2900 is getting better by the day

I know its been posted, but quite a few people are still on the fence-looking for a new gpu, heres what 1 driver update did http://www.hkepc.com/bbs/hwdb.php?tid=790821&tp=AMD-R600-HIS&rid=790821 This is usually a good site. Will it beat out the GTX? No. Will it win some benches in some games? Yes. If the XTX hadnt gone down in flames (or almost) being sooo hot, I think it would have won against the GTX. The 2950 will be coming out, at 65nm as reported, itll clock high. If they get their leaking gpu problems behind them, were going to see a nice card. Flames or comments are welcome
60 answers Last reply
More about 2900
  1. i think this link is faster:P

    http://www.hkepc.com/bbs/hwdb.php?tid=790821&tp=AMD-R600-HIS&rid=790821

    Quote:
    The 1950 will be coming out, at 65nm as reported, itll clock high. If they get their leaking gpu problems behind them, were going to see a nice card. Flames or comments are welcome


    you mean the 2950??
  2. Maybe if I buy one soon itll help in my posting :oops:
  3. I don't know witch benchmark to trust anymore...One site says that it is about 8800GTS & the other says that it's faster than 8800GTX... :?
    I think we need some one here to go out and buy a HD2900XT! :twisted:
    Oh man! This is going to take the whole day to post this reply! :(
  4. I'll buy it! just give me the money :roll:
  5. If you read the article, its the newest drivers. And they say theres more better drivers to come.
  6. Shame it looks so ugly, but I suppose when it's hidden in your machine..... :) Just hate red cards, even though I've owned one or two.... what's wrong with black?
  7. I was somewhat dissappointed when it first arrived. But the best of hopes are coming true. I dont need it to beat the GTX, but its hanging close now
  8. Great Find Jaydeejohn! A lot of us were thinking driver improvements would make a big difference with the 2900 XT and these tests seem to confirm it. Real interesting how it just about creamed the 8800 GTS 640 and came real close to the GTX (and even beat it a couple of times). With the 8.38 drivers, it looks like the 2900XT has the sweet spot for price/performance. Now... if somebody will only make it a little quieter and cooler, I may have found my next video card.

    Rob (obviously not of SLI fame)
  9. Oops - one additional thought (which also gets my post count up a little bit). I'm finally glad to see a site start using something other than an NVIDIA motherboard to test an ATI videocard. Although I can't read Chineese, it looks like they used a MOBO with the Intel 965 chipset. I got so sick of seeing every review of the 2900XT being conducted on 680i chipset motherboards. I know - a lot smarter people than me say that an ATI videocard should perform fine on an NVIDIA chipset, but I'm still suspicious. I think the 965 chipset is a great performer and at least has apparent neutrality when comparing the videocards.

    Rob
  10. Quote:
    Shame it looks so ugly, but I suppose when it's hidden in your machine..... :) Just hate red cards, even though I've owned one or two.... what's wrong with black?


    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need
  11. Great find ... :D

    I'm so sick of all the prophets on these forum these days, who suddenly became GPU experts and all they say is how much the R600 sucks and that it will be the end of ATI etc …

    The truth is that the initial drivers were pretty lousy … but this was the case in previous launches from ATI … they got better and better with each new driver release (remember OpenGL performance!!!)

    This is certainly no excuse for them but common people … just wait at least for the drivers to stabilize before start prophesizing the doom of R600 and ATI (AMD present)
  12. Yeah, if you've got a light source beaming on it... which would be very unlikely. I'm pretty sure colour doesn't make any difference. It might have something to do with the ridiculous amount of electrical energy it converts to heat...
  13. From what I'm seeing in the posted article it looks like the performance with AA and in higher resolutions has improved allot... I was definitely expecting that...
  14. If 2900xt really got this much better from a driver update only a few days after release, it was pretty stupid of AMD not to wait for the new driver.

    Now many people has already made up their opinions about the card and not everyone will have the interest to read new reviews.
  15. Quote:
    If 2900xt really got this much better from a driver update only a few days after release, it was pretty stupid of AMD not to wait for the new driver.


    They'd already delayed a long while, and had to deliver according to the schedule they'd already finalyl commited to.

    Quote:
    Now many people has already made up their opinions about the card and not everyone will have the interest to read new reviews.


