FX55 + GSKill...holy smokes!

[trans am]

here is a little unofficial rule:
you can't exceed total fron side bus speed around 1110~1120mhz(depends on the board)

for example, like my msi k8n neo2, I can run it at 280fsb with 4x htt, that's exactly 1120mhz..

is there a similar rule for 754? maybe can't exceed 900-920 total htt? This is very interesting.

[WesM63]

is there a similar rule for 754? maybe can't exceed 900-920 total htt? This is very interesting.

Idealy yes. In theory you don't want to exceed the max HTT supported by the board, NF3-150 800mhz, NF3-250 1ghz. However, you might be able to exceed it by a few mhz and be ok.

[some_young_guy]

is there a similar rule for 754? maybe can't exceed 900-920 total htt? This is very interesting.

Yes, it's the same for socket 754. For my old Giga K8NS Pro, I think the limit was 1020 mhz because I was stuck at exactly 255 HTT X 4 LDT, and 340 HTT X 3 LDT.

[Onepagebook]

Seems like I stepped on someone toes when barking at Gskill

OnePageBook,

those sticks were tested on A64 S754 mobo and 3700+ cpu, A64 S939 mobo and FX-53, Asus P5P800 and Intel P4 550 processor. With identical results.

You think it's still the freaking mem controller?!?

So I say case closed and Gskill didn't impress me.




Your running Sandra buffered. Mark the "Disable buffering" tab and you'll start getting the same results.

You know what? I am really lazy to explain all the OC things to the ppl who got bias on certain thing and just resist the own thoughts. If those results are identical for you, then all I have to say is I am so sorry that Gskill did not impress you but not you don't have 20 sticks ram to pick the best to review.
the sample for all the reviewers we snet were random select, nothing special.
Honestly, for me that's not enough for only test on 3 cpu. The point is not how many different cpu you test, it same type cpu but different one.
I did not mean to doubt your thought, but if you think that way, you are way far from OC filed.

[Hilppa]

Testing RAM with 3 different setups pretty much eliminates every possibility that is caused by other then the RAM itself. There is no way you should test it with 20 same type CPU's :banana4: What have you been smoking?!

If 3 different setups isn't enough to find a working combination with certain memories, I think it's pretty much a fault in the mem, don't you?

With these setups we have succesfully ran BH-5 / CH-5 at more than 260Mhz CAS 2.0 2-2-5, I wonder how it could be something else than the ram that's causing this lack in performance. I can't believe how persistently you are claiming that the fault is somewhere else.

It is the correct way to send products to reviewers, random pick, and just so happened that performance varies among gskill-memories, like with all the other memories. In fact, it seems this is quite the regular OC you get out of these sticks, atleast with the fast timings.
Check it out here


I really don't need you to explain me OC things

This is just the way I feel about the Gskill memories I tested.
Case closed, I stick with my BH-5 and run them happily everafter at 260Mhz CAS2.0-2-2-5, until you show me a Gskill that can beat that.

[Zeus]

[off-topic]OnePageBook, did you make the illuminated heatspreader in your sig yourself? I really like it, it's a true masterpiece! [/off-topic]

On-topic: Is this GSkill better than OCZ rev2, after all they're both TCCD chips on the very same 808 brainpower PCB.

I sure like how this Gskill has this double speed rating, 2,5-3-3-7@ PC4400 speeds guaranteed is nice, 2,5-3-3-7@ PC4800 speeds would be awesome.

Any chance Gskill will come up with some?

[Onepagebook]

Testing RAM with 3 different setups pretty much eliminates every possibility that is caused by other then the RAM itself. There is no way you should test it with 20 same type CPU's :banana4: What have you been smoking?!

If 3 different setups isn't enough to find a working combination with certain memories, I think it's pretty much a fault in the mem, don't you?

With these setups we have succesfully ran BH-5 / CH-5 at more than 260Mhz CAS 2.0 2-2-5, I wonder how it could be something else than the ram that's causing this lack in performance. I can't believe how persistently you are claiming that the fault is somewhere else.

It is the correct way to send products to reviewers, random pick, and just so happened that performance varies among gskill-memories, like with all the other memories. In fact, it seems this is quite the regular OC you get out of these sticks, atleast with the fast timings.
Check it out here


I really don't need you to explain me OC things

This is just the way I feel about the Gskill memories I tested.
Case closed, I stick with my BH-5 and run them happily everafter at 260Mhz CAS2.0-2-2-5, until you show me a Gskill that can beat that.

wanna beat your 260, 2-2-2-5?you really want that?ok! get your eyes ready!
if you still think this crap clock nothing, I will zip my sorry mouth and not argue with you any.
EDIT: oh sorry, I forgot to say it, DDR640? still piece of cake for this pair of ram, and this pair is just old gskill TCCD not even sort for "LE" or LA model.

