New EVGA 8800GT's - got Samsung Vram!

deviantek

Weaksauce
Joined
Jan 2, 2008
Messages
98
UPDATE: I applied the 1.1 voltage bios flash and am now able to push my GPU to 760/1860/1070 hoorah :)

695/1800/2120
Using Riva Tuner & ATiTool
Video Card cooling accessories: Accelero S1 with included vram heatsinks + 4 copper thermaltake heatsinks for the mofsets

Hi all,

I recentlly put an order through newegg for a new evga 8800GT 512mb (base model @ 269.99). I received it Thursday the 10th and finally got around to taking off the heatsink to install the Accelero S1.

(NOTE: I didn't have a camera on hand :()

In any case:

The version I got was the newer one with the revised heatsink/larger fan.

When I took it off, I was surprised to see that there was no gunk or goo on any of the vram modules or mofsets.... strange? It looks like with the double sided heat sink tape they used they forgot to take off the the think plastic side that applies to the card itself, thus making it gooey and sticking. Either that, or there's some new NASA heatsink compound that is non-stick and orange plastic looking.

So, being that there was no goo crap covering any of the ram, I examined it and to my surprise I see that it's all samsung ram (below is what was on each vram):

Samsung 749
K4J52324QE-BJ1A
ADJ42GBO

I applied the accelero s1 heatsinks to the modules, and some spare copper sinks to the moffsets, and proceeded to install the accelero s1. (THIS THING IS HUGE and I'm having a bit of trouble crunching it into my CM Stacker 830)

So, as I write this, i'm busy installing forceware drivers and rivatuner/atitool, and I'll see what I can come up with as far as overclocking is concerned.

p.s. can someone please tell me if that samsung ram I listed above is 1.0, 1.1, or 1.2 NS?

Thanks,
 
If you go to the samsung semiconductor site, there's a part # search on the upper right. Those, because they have a 1A suffix, seem to be 1.0ns RAM, so they should do 2000MHz no sweat. Good luck overclocking it, let us know how it turns out.
 
I guess I got samsung ones too because I had the bigger fan, black pcb, red no gooey tape.
 
I guess I got samsung ones too because I had the bigger fan, black pcb, red no gooey tape.

If you look closely, the tape on the heatsink side is gooey white, but for some reason the side that contacts the 8800GT board is red non-stick. Is it supposed to be that way?
 
Black PCB? No, it's still the green PCB w/ the newer open-ended larger fan heatsink.
 
Went ahead and added my OC results thus far. I'm using the latest riva tuner & ATiTool to perform the testing (as well as some 3dmark06 runs). Running it on vista x64
 
The great thing about the 8800GT's is you can give them a little voltage bump from 1.05V to 1.1V via a simple BIOS flash. You might want to try it since you already have excellent cooling :D.
 
The great thing about the 8800GT's is you can give them a little voltage bump from 1.05V to 1.1V via a simple BIOS flash. You might want to try it since you already have excellent cooling :D.

Yeah, I heard about this though I can't remember where I saw it. Would you mind linking the page for the flash utility/bios update?

Thanks!
 
Anyone have a link comparing the cooler styles? I got my EVGA 8800GT last week from NCIX but I didn't know they were shipping different versions.
 
It's essentialy this, but the card with the bigger FAN having the heatsink open-ended by the 6pin power connector on the eVGA:

leadtek_px8800gt_06.jpg
 
My story:

I ordered an evga 8800gt SC and when I recieved it, it crashed in anything in 10 seconds. I then checked the temps but they never went above 60 C. I then underclocked the ram to 900. That gave me about 20 minutes of playing time until a crash. Might be a temp problem still? SO then I slapped on an accelero s1 which gave me 35 C load :) but still crashed..:mad:

I then rmaed it to newegg. When they recieved it they said they had no more evga 8800GT's so they were just going to give me a refund :mad:. I then complained because it was just wasted time :mad:. I tryed to just get a xfx one in exchange but they would not agree...even thought they cost exactly the same..The next day I got my refund. I then order an xfx 8800GT Extreme Green pcb smaller fan

I then got an email from someone from newegg saying that they bumped me up to 2nd shipping and gave me a 25 dollar coupon on my next purchase :)

When I recieved the GT I was supprised when it was the black pcb, bigger fan one :)
I then slapped on the accerlero S1 and I noticed that they were using non gunky red tape for the memory. Didn't check the brand though. 30C idle temps :) @ v1.1 740/1800/1000
 
695/1800/2120... badabump! shame I'm having such shit luck on the core.... I don't know why either. I have an accelero s1 on this baby. I wonder if I made bad contact with the GPU when installing the heatsink?
 
