or so I'm very happy I bought them and will definately keep them for a while (at least until I make the switch to i7).

In regards to the IC's used on these modules I'm not 100% sure because the heatspreader is difficult to remove (much harder than my last kit of GSKill RAM that had the clips instead of the adhesive thermal pad used now). However I was able to carefully pull-up one side of one of the modules and they look just like this I think (but its hard to say without fully removing the heatspreader)

If I am able to safely remove the heatspreader I will take a few high-resolution photos and post them ASAP! I can also say that this kit scales pretty well with increased voltage (just like Denny's results show at the beginning of this thead). I didn't go past 2.10v in BIOS (which overvolts to about 2.18v-2.21v) but I feel confident that it would have continued scalling for a bit.

You had also asked about Vista and if these modules run on it, well my kit is running very strong on both WinXP SP2 and Vista 64-bit SP1. I haven't had a single issues with either OS and I've done lots of benchmaking and gaming on both OS's. I'm surprised your not having any luck with Vista using those Evo One modules but its clear from your picture above that the 4GBPI's are using different IC's (at least from what I saw and what I think they are). I'm even more interested in pulling this darn heatspreader off but like everyone else I also don't want to kill my very expensive RAM! Let me know if you want any specific Vista/XP results or tests and I'll do the best I can.

No need to pull your spreaders -n7 linked me already, they're ProMOS. Plus I wouldn't want you to risk $300 lol! There was another guy in Taiwan with these EVO's and he got 'em running DDR2-1200 @ 2.0v in XP 64 just fine, but he's on DFI. He said he couldn't boot into Vista at that speed. So I really don't know...it's probably just the chips, not binned high enough or they're just not 100% stable for Vista. After seeing the results on those PI's, I'm getting crazily anxious about grabbing a set.....that's how I went broke....but just can't resist it...

If you could show me some Everest and game benchies with those in Vista I'd really appreciate it! Thanks man~

[Roger_D25]

Guys - You probably read through my post but just incase you didn't I want to be clear. The picture I posted above is not an actual picture of my modules, I just copied that picture from earier in the thread because it was almost exactly the same from what I viewed of my own modules. With that said however its very clear their ProMOS without a doubt, definatly not PSC.

LuckySevn - I've only had my Rampage Formula for about a week now so I'm not sure about this but maybe you could try using "SetFSB" to raise the RAM speed once your booted into Vista (not sure if there is a SetFSB for our board though)? Its been a while since I used that program so I kinda forget if it would even be possible to do that? I'm just about to start watching a movie with the girlfriend (don't have a TV so we use my computer) so I'll have to post some benchies tomorrow afternoon. Fortunatly once I'm done with work I've got the night off for testing and tuning!

-n7- Thanks for confirming my thoughts about the 4GBPI IC's, much apprecaited! I tried searching for more detailed information about the IC's (like you can for Micron chips by going to Micron's website) but didn't have much luck). Not much to learn, I was just curious about the tech information on the chips themselves.

EDIT: I just noticed LuckySevn that your running 8GB of RAM (4 x 2GB), no wonder your having some difficulties booting into Vista at 1200Mhz! Having 4 2GB high-density modules puts alot more stress on the memory controller, I bet you wouldn't have much trouble doing it with only 4GB (that is all I'm running, just a 2x 2GB setup).

[LuckySevn]

Thanx for the pic confirming your kit is PSC; i'll add it to my collection

We figured out the ICs of these G.Skills a few pages back
See Miravo confirm they are ProMOS here: http://www.xtremesystems.org/forums/...&postcount=146
(This was the post containing the review link: http://www.xtremesystems.org/forums/...&postcount=140)

Another review linked here: http://www.xtremesystems.org/forums/...&postcount=180
In spite of the bad pics, also shows they are ProMOS.

Thanks for the linkies n7! I'm definitely getting a set now! Forget i7!