    Rubbish. The people that don't read don't know, and the people that do read keep reading.
  16. Quote:
    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need


    Maybe for absorbing radiant heat. But ....

    the card is generating heat and the goal is to cool than the card - so the card will not be absorbing heat, it will be emitting heat.

    Most heat from the card is transferred to the surrounding air by convection rather than by radiation. Radiant cooling at these temperatures is negligible.

    There is a belief that high performance car engines cool better if they are black. Do not know if this is fact or marketing but ...
  17. Quote:


    Rubbish. The people that don't read don't know, and the people that do read keep reading.


    spot on! :)
  18. Quote:
    Shame it looks so ugly, but I suppose when it's hidden in your machine..... :) Just hate red cards, even though I've owned one or two.... what's wrong with black?


    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need

    Don't think that would be a consideration. Think about it.....
  19. Quote:
    Shame it looks so ugly, but I suppose when it's hidden in your machine..... :) Just hate red cards, even though I've owned one or two.... what's wrong with black?


    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need


    Ugh.

    No, just no.

    Black also radiates heat faster than red, and since the card is hotter than the surroudings, a black one will lose heat faster.

    This is, however, rather irrelevant, as the casing is made of plastic and is therefore an insulator, meaning it's colour will have very little effect on the heat absorption.


    To Borat guy: infra-red is a type of light (electomagnetic wave), and objects give off infra-red, even when they aren't hot. When infra-red is absorbed, it turns into heat.
  20. Quote:
    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need

    My XFX 8800 GTS is BLACK! & it is Much cooler than my previews RED X1900XTX (at least 20~30c).
    :roll:
  21. I just got my 2900 XT, and it is pretty nice. I haven't got to do any real tests yet, but it plays great at 1280x1024 in Oblivion and M2TW (big suprise :lol: ).

    It also doesn't seem very loud or hot to me.

    The drivers are a pain though. They are bulky and difficult to install properly. I will probably try to reinstall them soon. Also, I had a weird 3D mark glitch. 3D mark gets stuck in an infinite scanning loop. It never ends the initial scan (it doesn't freeze, it just never ends). Other than that, the card runs beautifully in the real world. I am very impressed with it.
  22. Quote:
    Black absorbs heat and other energies alot moe that other colors.And heat is something the 2900xt doesn't need

    My XFX 8800 GTS is BLACK! & it is Much cooler than my previews RED X1900XTX (at least 20~30c).
    :roll:

    :lol: :lol:

    The internet just brings out the stupid side of people (or just the stupid people :?: ) Probably the way in which computer screens refresh which attracts dumassedness.


    Actually, this is a right downer for me; posts like this just illustrate the lack of education which is rampant in the world today.
  23. any chance any chinese speaking user can translate us?
  24. @ the red vs black 2900 debate:

    even if they make a black version, it will still end up being red, because it will be glowing from all the heat it puts off :P
  25. paint your XFX 8800 GTS red and we can end this debate once and for all.
  26. Quote:
    Ugh.

    No, just no.

    Black also radiates heat faster than red, and since the card is hotter than the


    I just can't let this much misunderstanding pass by without commenting.

    This is a typical mistake. I suspect people are confusing a few different topics and misunderstanding the concept of a black body and black body radiation. Black body radiation is simply the radiation of photons from all objects (not counting any radiation coming from external sources and reflected off) that are not at the theoretical temperature of 0 degrees K. The spectrum of the light (photons) emitted depends entirely upon the temperature - only very very hot things emit photons in the visible spectrum (that we can see; e.g. stars like our sun are visible because of their temperature. If you cooled them down, they would dim, but even once they became black to our eyes they would still be warm and emitting quite a bit of light, just outside of our range). Temperature and only temperature affects the spectrum - whether we are talking about plastic or metal or anything else (though naturally some materials would combust in an atmosphere like ours, at high temperatures).

