Graphics Card Question
Gaylord
replace the
Geforce 2 Mx400 that he currently has in for something a little better.
The best I can see for the money that he has is either a 128meg Geforce 5200
or a Radion 128meg 9200 which is the best one for him, he plays games mostly
if that makes any help.
Thanks
Morgan
Just don't buy a 5200 as the performance is not as good as the similar
priced ATI.
The only other consideration is a Ti4200 if you can get one at a good price.
They might be from an 'older' generation of cards but they perform superb on
today's games.
--
Regards
Morgan
My noisy drive is noisy no more...
www.flyinglizard.freeserve.co.uk
Hawkey
"Gaylord" <mych...@ithurting.mecom> wrote in message
news:bpnb0h$rjr$1...@newsg2.svr.pol.co.uk...
Phöčnix
> ATI 9xxx - just change the xxx to the amount of money that he can
> afford. Just don't buy a 5200 as the performance is not as good as
> the similar priced ATI.
>
I would be interested in seeing appropriate support for this statement.
I looked at this recently and the 5200 shades the 4200 in benchmarks I saw.
I only saw one review that benchmarked them together but though I just had a
quick look I can't find it again.
I wasn't looking for a gaming card so I didn't register the specifics.
Neither is a 'gamers card' per se.
What is your friends actual budget (inc vat and delivery)? Does he need TV
out, DVI or other dual monitor support?
Answer those two and I'm sure someone will be able to recommend a specific
'buy' that will give him the best gaming performance.
> The only other consideration is a Ti4200 if you can get one at a good
> price. They might be from an 'older' generation of cards but they
> perform superb on today's games.
Different price band (and different league).
ATI 9200 £35
Nvidia5200 £39
Nvidia4200 £70+ which brings in the 9600SE and FX5600 all about 65% increase
in FPS over the two you have been considering.
Trevor Best
I would agree with Morgan, look to the DX8 range of cards, Geforce 4
Ti (not MX) or Radeon 8500, which can be picked up pretty cheaply. For
DX9, the low end cards are just that, low end and would struggle
almost as much as his old MX on DX8 games (most games are still DX8)
although the midrange DX9 stuff is pretty good, Geforce 5600/5700 or
Radeon 9600 if he can stretch to it. The plain vanilla 5600 dioesn't
get much acclaim but I'm happy with mine.
--
A)bort, R)etry, I)nfluence with large hammer.
Phöčnix
> Morgan wrote:
>> ATI 9xxx - just change the xxx to the amount of money that he can
>> afford. Just don't buy a 5200 as the performance is not as good as
>> the similar priced ATI.
>>
>
> I would be interested in seeing appropriate support for this
> statement.
>
obviously**)
> I saw.
Nick Le Lievre
news:bpnise$18h$1...@hercules.btinternet.com...
> was it the sharp edges on the card that killed him?
Hehe I was gonna say and having a very limited budget gfx card has to do
with dying how ? Perhaps one could figure in such poor fps as to induce
suicide ? In the after life he wants a better graphics card.
TMack
"Gaylord" <mych...@ithurting.mecom> wrote in message
news:bpnb0h$rjr$1...@newsg2.svr.pol.co.uk...
He will get no real benefit from DX9 compatibility as there are few games
that use it so far and these will still run on DX8 cards. With low end DX9
cards they may be able to render images using DX9 but they do it so slowly
that games will not perform acceptably. I agree with the advice from others
who have replied - consider good DX8 cards. There are currently 3 used 64MB
Radeon 9000 pro cards in ebuyer's 'blowout deals' section at £50.07 inc VAT
and 10 128MB Radeon 9000 pro cards for £58.24 inc VAT. These will
out-perform the FX5200 and represent very good value. Alternatively look
for a used 8500 (275/275 core/memory) or even 8500LE (250/250 core/memory -
check before you buy as there are some lower spec cards on the market) which
are essentially the same as the 9000 series and which will offer performance
comparable with the FX5200. These have been available on ebay at around the
£40 mark recently although there is only one over-priced example currently
showing.