[Roger_D25]

Good choice, i7 will be around a while and will most likely get better with time. So by the time your ready to upgrade it will be cheaper, better developed, and there will be more community support here for it (at least that is how I justified paying this much for that RAM), lol! Now the only problem you'll have LuckySevn is finding a kit in stock somewhere (I lucked out by finding a few kits at newegg but its real hit or miss). Make sure you keep us updated as soon as you get a kit!

[LuckySevn]

I like your justification on that. lol!

Yea I suppose I shouldn't get i7 when it just barely launched anyways. But dude those Pi's are nowhere to be found lol...Well, setup Egg Auto-Notify...hopefully it hits Egg's dock soon and I be the 1st to be notified. About their Auto-Notify thing though...very unreliable. Happened to me with HD3870 and HD4870. Both times Egg notified me last min, when I get on the very same day I received the notice, all outta stock again. -_-

[bigslappy]

auto notify worked for me , that's how i got mine .

[Highgertz]

-n7-
I bought the GA-EP45-DS5, it works a little worse with Perfomanc Level compared to the DFI LT X48-T2R, the rest as well overklok improved Xeon 3360 C1 - 4.42Mhz FSB 520Mhz
http://valid.x86-secret.com/show_oc.php?id=445916



Photo my G. SKILL F2-9600



I doubt the origin of these chips but in any case, they are the best

[geok1ng]

excellent memories, great review, but at best very promising numbers: it is amazing how every single post of memory(criminal/yoko aside, of course- yoko is one of the best memory tweakers in the net) in Xtremsystems focus on E-penis numbers instead of real life best performance setups.
on jan/2008 Anandtech published a seminal article on Intel memory timings and options:http://www.anandtech.com/mb/showdoc.aspx?i=3208&p=9 with the exception of one high FSB/low multiplier screenshot, 10 months later, using a very top of line/US$300 memory kit noone has beated the Everest numbers running 4(FOUR!!!) 1GB sticks!?
Why my fellow Xtrems? Why? Because instead of focusing on tRD (AKA performance level) people are going for raw Mhz and lesser important subtimings, even 10 months after the receipt for maximum memory performance on the C2D plataform was published. Allow me to quote AT:

[I][When viewed together the task of picking out a couple of the best choices becomes rather simple. A+ ratings, of course, go to the two configurations colored in light green (5:4 for DDR2-1000 CAS4 and 3:2 for DDR2-1200 CAS 5). You may be surprised, as we mentioned earlier, that they both use a humble 400MHz bus speed. If forced to choose another option, close second place awards might go to 450Mhz FSB - 1:1 and DDR2-900 CAS 4, or 4:3 DDR2-1200 CAS 5 (assuming your motherboard is capable of stable operation with the sometimes poorly implemented 4:3 divider). Just do not forget - whatever memory option you choose, be sure to remember the importance of tRD./I]

In the end we decided to run our Intel Core 2 Extreme QX9650 at a final FSB of 400MHz with a multiplier of 10.0x at an even 4GHz. Our 4x1GB of OCZ DDR2 PC2-8000 Platinum Extreme Edition memory was set to run at DDR2-1000 (5:4) CAS 4 with a tRD of 5. Ultimately, the ASUS Rampage Formula provided us the ability to build a well-tuned system. Without question, ASUS' addition of BIOS options for direct tRD manipulation was instrumental to our successful overclock. When paired with top-end GPUs from either ATI or NVIDIA, our rig becomes a formidable gaming platform for enjoying today's titles and beyond.

the lower the TRD, the higher the memory read performance and vice versa. This change in TRD also drives vMCH, with lower values requiring substantially more voltage than the more "relaxed" (higher) values. We were able to push our ASUS Rampage Formula to a very impressive 450MHz FSB using a tRD of 5, which correlates to a TRD of about 11.1 ns. Because this required about 1.60V vMCH, for all intents and purposes we would consider this to be the practical limit of scaling, at least with conventional cooling. Above this voltage, we find that the X48 quickly turns into a little furnace during periods of prolonged load. Our recommendation is the liberal use of extra cooling capacity if you feel the need to push any further. DESTROYING THE MYTH THAT WATERCOOLING THE NB IS USELESS!

