-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
Radeon 2950pro
Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Keysplayr
Elite Member
Sounds good. But how many here think the 256-bit bus will have little to no effect on the performance of this product? I mean, when ATI announced the R600 would have a 512-bit bus, everyone drooled. Now, cutting that in half doesn't seem like a big deal. ::shrugs::
If AMD can compensate with higher clocked memory, that's great.
If AMD can compensate with higher clocked memory, that's great.
The 512-bit memory controller wasn't a bad idea in and of itself. The problem is that the R600 doesn't perform AA efficiently and fails to take advantage of the wider bus. It doesn't make any sense...an 8800 card with a 384-bit memory controller is *better* with AA than the ATI cards with a 512-bit bus. 😕
Keysplayr
Elite Member
Judging from the specs, and if indeed the 2900XT would not have been memory starved at 256-bit memory interface, the 2950pro will indeed be faster than a 2900XT due to it's increased core clock. But again, only if it is not bandwidth starved with a 256-bit bus. You can't really go by price point because as I said, with these modifications, the 2950pro will be dirt cheap to manufacture compared to a 2900XT.
My understanding of this situation is not quite clear either, as you can see.
This is a truly smart move by AMD. Bout time eh? Better late than never.
I'm taking these specs with a huge grain of salt.
Unless there is a new flagship to go with this "pro", there's no way they'd sell their highest performing card at $250. They might as well just drop out of GPU business. Not to mention Nvidia will undoubtedly wipe the floor with this offering and charge $1000 for their high end.
I fully expect this 2950 Pro to sport a reduced number of shaders and TMUs.
Unless there is a new flagship to go with this "pro", there's no way they'd sell their highest performing card at $250. They might as well just drop out of GPU business. Not to mention Nvidia will undoubtedly wipe the floor with this offering and charge $1000 for their high end.
I fully expect this 2950 Pro to sport a reduced number of shaders and TMUs.
I made the same mistake...
The info on R680 is pretty sketchy. Basically people are saying that ATI/AMD decided to fix the AA issue and give R680 dedicated AA hardware. The other thing they're changing is the shaders. They will be the same type that the 8800 cards use, and there will perhaps be more of them at a higher clockspeed. AFAIK they're keeping the 512-bit memory bus.
Does anyone have a release date for R680? I thought it was supposed to come out this summer...is there any chance that it will be released along with Barcelona?
cmdrdredd
Lifer
bryanW1995
Lifer
yes, but intel chips cost 266 in batches of 1000 for newegg et.al to purchase. they are currently having a field day marking them up b/c there is nothing from amd OR intel that competes with it. the radeon 2950 pro will be a middle-of-the-road card at best so is unlikely to command a strong premium. Also, I don't believe that video cards are priced in batches of 1000 like cpus.
bryanW1995
Lifer
I hate to get my hopes up, but this might be a step in the right direction for ati/amd.
I'm ready to give up hope. People are now asking what the 'point' of Fusion is. Apparently the chip is over a year away, and won't even make a revolutionary difference.
You look at ATI's roadmaps, and they have an OK looking midrange card coming (2950pro) and perhaps an OK high end card. The trouble is that both will be way too late to market...AGAIN AMD is late!
Unless Barcelona is a huge success, my guess is that AMD is pretty much finished. If Samsung buys them, expect them to make chips for TV sets and maybe videogame consoles. Forget about the X86 CPU market and the discreet graphics market.
That's way too late for this chip to have any kind of impact IMO.
It will probably perform along the lines of the 320mb 8800GTS on a very very good day. Nvidia will have had its new high end gear out for 2 months before this card even hits market, which will make it, at best, 1/4 the speed of the G92. So, if it sells for $249, then Nvidia might as well charge around $1000 for the G92. :thumbsdown:
shadowofthesun
Member
R680 on scalar shaders? Come on. I seriously, seriously doubt this story.
First of all, R600's vector shaders are far from underpowered. Both G80 and R600's architectures have a long way to go before maturity. I suspect at least one or two more generations on each before we see drastic changes to either. With driver revisions, lower nodes, refined processes, and some fine-tuning, both these architectures have alot more headroom before either of them needs to be replaced.
Secondly, coding drivers for these two different architectures would be brutal. Imagine ATI/AMD supporting 3 different architectures at once (X1xxx, R600, R680). Say goodbye to monthly driver releases.
Finally, assuming that they do make this huge jump, the reason would be that they believe that nVidia's scalar shaders perform better (which isn't an assumption anyone should be making this early in the DX10/SM4 generation). If they do think so, they will probably release an entirely new lineup based on this updated architecture- not R680, but R700 with corresponding mid and low-end parts. There's no reason for them to conceede that their opponent's architecture is better and only update one part of their lineup to be competitive.
Honestly, if AMD plays their cards right, the current vector shader architecture will last for some time. I doubt they would give up so soon.
First of all, R600's vector shaders are far from underpowered. Both G80 and R600's architectures have a long way to go before maturity. I suspect at least one or two more generations on each before we see drastic changes to either. With driver revisions, lower nodes, refined processes, and some fine-tuning, both these architectures have alot more headroom before either of them needs to be replaced.
Secondly, coding drivers for these two different architectures would be brutal. Imagine ATI/AMD supporting 3 different architectures at once (X1xxx, R600, R680). Say goodbye to monthly driver releases.
Finally, assuming that they do make this huge jump, the reason would be that they believe that nVidia's scalar shaders perform better (which isn't an assumption anyone should be making this early in the DX10/SM4 generation). If they do think so, they will probably release an entirely new lineup based on this updated architecture- not R680, but R700 with corresponding mid and low-end parts. There's no reason for them to conceede that their opponent's architecture is better and only update one part of their lineup to be competitive.
Honestly, if AMD plays their cards right, the current vector shader architecture will last for some time. I doubt they would give up so soon.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 24K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