Tony
TMack
"TMack" <REMOVETHECA...@bigfoot.com> wrote in message
news:bpnoao$1q9631$1...@ID-67499.news.uni-berlin.de...
SNIP!
>Alternatively look
> for a used 8500 (275/275 core/memory) or even 8500LE (250/250
core/memory -
> check before you buy as there are some lower spec cards on the market)
which
> are essentially the same as the 9000 series and which will offer
performance
> comparable with the FX5200. These have been available on ebay at around
the
> £40 mark recently although there is only one over-priced example currently
> showing.
Correction - there are several 8500 and 8500LE cards on ebay - I was only
looking in the 'buy it now section' when I made the original statement.
Tony
TMack
"TMack" <REMOVETHECA...@bigfoot.com> wrote in message
news:bpnoao$1q9631$1...@ID-67499.news.uni-berlin.de...
SNIP!
>Alternatively look
> for a used 8500 (275/275 core/memory) or even 8500LE (250/250
core/memory -
> check before you buy as there are some lower spec cards on the market)
which
> are essentially the same as the 9000 series and which will offer
performance
> comparable with the FX5200.
One further correction (OK - I'll get my coat.........) I have just
remembered that the 9000 series only have one texture pipeline whereas the
8500 series have two which make them a bit faster althought the cards are
similar in other respects (the 9100 and 9200 have 2 texture pipelines). the
9000 pro will still substantially outperform an FX5200 with standard
settings but the FX5200 is better than the 9000 pro with antialiasing
enabled (though why anyone would want AA enabled on a slowish low end card
is another question). Overall the 8500 series are the best bet. Last word
of warning - avoid anything that has 'lite', SE or similar in its name as
this usally means substantially reduced spec and performance.
Tony
Phöčnix
This comparison shows similar performance to the FX5200
It comes out worse in the other benchmarks. So is it really worth paying
more for a second hand one than for the other brand new FX5200.
TMack
"Phöčnix" <mindyouownbusiness> wrote in message
news:3fbfd3b0$0$13348$cc9e...@news-text.dial.pipex.com...
The graph on Tom's site shows Radeon 9000 NOT 9000 Pro. Try these
http://www.hothardware.com/hh_files/S&V/aiw_9000_pro(3).shtml
http://www.hwupgrade.it/articoli/879/1.html
The 9000 Pro scores consistently higher than the FX5200. However, I would
still go for an 8500 which is substantially better than either.
Tony
Phöčnix
I was referring to the 8500 you recommended.
> Try these
> http://www.hothardware.com/hh_files/S&V/aiw_9000_pro(3).shtml
These don't show the 8500.
> http://www.hwupgrade.it/articoli/879/1.html
>
Nor does this. But it does bear out what I said about the FX5200 v Radeon
9200 earlier in the thread. Thanks
> The 9000 Pro scores consistently higher than the FX5200.
At a comparable price?
> However, I
> would still go for an 8500 which is substantially better than either.
>
Not according to the comparison I posted for you.
> Tony
TMack
"Phöčnix" <mindyouownbusiness> wrote in message
I think you need to look again at the link that you posted - it shows the
8500 out-performing the FX5200 at all resolutions in UT2003 if AA is not
enabled. There are also some VERY peculiar scores being shown for other
cards in other benchmarks. For example, the Q3 Team Arena scores show the
Geforce MX440 out-performing a Radeon 9500!!!!!!!!! (It has been said that
Tom has sold his soul to NVidia in return for sponsorship of his site.) I
have NEVER seen any other report of an MX440 (a tarted-up Geforce 2) coming
even remotely close to a 9500 so it looks like something is not right here.
