Discussion in 'Videocards - AMD Radeon' started by soul8103, Aug 29, 2005.
Now that I've given up on my x800XT PE (and I was so happy when I bought it..it looked so shiny and powerfu)....which card should I get--the 6800GT or the 6800 Ultra?
I'm not a benchmarking-fanatic, so I don't care if there is a 3 fps difference or a 200 3dmark05 point difference--which one will give me more bang for my buck?
Coming from a XT PE, id go for a 6800 Ultra (or possibly 6800GT Golden Sample) as that would be the equivalent card (sort of, not counting the 6800 Ultra Extreme Edition). Saying that, most 6800GTs will overclock to Ultra, but id be a bit pissed if the one I got didnt.
I took into account your advice and went with the Ultra instead of the GT--it'll cost me 100 bucks extra--but I don't want to have the feeling that I'm downgrading rather than upgrading my computer from my current setup. If I continue to have these ridiculously low FPS--I'll either jump out of a window, or throw my computer out of it I'll tell you how it goes after it arrives on Tuesday--FYI, it's an MSI NX6800 Ultra.
Let me be the first to wish you tons of luck Mr Pro.
I'm with you on that sh!t for sure.. my computer will have a boulder tied to it and dropped to the bottom of the lake.
Soul: Glad to hear you're not giving up on the gamine rig just yet. I plan on doing the same, making an entire new rig just for gaming. However, I'm planning on going travelling first, so that will have to wait for a year and a bit. All the best with it though.
I'll be on vacation for the next week and a bit. This will definitely be swimming in my mind here and there, and I'm very interested to see what happens with Mr. Pros new card.
Guess all I can say is whenever your guys' fps drops down to a new low, just roll one up and ease your mind. No sense losing sleep/braincells/erections over it
Maybe it's time to let go of this whole issue and wait for the card to arrive...but I was just wondering what CPU scores you guys get in 3dMark05...I've been thinking that the reason why we may be getting slow performance in games but not in 3dMark05...may be because the latter specifically tries to use GPU power, not CPU power, in the graphics demos. Is it possible that the CPU in our computers may be malfunctioning/overheating etc., thus limiting what we can get in games--but not in 3dmark05? My Mobo temperature is at 35 when playing games, my CPU is at 60--but that's within normal bounderies and shouldn't casue problems, right? Is it possible that I've made a mistake while installing the CPU (which I've done by myself, not wanting to get a technician..the cheap arse that I am)...or would it just plainly not work at all if that was the case?
No,it would however produce more heat if the hsf is not seated properly.One thing to take in account is that the videocard temperature can rise and drop very quickly.So when playing a game it can go very high,but when exiting to the desktop there can be already a drop of 10 celcius.
A system temp of around 35 and a cpu temp that's almost double would seem to point at either a wrong bios reading or bad contact from the hsf to cpu surface.
It's more common to have a difference of 10/15 celcius between cpu and system temps.
However,my videocard can reach 85celcius and still maintain the fps in a game.The darn thing runs between 70 and 75 celcius when i play farcry but one time i fiddled with the fan speeds and forgot to set it back resulting in too low rpm,that's when it reached 85C.,i noticed that when checking atitool after the game.
One thing is for sure,i can oc my pc or use other drivers,my 3d05 will give higher scores.It can go at least 500 points higher.
But it does not translate in my games.
For now i'm settled at 5300 3d05.The x800xl should be close behind that score at stock settings on a similar rig.
Yeah, now that I think about it, the 30 degree cpu/system difference seems suspicious...I have all the latest BIOS, so it probably isn't faulty BIOS reporting. HSF, I take it, is a heat-sink-something--how do I check whether it's connected well to the CPU (after physically detaching the two, that is--although I don't recall how to open those tricky hook-things that connect my heatsink to the CPU)? I'm becoming increasingly suspicious of the CPU as the source of my woes--in crowded areas in Fable and GTA, I take a 20 fps performance hit even if my character is looking at a wall--which indicates that it's the CPU computing the position of the other NPC characters--rather than the GPU actually slowing down the game because it's drawing them. Is there any reliable CPU test/benchmark out there to show me whether my CPU may be the bottleneck b/c it's malfunctioning, etc..?
P.S-also, is there an utility to measure GFX card temperature?
Yeah I would also like to runs some tests of that sort, but saying that we all have simliar problems with similar hardware, coincidence, or maybe we all have faulty cpu's, who knows. 1 thing ive noticed on my rig though is that the cpu speed in 3d05 tests does seem pretty slow.