Moving on to the analysis at 450Mhz FSB, although we are able to show impressive memory read speeds at this same bus frequency using a 3:2 divider for DDR2-1350 CAS 5 at a tRD of 5, this configuration was far from stable on stock cooling alone - as was the memory speed. Most likely, the need to loosen tRD to 6 will be inevitable at this bus speed. Right from the start, we can see that TRD takes a hit as our minimum values thus far came at the lower 400Mhz FSB. What's more, possible memory configurations at this FSB do not seem to offer any real improvement over those at 400MHz. The first option worth considering, DDR2-900 at CAS 4, can be a little slow, especially considering that a tRD of 6 is not allowed. The DDR2-1200 choice will provide (at best) only miniscule gains over the same memory speed at 400MHz FSB because of the additional throughput efficiency made possible by the higher bus speed. Then again, the increase in TRD to 13.3ns, up from the minimum value of 12.5ns at 400MHz FSB, may completely negate any chance of a performance gain. Our last contender, DDR2-1080 at CAS 4, might be another neutral choice - on one hand there could be a small gain in performance over DDR2-1000 (CAS 4) seen at 400MHz FSB, but again the higher TRD may be this configuration's undoing. THATS THE REALM OF MOST GOOD SCREENSHOTS HERE-tRD of 6 and 1200Mhz speeds CAS5- whats the point in paying US$300 in a 2x2Gb kit if a 10 months old 4x1gb kit can do the same or better?

this last quote shows what i am expecting from Xtrem users: to make a DDR2 system stable 24x7 with the dream setup: 3:2 divider for DDR2-1350 CAS 5 at a tRD of 5- go for it boys!

So it has been possible to run 4GBs (4x1gb) at 1000Mhz AND keep the CPU at 10x400Mhz for over 10 months, the real challenge for new 2x2gb memory sticks is to keep tRD at 5 with the new CPUs that can do 10x450Mhz on air. On the article AT points out that the system can not run 3:2 divider for DDR2-1350 CAS 5 at a tRD of 5 STABLE- new memory chips and top P45 mobos should try to break this barrier, not reinvent the wheel with tRD 6 CAS 5 1200Mhz+ numbers. The 1.60V vMCH is a nice starting point for (ab)users of these amazing memory modules!

[Roger_D25]

I've been doing some various testing tonight with this new RAM trying to get decent Read/Write/Latancy numbers while still being somewhat stable and I just got the best I've been able to do so far. Keep in mind this isn't impressive compared to what most of you guys do but for me this is great! This RAM is very impressive I think and I've loving this new Rampage Formula of mine! Besides the high clockspeed run below I did alot of testing at lower clocks and tighter latencies but so far this yeilded the best numbers. Plus as I'm sure most of you guys know already the most important setting so far seems to be "Common Performance Level" or "Performance Level", it easily has the biggest impact on memory performance. If I was watercooling my NB I'd would have tried lowering the PL from 7 (which is what I used for below) to 6 but to get any stability I needed to use more NB voltage than I felt comfortable with.

Yeah read the sam article and that is why I posted this with my above results. As you mentioned its great to have screenshots showing uber high FSB, RAM clocks, etc.. but in my own testing (as you mentioned) I've found the best "actual" performance when running at 400FSB along with lots of tweaking! Granted this may not yeild the highest benchmark numbers it does yeild the best overall performance/reliability/low voltages. You make a great point man and one that I wish I realized a long time ago! I am by no means an expert when it comes to this stuff but I'm quickly finding that using the methodology you mentioned in your post above really does net the best performance (even though I havne't been able to get the settings you mentioned stable yet). Thanks for posting that information, its something I wanted to post but couldn't quite articulate what I was trying to say (but you did)!

[Roger_D25]

UPDATE: Just wanted to let you guys know that newegg currently has a few kits of GSkill's PC2-9600CL5D-4GBPI modules available, get em while they last!

[ChaosAD]

Does anyone know if its possible to find this kit in Europe?