Of course NVidia 'optimise' their drivers (in some cases by removing
capability and reducing image quality) for certain games and benchmarks so
it is just possible that these scores are 'genuine'. However, no sane
person would buy an MX440 in preference to a 9500 on the basis of these
benchmarks - the 9500 (basic non-pro version) is an infinitely superior
card. (In 3DMark 2003 MX440 has a top score of 425 - I could get around
1400 with my basic non-pro 9500. In AquaMark 3 I was able to get a score of
26,097 with my overclocked but unmodified non-pro 9500 whilst the highest
score for an overclocked MX440 was 12,766)
The 8500 is consistently substantially faster than the 9000 Pro in
benchmarks (unless FSAA is being used - but like I said, why use FSAA with
a lowish end card which is likely to need all the raw speed it can deliver
with newer games). In fact the slower 8500LE is still faster than a 9000
Pro. See the following:
(watch out for wrapping of links)
http://www.tech-report.com/reviews/2002q3/radeon-9000pro/index.x?pg=4
http://www.sharkyextreme.com/hardware/videocards/article.php/3211_2225351__6
http://www.firingsquad.com/print_article.asp?current_section=Hardware&fs_art
icle_id=1211
Now if the 8500 series is faster than a 9000 Pro, which in turn is faster
than an FX5200 then it stands to reason that an 8500 is considerably faster
than an FX5200 (if AA is not being used). You may also wish to use the
'compare projects' facility with 3DMark's online results browser - the top
score recorded ion 3DMark 2001 for an FX5200 is 11,662. The top score for
an 8500 is 16,062. Now the card scoring 16,062 could be a highly
overclocked supercooled freak but there are also loads of scores over 15,000
for the 8500 series. In 3DMark 2003 - top score for 8500 series is 1828
with a total of 10 scores in excess of 1500. For the FX5200 the top score
is 1540, followed by one of 1440 and the rest are below 1300. All quoted
scores are using default setting for the benchmarks.
So which looks like the better bet - an 8500 or an FX5200?????
Tony
Jonathan Eales
graphics card for £19.98. My local Reading branch has a pile of them
mispriced at £14.99.
Best bang for buck in today's graphics cards!
Jonathan
"Gaylord" <mych...@ithurting.mecom> wrote in message
news:bpnb0h$rjr$1...@newsg2.svr.pol.co.uk...
Phöčnix
> "Phöčnix" <mindyouownbusiness> wrote in message
> news:3fbfe3cd$0$13349$cc9e...@news-text.dial.pipex.com...
>> TMack wrote:
>>> "Phöčnix" <mindyouownbusiness> wrote in message
>>> news:3fbfd3b0$0$13348$cc9e...@news-text.dial.pipex.com...
>>>> TMack wrote:
>>>>> "TMack" <REMOVETHECA...@bigfoot.com> wrote in message
>>>>> news:bpnoao$1q9631$1...@ID-67499.news.uni-berlin.de...
>>>>>
>>>>> SNIP!
>>>>>
>>>>>> Alternatively look
>>>>>> for a used 8500 (275/275 core/memory) or even 8500LE (250/250
>>>>>> core/memory - check before you buy as there are some lower spec
>>>>>> cards on the market) which are essentially the same as the 9000
>>>>>> series and which will offer performance comparable with the
>>>>>> FX5200.
>>>>>
Oh behave, outperform? They are more or less identical. A 1% difference in a
benchmark is negligible and hardly a basis for your recommendation nor
refutes my observation that the cards show similar performance in this
report. Stick to the question. It's not a trick.
I deliberately posted the page which showed the 8500 at it's best to avoid
such imature bias. I could have pointed to http://tinyurl.com/w6ws which
shows the 8500 in a worse light.
If you back up your recommendation do so. I am not picking an argument with
you for the sake of it. I am questioning your advice as it is unsupported
and the only comparison I have seen suggests it is poor advice.
As I said I had been looking at the same cards as the OP and would be
interested in decent alternatives. You recommend a secondhand 8500 over a
new FX5200 or R9200.
Don't start getting pedantic or take umbridge just because I asked you to
support your recommendation and/or refute contradictory articles. It does
not support your case and just makes the thread pointless.
>
> The 8500 is consistently substantially faster than the 9000 Pro in
> benchmarks
> (unless FSAA is being used - but like I said, why use
> FSAA with a lowish end card which is likely to need all the raw speed
> it can deliver with newer games).
Let's take that as a given. Ensure FSAA is turned off if using low end
cards.