    Now, as for the radiation of energy from an object depending upon it's color, this is absolute rubbish. Objects that appear black to our eyes do not reflect (but rather absorb) light within the particular portion of the spectrum of light that our eyes detect - they still radiate light, as a black body (depending upon their temperature), they still transmit light (if they are very thin, light may or may not, depending upon the substance and the thickness, shine *through* an object - e.g. hold paper up between your eyes and a light), and they still reflect light ... they just don't reflect it in your visible spectrum. Likewise, objects that appear blue reflect blue photons (that is a certain frequency/wavelength range - wikipedia lists it as 450–495 nm) but absorb photons that have larger than smaller wavelengths (within the visible spectrum).
    In other words, the color that you see is a mere reflection of the different frequencies reflected by objects - when it comes to radiating heat, this simply does not factor into any of the equations.
  27. Actually they're thinking about the tidbit they picked up on why blacktop gets so hot in the summer, without ever really delving in or understanding it. It's simple. The colors we see are actually just parts of the spectrum blended from a light source, usually the sun. No particles actually have a property of "color", color is part of the spectrum of light. What we see isn't a property of an object, but our eyes are able to sense different parts of the light itself. This light "spectrum" or range of colors is actually different wavelengths, much like a radio signal frequency. Our eyes are only capable of detecting a specific frequency wavelength range, but there are shorter and longer wavelengths as well that our eyes can't detect. Cats and other animals are able to see in the dark because they're eyes are more sensitive to certain wavelengths, while we can use infrared technology to detect even shorter wavelenghts and create a visual image with technology that our eyes CAN see. The reason that a white T-shirt does not absorb as much heat as a black one does, is because the properties in the fabric or dye cause the T-shirt to reflect more of the spectrum or short wavelengths, rather than absorb them. Think of carpeting in a sound studio muffling sound vibrations, or the ripple effect of a pebble dropped in the middle of a lake. By the time the ripple reaches the lake's edge it's tiny. A black object will absorb more of the sun's light, which is actually a form of radiation. Radiation is energy, just the same as everything we know. Matter is energy, you can split water molecules to get energy, just as you can fuse them to get energy. (Research a hydrogen bomb or fusion to learn more about energy) Anyway, color only makes a difference with heat when it comes to light. A house with a white roof will be cooler in the summer, but a black roof will warm it in the winter. It's why neutral colored cars for climates that have both extremes usually prove to be the best, and in hot climates on the equator a white car is better. In cold antarctica a vehicle with a black hood would start just a bit sooner because the hood would absorb more light, which is energy, which is heat.
  28. Quote:
    paint your XFX 8800 GTS red and we can end this debate once and for all.

    OK OK Here just for you guys! :wink:

    WOW, the temps go down from 70Idle/82Load to -100Idle/-50Full load! :o
    Ice cream anyone? 8)
  29. The Hitchikers Guide to the Galaxy has brougt us the babelfish, also on the net.

    Try typing the url into the translator found at babelfish.altavista.com I've found, in this case, that the Traditional Chinese to English does a good job :)

    Enjoy!
  30. I feel ignored. But it's cool. I'll just rack up my count with a double post! :-)

    Actually they're thinking about the tidbit they picked up on why blacktop gets so hot in the summer, without ever really delving in or understanding it. It's simple. The colors we see are actually just parts of the spectrum blended from a light source, usually the sun. No particles actually have a property of "color", color is part of the spectrum of light. What we see isn't a property of an object, but our eyes are able to sense different parts of the light itself. This light "spectrum" or range of colors is actually different wavelengths, much like a radio signal frequency. Our eyes are only capable of detecting a specific frequency wavelength range, but there are shorter and longer wavelengths as well that our eyes can't detect. Cats and other animals are able to see in the dark because they're eyes are more sensitive to certain wavelengths, while we can use infrared technology to detect even shorter wavelenghts and create a visual image with technology that our eyes CAN see. The reason that a white T-shirt does not absorb as much heat as a black one does, is because the properties in the fabric or dye cause the T-shirt to reflect more of the spectrum or short wavelengths, rather than absorb them. Think of carpeting in a sound studio muffling sound vibrations, or the ripple effect of a pebble dropped in the middle of a lake. By the time the ripple reaches the lake's edge it's tiny. A black object will absorb more of the sun's light, which is actually a form of radiation. Radiation is energy, just the same as everything we know. Matter is energy, you can split water molecules to get energy, just as you can fuse them to get energy. (Research a hydrogen bomb or fusion to learn more about energy) Anyway, color only makes a difference with heat when it comes to light. A house with a white roof will be cooler in the summer, but a black roof will warm it in the winter. It's why neutral colored cars for climates that have both extremes usually prove to be the best, and in hot climates on the equator a white car is better. In cold antarctica a vehicle with a black hood would start just a bit sooner because the hood would absorb more light, which is energy, which is heat.
  31. Thought as much. ATi has previously released cards that had "immature drivers", but one or two drivers down the line they got some perf back.

