Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Radeon HD 2400 XT 128MB vs Radeon 9800 Pro 128MB
- Thread starter Archmagination
- Start date
- Sort by reaction score
I had a 256MB Radeon 9800 Pro in my (now deceased) PowerMac G5. Based on published reviews of the HD 2400 XT and the HD 2600 Pro, my frame rates in Call of Duty 2 weren't really any worse. It was the major reason I didn't pick up one of the new iMacs.
I'm guessing they did, now that they know people have the ability to install software to read the gpu bios...
I'm thinking of this benchmark test:
http://www.extremetech.com/article2/0,1697,2151752,00.asp
The Radeon HD 2600 Pro is getting an average FPS of 22 at 1280x1024 (no AA/AF), 14 FPS at 1280x1024 (4xAA/8xAF), and 14 FPS at 1920x1200 (no AA/AF). People are going to want to kill me for saying this because I can't give exact benchmark numbers, but I played the game through on my Radeon 9800 Pro equipped G5 at 1280x1024 and subjectively it would have felt in a similar range of performance (ie: between 15 to 25 FPS, depending on the scene and what was happening).
It's playable (if you don't have high expectations...and I fully realise the iMac isn't aimed at hard core gamers), but I was really hoping for a substantial jump in performance after 4 years.
Bear in mind that all those benchmark tests are in Windows Vista. Games typically lose 10 - 15% of their frame rates when compared to running in WinXP.
2400XT isn't all that fast
At 1024 x 768 or 1280 x 1024, the formerly ToL 9800XT should be roughly equivalent to the 2400XT, depending on the age of the game. The 2400XT is roughly the performance of the x1600 pro, which is roughly the performance of the 9800XT. Given that neither chip can push frames much beyond those resolutions, higher res 3D is a moot point.
The reason why the 9800 XT beats out 2400 XT as resolutions rise is because of the 4x wider bandwidth. The 9800 XT has a 256 bit bus vs the 2400 XT's 64 bit bus. The wider the bus the more data that can go through the processor. In this case, the 2400 XT's bus is the limiting factor, whereas the 9800 XT's processor is the limiting factor in the other.
For what it's worth, my quicksilver Dual 1 GHz with a Radeon 9700 pro benchmarks Quake 4 at ~25 fps. CNet today benched the 20" iMac with the 2600 pro at 39 fps. Considering the 2600 pro is a significantly faster card than the 2400 pro, I wouldn't be surprised to see the 2400 pro post scores that are about the same as my old (but admittedly heavily upgraded) tower.
ATI and nVidia's top end will always be 4-10x faster than their bottom end chips, due to a neutering of the chip (1/3 as much processor power in the GeForce 8600 vs the 8800, for example) and cutting down of memory buses (the 8800 has a 384 bit memory bus, the 8600 has a 128 bit memory bus, the 8400 has a 64 bit memory bus). My numbers are for nVidia, but ATi is comparable - actually a little worse in this generation.
At 1024 x 768 or 1280 x 1024, the formerly ToL 9800XT should be roughly equivalent to the 2400XT, depending on the age of the game. The 2400XT is roughly the performance of the x1600 pro, which is roughly the performance of the 9800XT. Given that neither chip can push frames much beyond those resolutions, higher res 3D is a moot point.
The reason why the 9800 XT beats out 2400 XT as resolutions rise is because of the 4x wider bandwidth. The 9800 XT has a 256 bit bus vs the 2400 XT's 64 bit bus. The wider the bus the more data that can go through the processor. In this case, the 2400 XT's bus is the limiting factor, whereas the 9800 XT's processor is the limiting factor in the other.
For what it's worth, my quicksilver Dual 1 GHz with a Radeon 9700 pro benchmarks Quake 4 at ~25 fps. CNet today benched the 20" iMac with the 2600 pro at 39 fps. Considering the 2600 pro is a significantly faster card than the 2400 pro, I wouldn't be surprised to see the 2400 pro post scores that are about the same as my old (but admittedly heavily upgraded) tower.
ATI and nVidia's top end will always be 4-10x faster than their bottom end chips, due to a neutering of the chip (1/3 as much processor power in the GeForce 8600 vs the 8800, for example) and cutting down of memory buses (the 8800 has a 384 bit memory bus, the 8600 has a 128 bit memory bus, the 8400 has a 64 bit memory bus). My numbers are for nVidia, but ATi is comparable - actually a little worse in this generation.
I really find it sad that a 4-5 year old top of line card is better than than the current midrange cards... its really pathetic. Looks like I will have to spring for the $1499 Imac.
the reviews are not all in agreement. Some seem to say that the 2000 series is a pretty good competitor to nvidea's 8000 series for a decent price, while others bemoan the 2000 series as essentially hopeless.
But those are just commentary. Based on my own review of the benchmarks, I'd say that the 2000 series cards are disappointing, but not exactly the disaster that many here have been suggesting since the imacs were released. It's a shame that there aren't better options, but I would suggest that people calling for the nvidea 8000 series take a look at the benchmarks in this thread - it ain't all that much better.
But those are just commentary. Based on my own review of the benchmarks, I'd say that the 2000 series cards are disappointing, but not exactly the disaster that many here have been suggesting since the imacs were released. It's a shame that there aren't better options, but I would suggest that people calling for the nvidea 8000 series take a look at the benchmarks in this thread - it ain't all that much better.
X1600 came out in November of 2005.
The 9800 came out in May of 2003
The 7600 GT came out in March of 2006
The X1600 was included with the first Intel iMac which came out in January of 2006.
In September 2006, they added the 7600GT option.
http://www.legitreviews.com/article/547/7/
Sure the game is X3 but the X300 (PCI-E 9600) beats it... so yeah I think its safe to say the 9800 Pro is better and it makes sense even, the X1600 was barely like a 6600GT and the 6600GT was a little better than the 9800 Pro, the HD 2400XT is considerably worse than the X1600.
Also those benchmarks show the 7600GT is quite a bit better than the HD 2600XT (except in R6 Vegas) and a LOT better than the HD 2600 Pro (even in R6 Vegas) so yeah... its disappointing this so called graphics upgrade is actually a downgrade except for the 20" model with the HD 2600.
Sure the game is X3 but the X300 (PCI-E 9600) beats it... so yeah I think its safe to say the 9800 Pro is better and it makes sense even, the X1600 was barely like a 6600GT and the 6600GT was a little better than the 9800 Pro, the HD 2400XT is considerably worse than the X1600.
Also those benchmarks show the 7600GT is quite a bit better than the HD 2600XT (except in R6 Vegas) and a LOT better than the HD 2600 Pro (even in R6 Vegas) so yeah... its disappointing this so called graphics upgrade is actually a downgrade except for the 20" model with the HD 2600.
It's not uncommon for a game to play particularly well on a certain card. X3 seems to play so well on the x300 that I'd think they had the hardware in-house to work with it. It's not so much a negative for the 2000 series.
As far as the benchmarks go, that is a desktop x1600, not the one in the previous iMac.
I do consider it a graphics update - these cards have newer features and will be able to play new games better than the x1600 mobile the iMac had before, albeit at lower resolutions...