> In fact the slower 8500LE is still
> faster than a 9000 Pro. See the following:
> (watch out for wrapping of links)
> http://www.tech-report.com/reviews/2002q3/radeon-9000pro/index.x?pg=4
>
http://www.sharkyextreme.com/hardware/videocards/article.php/3211_2225351__6
>
http://www.firingsquad.com/print_article.asp?current_section=Hardware&fs_art
> icle_id=1211
>
> Now if the 8500 series is faster than a 9000 Pro,
That would seem the case.
> which in turn is
> faster than an FX5200
Not so sure. Having shown your true colours as someone who has taken a
position and is prepared to mislead/misreport I took another look at your
previous link. The only other one apart from Tomshardware that links your
thread together. Some of these show the FX5200 outperforming the 900 pro by
10% or so. You seem to have linked to the one that supports your case and
ignored the others. You can hardly rely on this as outright proof.
> then it stands to reason that an 8500 is
> considerably faster than an FX5200 (if AA is not being used).
This may or not be the case. Your (biased) logic is not convincing. If you
had shown yourself to be objective I may have been more inclined to agree.
That the 8500 may be better than the FX5200 I am inclined to concede. I
reserve judgement on whther that difference is considerable or not.
> You
> may also wish to use the 'compare projects' facility with 3DMark's
> online results browser - the top score recorded ion 3DMark 2001 for
> an FX5200 is 11,662. The top score for an 8500 is 16,062. Now the
> card scoring 16,062 could be a highly overclocked supercooled freak
> but there are also loads of scores over 15,000 for the 8500 series.
> In 3DMark 2003 - top score for 8500 series is 1828 with a total of 10
> scores in excess of 1500. For the FX5200 the top score is 1540,
> followed by one of 1440 and the rest are below 1300. All quoted
> scores are using default setting for the benchmarks.
>
You have shown yourself to cherry pick articles and string them togther to
support your arguments. I would not be surprised if someone could not do
similary and show a very different picture.
Given your bent, I am inclined to say statistics can show what you want them
to show. However I would say that if a sufficient number of comparable
systems with non-oc cards indicate the 8500 to consistently and
significantly outperform the FX5200 across the benchmarks then I would be
inclined to agree with your hypothesis.
> So which looks like the better bet - an 8500 or an FX5200?????
>
I would be surprised if it is not the 8500. But whether difference is enough
to support your view that a secondhand 8500 would be a better purchase than
a new FX5200 or R9200 I am unconvinced, but it is certainly not
substantiated here.
> Tony
TMack
"Phöčnix" <mindyouownbusiness> wrote in message
Fine - go buy an FX5200 if that makes you happy. Or pay more for a 9200 and
get no advantage over an 8500 if that makes you happy. I am not "cherry
picking articles", I have been trying to give you the best advice that I
can, for free, becuase I initially had some interest in the topic and I
thought your intention in your original article was to get some informed
advice. If you wanted an argument why didn't you just say so?
I certainly haven't been "cherry picking" the 3DMark scores - they are 'on
the record' for anyone to look at - or are Futuremark in on the conspiracy
as well??? If you want to ignore them, fine. If you think I have some
ulterior motive please ignore everything that I have said.
Neither you or nor anyone else will be able to find evidence that the FX5200
performs consistently better across a range of games or represents better
'bang for the buck than a less expensive second hand 8500 because the
evidence simply isn't there. However, the last piece of advice I will give
you is to ask in comp.sys.ibm.pc.hardware.video whether anyone thinks that
the FX5200 represents good value compared with a substantially less
expensive second hand 8500. The FX5200 isn't necessarily a bad choice - it
just isn't as good as some of the alternatives.
Tony
TMack
SNIP
> Fine - go buy an FX5200 if that makes you happy. Or pay more for a 9200
and
> get no advantage over an 8500 if that makes you happy. I am not "cherry
> picking articles", I have been trying to give you the best advice that I
> can, for free, becuase I initially had some interest in the topic and I
> thought your intention in your original article was to get some informed
> advice. If you wanted an argument why didn't you just say so?
SNIP
I have just noticed that you didn't actually start this thread so maybe you
were just looking for an argument when you joined in.
Tony
TMack
"Jonathan Eales" <Jon....@Virgin.net> wrote in message
news:y61wb.5667$4Y6....@newsfep4-winn.server.ntli.net...