[ar7786]

Onepagebook, how do you run MacOS on a PC?

[jamesfuh]

Onepagebook, how do you run MacOS on a PC?

haha...

The magic way is ...with your elbows knock to the screen windows then it might be forced to a MAC...

absent minded...

[jmke]

Testing RAM with 3 different setups pretty much eliminates every possibility that is caused by other then the RAM itself. There is no way you should test it with 20 same type CPU's :banana4: What have you been smoking?!.

These TCCD chips are made for extreme overclocking with relaxed timings, at CL2 2-2-5 you will hit a wall pretty soon, they are not BH-5 equals with tights timings.. but. When you relax the timings to CL2.5 3-3-7 you will find the limit of your Motherboard/CPU very soon.

When Onepagebook told you to try and test different CPU's he was referring to the A64 platforms; since the memory controller is on the CPU, the quality of the CPU will define your max FSB overclock, my first A64 3000+ S754 did not do 265Mhz stable with 2x512Mb TCCD sticks, now with a newer A64 3200+ I'm able to hit stability up 270Mhz; and 300Mhz with only 1 stick.

anyway, you seem to be missing the point of Onepagebook's remarks imho;

[G H Z]

Sorry,

Is this a Clawhammer or Sledghammer core?

[QuadDamage]

I like new Corsair stuff i just got. Likes volts too Haven't had time to really try.

[onethreehill]

Sorry,

Is this a Clawhammer or Sledghammer core?

FX-55 is a ClawHammer

[G H Z]

Thats what I thought, I guess Newegg is just stupid sometimes

But why do they sell the OEM FX-55(listed as Clawhammer 939) for $899, and the retail FX-55(listed as Sledgehammer 939) for $849

[SIK_L_CELL]

Why does the SPD tab on CPU-Z show my model number as 3-4-4-8Gskill? Could this part be mis-labeled?

[SIK_L_CELL]

These are supposed to be the 2.5-3-3-7 sticks. Also I am not blown away by these sticks. Not too stable at 275. I will give it a few days to tweak. After that I think I am going with the OCZ 4200 plat.

[trans am]

These are supposed to be the 2.5-3-3-7 sticks. Also I am not blown away by these sticks. Not too stable at 275. I will give it a few days to tweak. After that I think I am going with the OCZ 4200 plat.

Hate to say I told you so but I did.
did you try higher volts with the ddr booster yet?

[SIK_L_CELL]

Hate to say I told you so but I did.
did you try higher volts with the ddr booster yet?

Yes you did! 3-3-3-8 seems stable up to 285.

[trans am]

Yes you did! 3-3-3-8 seems stable up to 285.

what volts? what if you try just 1 stick? Just to see. I think the memcontrollers in the fx55 is the best, so I highly doubt this is a "blame it on the mem controller" case.

[Hilppa]

wanna beat your 260, 2-2-2-5?you really want that?ok! get your eyes ready!
if you still think this crap clock nothing, I will zip my sorry mouth and not argue with you any.

Nice you have that FX-55, you can try beat this SuperPi time at same CPU clockspeed and your TCCD-ram. After that, I admit your Gskill performs better, even though it's cherry picked, and start looking for a pair of those.


Today we will try these Gskill's with a new 3000+ chip and Abit mobo, just to make it sure.

And are you going to say to a customer who can't get these sticks working, that he's gonna have to buy 20 CPU's of the same type and look for the one that works

[mdzcpa]

Hilppa,

Please do me a favor and take your issues with Gskill to another thread and quite crapping on my thread please.

This is MY thread about MY results with some Gskill that I ordered from Excaliber. Nothing about it is cherry picked. Its just off the shelf. And heres a reminder (not the best thing out there, but not bad either):




PLEASE don't crap on this thread anymore!

[jmke]

those results look very good; how much VDDR

[mdzcpa]

2.7v.

[Hilppa]

OFFTOPIC

PC_ICE, I thought this was a hardware discussion forum, where you are supposed to exchange views and opinions on different matters, and tell others about your experiences, so we all would know a little more. Seems like to you all think is the kinda place where somebody posts some scores and everyone's replying: "yeah!", "rock on!", "great scores!", "You are great [place name here]"...

not allowed to post bad results on something that supposed to be good? Not allowed to say your opinion, even though it differs from others? ok, seems like im out of here for good.

Oh, and sorry if I ruined somebodys day.

[Ubermann]

I think the problem here is that GSKILL ppl started showing extreme high FSB and sandra scores right from beginning.
Then everyone that ordered GSKILL was hoping for those scores but it seems like that failed for many of em.
You should have posted results with decent sticks instead of those incredible sticks.
Or else everyone will think they got fooled.
But as long as they reach whats they are rated for they cant complain, but its not so fun =(
For me i bought something i thought would reach 280, but it went beyond 300 so i got very very happy!