After installing the card i formatted my C drive. Once reinstalled my mobo wasnt being read properly. In device manager mobo devices were question marked. At the time I was hitting an all time low 0 fps in the cpu test without getting thoe drivers. After installing the missing devices the cpu speed went up (I can peak at 4 fps in the cpu test, avg around 2-3 on the very last demo). I was wondering if this may have anything to do with the issue at hand, and if u guys had similar problems if you formatted once installing the card, and had devices missing.
My cousin came over 2 days ago who knows a lot about computers as he is an engineer (working for aerospace atm, freakin unreal) and he said most likely its a ram issue, or a mobo issue. I figure that if a processor is faulty in anyway, booting into windows alone would be alone miracle right?...
On the ram side of thing i know me PC3200 ram only runs at 200 mhz, as it is on the bargain side, so do you reckon that ccould also use a shape up?
Jus throwing things insto the pipeline to see if any more suggestions are about.
With atitool 0.24 you can check min/max/average temperatures and also set various fan speeds.That is if your videocard has the hardware installed.My card came with all sorts of software except for hardware monitoring.When i wanted to overclock i read about the atitool (i don't have overclocking options in control center) and found it very usefull,you can also test for artifacts and let it search for max mem/core settings.
As for the cpu,on a skt939 it can get stuck in the corner of the cpu bracket,there is a small plastic edge in each corner sticking out.I had this when fitting my hsf (heatsink+fan idd) luckily it clicked in place after some pressure,i couldn't get the retention clip closed.But just before that,i was thinking : it sits firm so lets forget that extra lever.I'm glad i persisted with that.
This was with the standard hsf from amd.
In any case,it wont hurt if you remove the hsf and clean it togetter with the cpu surface.Then apply some thermal paste and make a correct fit.If applied too much paste things can get hot again,use not enough and it's the same story.
Reading that you guys have an agp card makes me also wonder if that's what is causing the troubles.
Try some older mainboard drivers,or different via hyperion drivers (if applicable)
Check in registry if you have multiple monitor references,after every new driver install for my vidcard i get an extra one in registry,this messes up my monitor frequency in games etc.
Searching for the brand name will get you there.
It's also worth to check if there is the correct line in registry for your graphics card,it's somewhere in hkey_local_machine\currentcontrolset\enum\pci
If your card supports multiple monitors there should be 2 lines.
A few questions (from the technically retarded):
1. ATI tool shows that my GFX temperature doesn't go over 50 (that's the peak w/ a heavy graphics load in FAble)...is this normal?
2. By 'mainboard drivers' you mean chipset INF. drivers? E.G. in my case Intel 865PE...
3. I don't quite know what you mean with the multiple monitor registries, being a complete registry-newb--but I'll link to a screenshot of the registry results after I searched for "ATI"--the ones that popped up in the "currentcontrolset" category...tell me if anything looks fishy, please
4. In fable and other games...small areas have very high FPS, no matter what is shown on the screen (e.g. a dozen characters) wheras large areas have low fps (with the same number of buildings/npc's actually VISIBLE as in the first case)....it's like a difference between 80 fps and 20 fps! Am I not reading htis correctly--or is this really a sign of a CPU related problem? NB: I have a socket 478 P4 3200mhz.
Registry SS URL: http://img153.imageshack.us/my.php?image=reg0km.png
It's not the registry entry i meant.
This is what i meant.
Well,i suppose it's not that important,but perhaps you upgraded from a nvidia card and some leftover entries could be present yet.
As for the temperature,i think i should worry more,since under full load i go beyond 70celcius.Even with a fan blowing under and to the side of it, and a fan sucking out.
I don't really know what temperatures the x800xl should have.
In any case,the backplate on my card is hot to the touch,so the 70C seems realistic.
Hm,just came to my mind,is my card so hot because it's working to the fullest and yours not?Or did you replace the cooling device from your videocard for a better one?
I replaced the cooling device but the issue was present before i did...
the only proplem that you is that your x800xl sucks you should get an fx5500 like me it is the best
Well, thanks for that objective, qualified and above all--helpful opinion.
Member Guru you talk total cr@p m8,lol,do you know anything about x800xl's you newb?
A fx5500??? ROTFLMAO!!!
My Sapphire X800XL Ultimate (w/Zalman cooler) hits 56C max when playing DoW for hours.
Just wondierng how your making out Soul, and Mista..
Got back from holidays and I'm considering selling my XL to the highest bidder and grabbing a 6800GT..
has anyone here tried to swap out their card yet? If so what were the results?