> PCWorld has an offer (until 2/12/03) for an ABit 64Mb DDR nVidia MX440
AGP8X
> graphics card for £19.98. My local Reading branch has a pile of them
> mispriced at £14.99.
>
> Best bang for buck in today's graphics cards!
I have to agree with that. It may not be 'state of the art' but at that
price they represent amazingly good value.
Tony
Phöčnix
> Tony
Phöčnix
> FX5200 performs consistently better across a range of games or
> represents better 'bang for the buck than a less expensive second
> hand 8500 because the evidence simply isn't there.
That's rather subjective isn't it. How mcuh better the 8500 may be than the
FX5200 may be quantifiable under certian circumstances but the price at
which a seondhand 8500 without a warranty represents good value is somewhat
subjective.
> However, the last
> piece of advice I will give you is to ask in
> comp.sys.ibm.pc.hardware.video whether anyone thinks that the FX5200
> represents good value compared with a substantially less expensive
> second hand 8500.
Where is this source of substantially cheaper 8500s. Certainly not E-bay
which is what you suggested.. The one you referred to yesterday sold for £51
delivered. This is *significantly* more than the two cards the OP
was asking about can be sourced for.
> The FX5200 isn't necessarily a bad choice - it
> just isn't as good as some of the alternatives.
>
I wonder if the OP is convinced?
> Tony
Tony Houghton
Jonathan Eales <Jon....@Virgin.net> wrote:
> PCWorld has an offer (until 2/12/03) for an ABit 64Mb DDR nVidia MX440 AGP8X
> graphics card for £19.98. My local Reading branch has a pile of them
> mispriced at £14.99.
They've discontinued that offer. I tried to get one for my
brother-in-law.
--
My real address is in the Reply-To and includes the .nospam.
See <http://www.realh.co.uk/contact.html> for more reliable contact addresses.
TMack
"Phöčnix" <mindyouownbusiness> wrote in message
> > Neither you or nor anyone else will be able to find evidence that the
> > FX5200 performs consistently better across a range of games or
> > represents better 'bang for the buck than a less expensive second
> > hand 8500 because the evidence simply isn't there.
>
> That's rather subjective isn't it. How mcuh better the 8500 may be than
the
> FX5200 may be quantifiable under certian circumstances but the price at
> which a seondhand 8500 without a warranty represents good value is
somewhat
> subjective.
Ultimately everything is subjective. My argument is that there are cards
available (9000 pro, 9200) that will offer overall better performance when
bought new at a similar price to the FX5200 and there is a card (8500
series) which may be available at lower second-hand prices and which which
may be even better.
> > However, the last
> > piece of advice I will give you is to ask in
> > comp.sys.ibm.pc.hardware.video whether anyone thinks that the FX5200
> > represents good value compared with a substantially less expensive
> > second hand 8500.
>
> Where is this source of substantially cheaper 8500s. Certainly not E-bay
> which is what you suggested.. The one you referred to yesterday sold for
Ł51
> delivered. This is *significantly* more than the two cards the OP
> was asking about can be sourced for.
Trying to change the argument to price instead of performance now? Ebay is
not a shop so prices will vary but I bought an 8500 recently (for a PC I was
building for someone) at a price of Ł39.00 so such bargains can be found
> > The FX5200 isn't necessarily a bad choice - it
> > just isn't as good as some of the alternatives.
> >
> I wonder if the OP is convinced?
I have no Idea - I only offered advice, its entirely up to him whether or
not to take it.
Tony
TMack
"Phöčnix" <mindyouownbusiness> wrote in message
> TMack wrote:
> > "TMack" <REMOVETHECA...@bigfoot.com> wrote in message
> > news:bpqhhi$1q9ja0$1...@ID-67499.news.uni-berlin.de...
> >
Such as? So far I have offered opinions, backed with evidence whilst you
have failed to substantiate many of your insinuations and assertions. If
you think I was gilding the lily with my comments about Toms Hardware site
go and search on google groups and you will find much criticism of Tom for
taking payments from NVidia - hence my scepticism about the validity of some
his benchmarks.
Tony