Awesome :)

Unfortunately however, I have no floppy or usb flash drive handy and I'm running vista x64 ~_~

Is their a way to simply flash the bios via software while in windows 64 bit?

Store your edited bios on an easy to find (and type) place on your HDD... simply on the C:\ drive works great. create a boot CD with Nero and when you get into your DOS prompt using that, run your nvFlash with your new bios.

Hope this helps... the added .05v will for sure :)

Grey
 
I wish I could figure out how to measure my temps, but no programs I've found thus far support the temp measurement in vista x64. Currently using rivetuner, gpu-z, & atitool. If you know how I can measure the temps please let me know.

Thanks,
 
What temps are you getting with 760 core? How did you set up the S1? Passive? Fans?

I have the s1 included heatsinks on the vram chips, and thermaltake copper sinks on the moffsets and exterior side hot spots.
I have 2 aspire 120mm Fan's zip tied on to the card.
 
UPDATE: I applied the 1.1 voltage bios flash and am now able to push my GPU to 760/1860/1070 hoorah :)

695/1800/2120
Using Riva Tuner & ATiTool
Video Card cooling accessories: Accelero S1 with included vram heatsinks + 4 copper thermaltake heatsinks for the mofsets

Hi all,

I recentlly put an order through newegg for a new evga 8800GT 512mb (base model @ 269.99). I received it Thursday the 10th and finally got around to taking off the heatsink to install the Accelero S1.

(NOTE: I didn't have a camera on hand :()

In any case:

The version I got was the newer one with the revised heatsink/larger fan.

When I took it off, I was surprised to see that there was no gunk or goo on any of the vram modules or mofsets.... strange? It looks like with the double sided heat sink tape they used they forgot to take off the the think plastic side that applies to the card itself, thus making it gooey and sticking. Either that, or there's some new NASA heatsink compound that is non-stick and orange plastic looking.

So, being that there was no goo crap covering any of the ram, I examined it and to my surprise I see that it's all samsung ram (below is what was on each vram):

Samsung 749
K4J52324QE-BJ1A
ADJ42GBO

I applied the accelero s1 heatsinks to the modules, and some spare copper sinks to the moffsets, and proceeded to install the accelero s1. (THIS THING IS HUGE and I'm having a bit of trouble crunching it into my CM Stacker 830)

So, as I write this, i'm busy installing forceware drivers and rivatuner/atitool, and I'll see what I can come up with as far as overclocking is concerned.

p.s. can someone please tell me if that samsung ram I listed above is 1.0, 1.1, or 1.2 NS?

Thanks,

Do you have a before/after stock bench and overclocked bench?
 
Awesome. I have a similar set up but with Thermalright sinks on my mosfets. I'm definitely going to do the BIOS voltmod to get more juice out of her :D
 
I wish I could figure out how to measure my temps, but no programs I've found thus far support the temp measurement in vista x64. Currently using rivetuner, gpu-z, & atitool. If you know how I can measure the temps please let me know.

Thanks,
Rivatuner should measure temps fine. Go to the hardware monitor, it should be right there.

Great clocks you got, what kind of improvement in applications are you seeing? :cool:
 
Rivatuner should measure temps fine. Go to the hardware monitor, it should be right there.

Great clocks you got, what kind of improvement in applications are you seeing? :cool:

It was quite noticeable in 3dmark. I think it went up around 10+FPS.

However I haven't had much time running other apps, as it looks like I'm having troubles with the OCZ DDR-400 ram I recently ordered... but that's another story. hehe.
 
Do you have a before/after stock bench and overclocked bench?

I'm afraid that I don't, sorry. I had the card and the S1 sink come in the same day and figured I'd save myself sometime and install the S1 before putting the card in my case.
 
I'm afraid that I don't, sorry. I had the card and the S1 sink come in the same day and figured I'd save myself sometime and install the S1 before putting the card in my case.

Well, I mean the performance differences between a stock clock and your overclock. The reason I ask is because I have a watercooled 8800, and I got to a top OC of 620/905 - on the first core revision - max temp 39c under load. Unfortunately it smoked itself and I'm receiving the new updated core which overclocks like a madman, and on water I expect to go pretty high - just wondering if the performance improvements were minimal - I sometimes couldn't even see a difference at all in games - especially with world in conflict - the max frame rate would increase 5fps, but the average would decrease. I've been OC'ing cards for a while and I really just enjoying seeing before/after.
 