    With the R600, I think they will have their hands full for quite some time, as will NV. The threading aspect is just getting mind boggling right now. I'd hate to do drivers for these cards... :lol:
  32. JonnyDough, please read mattc's post again before you make that a triple posting. No offense, but you making many generalizations and they are for the most part incorrect. Please see also this post on a similar issue a few months ago...

    JumpingJack and Hotfoot discuss radiative heat transfer and emissivity
  33. im confused. i thought color only mattered because black absorbs a wider range of light. but if the card is in the case what light can it really abosorb. the only light getting in really should be out side of the range that we see as color and should not depend on what we see as color. unless you have a light inside of your case i cant see how the color would really matter. maybe if the case had a plastic side, but you would still need to have a lot of direct light wouldent you?
  34. Not sure who is doing it but I think someone is crossing their photons with their electrons
  35. Even though it is scoring nice, I think the GTS is still a much better buy. Its cheaper, keeps up, runs MUCH cooler, and uses MUCH less power. I think ATi, even though the card is late, rushed it still yet. Their 80nm process must suck.

    I'm just glad my x1950Pro can still keep up. I am going to skip this round of cards, maybe the 8900 or 2950 will be more up my alley.
  36. Does it help with the power/heat issues?
  37. Quote:
    Thought as much. ATi has previously released cards that had "immature drivers", but one or two drivers down the line they got some perf back.


    Exactly. Both ATI and nVidia have pushed products onto the market with "immature" drivers. Increased performance from subsequent drivers releases is nothing new. It just seems that the 2900 is getting so much attention (hence a bad rep) because if the AMD/ATI situation.

    Not for nothing, but I'll hold on to my 7900GT until there are actually DX10 games on the retail shelves. And even then, I'll wait for the 8900 or 2950 and subsequent driver releases.
  38. Ultimately, no, the colour does not affect the loss (or gain) of heat to a significant degree.

    All I have to go on is what I have learned at school. Namely that white/reflective object reflect light (including infra-red) and don't radiate much. Dark/non-reflective objects will absorb and radiate heat (infra-red) much better. Thus a black radiator would be much more efficient than a white radiator.

    I understand that a black (visibly) object is not nessecarily black in the infra-red spectrum, but I was led to believe that a black object is more likely to be.

    Explain this to me then: we did an experiment where we had 2 brass cups, one painted black, one unpainted. Both cups were filled with hot water at the same temperature and were monitored. The black cup cooled more quickly than the unpainted.
  39. Quote:
    im confused. i thought color only mattered because black absorbs a wider range of light. but if the card is in the case what light can it really abosorb. the only light getting in really should be out side of the range that we see as color and should not depend on what we see as color. unless you have a light inside of your case i cant see how the color would really matter. maybe if the case had a plastic side, but you would still need to have a lot of direct light wouldent you?


    That's the point, unless direct sunlight is hitting the inside of your case it doesn't matter a crap what color the components are. LOL
  40. Quote:
    Explain this to me then: we did an experiment where we had 2 brass cups, one painted black, one unpainted. Both cups were filled with hot water at the same temperature and were monitored. The black cup cooled more quickly than the unpainted.


    My guess is that there were other variables, or maybe you used a black magic marker to color a styrofoam cup. Styrofoam insulates, however a black magic marker may have eaten away at the foam a bit, causing the insulation value to decrease. Ever try superglue and styrofoam? The two do not mix. The styrofoam melts. There is an acidic or caustic property/component in the glue. It's likely that the permanent marker has a similar component. :-)
  41. Quote:
    im confused. i thought color only mattered because black absorbs a wider range of light. but if the card is in the case what light can it really abosorb. the only light getting in really should be out side of the range that we see as color and should not depend on what we see as color. unless you have a light inside of your case i cant see how the color would really matter. maybe if the case had a plastic side, but you would still need to have a lot of direct light wouldent you?


    That's the point, unless direct sunlight is hitting the inside of your case it doesn't matter a crap what color the components are. LOL

    Wrong. It doesn't matter. But the reason it doesn't matter is because colour has only a small impact on absorbtion and radiation at such low temperatures. There IS light in a case, even if the case is sealed and the computer is off, because the components are radiating infra-red. As they heat up, they radiate more infra-red. Much more significant is the material of the component.
  42. Quote:
    Explain this to me then: we did an experiment where we had 2 brass cups, one painted black, one unpainted. Both cups were filled with hot water at the same temperature and were monitored. The black cup cooled more quickly than the unpainted.