Many RTS games are CPU limited as it is, so I don't think it would show as much difference in the first place (correct me if I'm wrong).
 
Have you ran 3dm06 yet? I'd like to see what score you'd get with the OC'd settings as opposed to stock.
 
Many RTS games are CPU limited as it is, so I don't think it would show as much difference in the first place (correct me if I'm wrong).

Yeah, I would say that's probably correct. To base real world performance OC results in an RTS game is a bit idiotic in my opinion.
 
Have run 3dm06 yet? I'd like to see what score you'd get with the OC'd settings as opposed to stock.

I'm currently @ work, so I can't right now, but I will later tonight. I've been too busy fiddling with my DRR-400 ram settings which decided to not post above 151mhz ... it's been quite frustrating as I was just running it at 216mhz on the same latency the other day. ugh.
 
Yeah, I would say that's probably correct. To base real world performance OC results in an RTS game is a bit idiotic in my opinion.

I have an E6850 @ 4.4GHz - I haven't noticed much difference in overclocking higher, so I would say it is definately video card related. Idiotic is uncalled for.
 
I have an E6850 @ 4.4GHz - I haven't noticed much difference in overclocking higher, so I would say it is definately video card related. Idiotic is uncalled for.

How can you not notice a difference from a reference clock of 600/1500/900? Unless you're comparing it to your "and I got to a top OC of 620/905"
If that's the case, you have something seriously wrong in your cooling set up.

I can visibly watch the 3dmark06 FPS jump up 10+FPS from stock speeds to the new OC clock.
 
How can you not notice a difference from a reference clock of 600/1500/900? Unless you're comparing it to your "and I got to a top OC of 620/905"
If that's the case, you have something seriously wrong in your cooling set up.

I can visibly watch the 3dmark06 FPS jump up 10+FPS from stock speeds to the new OC clock.

Yes, I see them jump in 3dmark. My reference clock is 576 - so from 576 core to 620 I see about a 5-10fps difference - it decreases depending on the load however. It's less substantial depending on the game. My point is, I've been unable to really flog my 8800GTS to see what a seriously decent OC will yield as far as real world results, so I'm curious as to the OP's before/after.
 
Yes, I see them jump in 3dmark. My reference clock is 576 - so from 576 core to 620 I see about a 5-10fps difference - it decreases depending on the load however. It's less substantial depending on the game. My point is, I've been unable to really flog my 8800GTS to see what a seriously decent OC will yield as far as real world results, so I'm curious as to the OP's before/after.

Well for starters, this discussion was about an 8800GT, and you failed to mention in your prior post that you were talking about an 8800GTS. People don't simply overclock for the #'s, they do it because it does in fact boost the frames and smoothness of your 3d application (whatever that application may be, the performance increase will vary + factor in the total increase of your OC yield).

As far as the before/after. I wouldn't know as I've the only program I've measured the 8800GT's FPS/performance gain in is 3dmark06 thus far.
 
Well for starters, this discussion was about an 8800GT, and you failed to mention in your prior post that you were talking about an 8800GTS. People don't simply overclock for the #'s, they do it because it does in fact boost the frames and smoothness of your 3d application (whatever that application may be, the performance increase will vary + factor in the total increase of your OC yield).

As far as the before/after. I wouldn't know as I've the only program I've measured the 8800GT's FPS/performance gain in is 3dmark06 thus far.



Well no shit people overclock for performance:rolleyes: I was wanting to see what those performance gains were for his before/after. I've measured score differences in 3d mark, and from 576 to 620 I see a 1000-1400pt increase depending on how aggressive it is.
 
Well no shit people overclock for performance:rolleyes: I was wanting to see what those performance gains were for his before/after. I've measured score differences in 3d mark, and from 576 to 620 I see a 1000-1400pt increase depending on how aggressive it is.

Last reply for you, as you're pulling my thread completely off-topic. In case you haven't noticed the "his before/after" you're referring to is ME. Maybe grab a cup of coffee? Try a thread based on your 8800GTS (which uses a completely different GPU core mind you).
 
Last reply for you, as you're pulling my thread completely off-topic. In case you haven't noticed the "his before/after" you're referring to is ME. Maybe grab a cup of coffee? Try a thread based on your 8800GTS (which uses a completely different GPU core mind you).

let me know how your before/after comparison looks.
 
Back
Top