    My guess is that there were other variables, or maybe you used a black magic marker to color a styrofoam cup. Styrofoam insulates, however a black magic marker may have eaten away at the foam a bit, causing the insulation value to decrease. Ever try superglue and styrofoam? The two do not mix. The styrofoam melts. There is an acidic or caustic property/component in the glue. It's likely that the permanent marker has a similar component. :-)
  43. Quote:
    Explain this to me then: we did an experiment where we had 2 brass cups, one painted black, one unpainted. Both cups were filled with hot water at the same temperature and were monitored. The black cup cooled more quickly than the unpainted.


    My guess is that there were other variables, or maybe you used a black magic marker to color a styrofoam cup. Styrofoam insulates, however a black magic marker may have eaten away at the foam a bit, causing the insulation value to decrease. Ever try superglue and styrofoam? The two do not mix. The styrofoam melts. There is an acidic or caustic property/component in the glue. It's likely that the permanent marker has a similar component. :-)

    Which part of the "they had 2 brass cups" you didn't understand? ;)
  44. Quote:
    Shame it looks so ugly, but I suppose when it's hidden in your machine..... :) Just hate red cards, even though I've owned one or two.... what's wrong with black?

    I have a 100% clear case. So, I'll rather have a UV reactive plutonium green card to match the UV reactive water cooling fluid. :twisted:
    It's nice getting a sun-tan staying indoors.
  45. Quote:
    JonnyDough, please read mattc's post again before you make that a triple posting. No offense, but you making many generalizations and they are for the most part incorrect. Please see also this post on a similar issue a few months ago...

    JumpingJack and Hotfoot discuss radiative heat transfer and emissivity


    If people go read the thread stagepuppy linked here, I think you'll have your questions answered. I started out this discussion with Jack on the assumption that radiating property is equal to absorbtion property. This is not true, as emmisivity is a function of wavelength. What you really want is a material that radiates well at the wavelength corresponding to the operating temperature. Fortunately for this purpose, most materials radiate quite well in the infra-red spectrum.

    However, I also did a back-of-the-envelope calculation in that thread that shows that the contribution to cooling that radiative heat transfer provides, even for fairly large surface areas, is very small compared to conduction and convection at stadard computer operating temperatures. Even if the colour of your heatsink did play a significant role in it's emmisivity for the infra-red spectrum, it wouldn't matter much.
  46. Quote:
    @ the red vs black 2900 debate:

    even if they make a black version, it will still end up being red, because it will be glowing from all the heat it puts off :P


    Maybe they made it red to hide the fact it glows red with heat! :D
  47. Quote:
    No particles actually have a property of "color"


    Unlucky, quarks and gluons have a property of "colour" 8)


    Back to the heat thing though, when i was in school, as well as the cups thing another one was filling a brass cube with hot water and the different sides were painted differently and you could clearly feel the difference between the sides by holding your hand near them. It's probably not such a tiny difference as people make out, but it's still unimportant compared to other factors (especially since the pcb shouldnt be that hot anyway) unless you need every last bit of cooling you can get, and black pcb may be inherently more expensive.
  48. The only thing worrying me is that it loose quite badly in Dx10 benches. I tought that the R600 was gonna get back it luster when Dx10 games were gonna come out. Anyway, I won't get any Dx10 card until Dx10 games start to pop out in sufficient numbers.
  49. Quote:
    The Hitchikers Guide to the Galaxy has brougt us the babelfish, also on the net.

    Try typing the url into the translator found at babelfish.altavista.com I've found, in this case, that the Traditional Chinese to English does a good job :)

    Enjoy!


    its still way too broken ...I aint dumb, I tried that before asking..

    Quote:
    The only thing worrying me is that it loose quite badly in Dx10 benches. I tought that the R600 was gonna get back it luster when Dx10 games were gonna come out. Anyway, I won't get any Dx10 card until Dx10 games start to pop out in sufficient numbers.

    you mean lost planet?
    because in call of juarez most of the time its at the GTX levels or a bit faster o_O
Ask a new question

Read More

Graphics Cards Gtx GPUs Graphics Product