• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

X1950 XTX review/performance

That DDR4 must be running with very slack timings, or the memory controller needs tweaking.

I'm sure an X1900XT-X with it's DDR3 clocked to 2Ghz would easily beat the X1950XTX. An extra 350Mhz ram speed should add more than a few FPS to most games.
 
dave2k2005 said:
WTH dude, you have an X1950?!?
lol, no mate I grabbed the screen from this review: Driverheaven

Sir Random said:
I'm sure an X1900XT-X with it's DDR3 clocked to 2Ghz would easily beat the X1950XTX

IIRC one of the leaked reviews clocked an x1900xtx's memory to 2GHZ and the card scored 200+ more 3dmark05 points than a stock x1950xtx, not that impressive really.
 
Gibbo said:
Well I am a little annoyed as was expecting more than this.

One on one against the 7950 GX2 the X1950 XTX is definetely not as quick in most games. Though its not miles behind and I've got a feeling there will be some nice improvements to come as at the moment that super-fast memory seems to be having practically no effect.

Though what I have noticed from the benchmarks is that X1950 XTX Crossfire seems as quick or quicker than Quad SLi, plus the fact more stable I'd imagine too and a lot less heat produced too. :)

Best bet is to go X1900 Crossfire, you can now do it for practically £400 if you purchase from OcUK as we got the best deals. From looking at all these benchmarks X1900 Crossfire is as quick as 7950 GX2 in Quad SLi but Crossfire is definetely more stable and works with all games. :)



If it was my money I'd be buying HIS X1900 Crossfire and a HIS X1900 XT, that why you get good cooling, silence and Quad SLi performance levels for around £450 Inc. VAT. :D :D

Gibbo

Previously You said You had seen How good the X1950xtx was ! and was happy to join in with the hype, Just explain to Us please how and why this has changed in a couple of Days.. :confused:

You have a responsibility to be more accurate than the rest of Us on this Forum.

How did You get on with Your X1950XtX ?

Thanks
 
Unless the Fan and memory are super efficient! How can the same Core at the same speed on virtually the same Board Use less Power ?

I too had heard rumors of that...more miss-information ;)
 
robgmun said:
So does this use less power then the 1900TX as roumered to be?

The x1900xt doesn't use more than 1,5w per memory MAX. 1,5wx8=12w, so there really isn't much to save by changing to gddr4. X1950XTX is supposed to use 30% less power for the memory, so only a few watts there. So in the end, X1950XTX uses about 2-3w less than X1900XT :p . But quite a lot less than X1900XTX though, not sure what they've done :confused:
 
Sir Random said:
That DDR4 must be running with very slack timings, or the memory controller needs tweaking. .

Fs speculate that the latencies are higher to avoid more heat/power consumption.Maybe newer drivers aimed at these cards might improve things.
I imagine that 1900xts will be avaliable for a while yet,but after this you have a 256mb 1950 or you have to go to xtx version for 512mb.
Enthusiasts are more likely to go for 512mb,so it looks like ati want us to dip into our pockets again.
 
with only 16 pipes and 650mhz clock speed, the x1950xtx doesn't seem to have enough fillrate to make use of its addition memory bandwidth gain. seems like its fillrate limited and not memory bandwidth limited.
 
Cyber-Mav said:
with only 16 pipes and 650mhz clock speed, the x1950xtx doesn't seem to have enough fillrate to make use of its addition memory bandwidth gain. seems like its fillrate limited and not memory bandwidth limited.

Can't really see that X1900xt's seem to respond better with OC memory than core on the tests I have done
 
Cyber-Mav said:
with only 16 pipes and 650mhz clock speed, the x1950xtx doesn't seem to have enough fillrate to make use of its addition memory bandwidth gain. seems like its fillrate limited and not memory bandwidth limited.

I agree, but it would be interesting if you say, increase core speed to 750mhz, and then see the difference :) . I'm also eager to see if the x1950xtx overclocks better or worse than x1900xt/xtx(they must have done something to keep power use down...)
 
JAKUS said:
Gibbo

Previously You said You had seen How good the X1950xtx was ! and was happy to join in with the hype, Just explain to Us please how and why this has changed in a couple of Days.. :confused:

You have a responsibility to be more accurate than the rest of Us on this Forum.

How did You get on with Your X1950XtX ?

Thanks

I had seen the presentation from ATI with game benchmarks and video's that was looking great. I think however its all down to the settings used which gave them good results and NV's card sub par results from what it would normally get. Typical propaganda and both NV and ATI do it all the time.

Though I've been playing with my X1950 XTX today and I am quite impressed. Yes its not much quicker than an X1900 XTX, but it is quicker, quieter and cooler. Plus it cost the same as what the XTX was.

I will post some benchmarks tomorrow but in my Conroe rig it scored:-
3D Mark 2006 - 7200 Marks
3D Mark 2005 - 14001 Marks
3D Mark 2003 - 24500 Marks
3D Maek 2001 - 57000 Marks

So the results are very good for a single card. I overclocked it too 680MHz core, 2.1GHz memory and the Conroe CPU was running at 3.75GHz to obtain those results.

For a single card solution I think the results are quite impressive and even though its only a little quicker than the X1900 XTX it is a damn good card at a good price. Don't forget the X1900 XTX is now going EOL and its only so cheap now because were clearing it out.

Still I won't deny that an X1900 XT 512MB is still a fantastic buy, especially the HIS models due to the fact they already have a cooler that is practically identical to the one of the X1950, albeit the one on the X1950 is slightly better. :)
 
sablabra said:
The x1900xt doesn't use more than 1,5w per memory MAX. 1,5wx8=12w, so there really isn't much to save by changing to gddr4. X1950XTX is supposed to use 30% less power for the memory, so only a few watts there. So in the end, X1950XTX uses about 2-3w less than X1900XT :p . But quite a lot less than X1900XTX though, not sure what they've done :confused:

GDDR4 consumes less power and runs far cooler. That is power savings, also the board is ever so very slightly differently laid out, so possibly a few tweaks here and there. But my sample is definetely cooler than what an X1900 XT runs out, for example I can hold onto the X1950 XTX whilst running at not get burnt even though its still hot. :)
 
OK, I haven't heard anything back from the webnote I sent requesting to change my x1950 order to a HIS x1900xt, so that amongst other things, I've decided to keep the order alive. For reasons stated in another post, I pratically swapped my borked x1900xtx AND received £40 for the x1950xtx, so its not like it cost me a penny. For sure I could have saved £100 buying the HIS XT but this is my last g-card upgrade before the 2nd gen DX10 hit, so. . . bah. Next system upgrade will be a quad core kentsfield which most of us know is due to hit in a couple of months, for now I'll have to make do with my little old 4000+. I'll be sure to(try)clock the bejeezus outta the x1950 and will post 3D benchys when it arrives :) Cheers.
 
Gibbo said:
I had seen the presentation from ATI with game benchmarks and video's that was looking great. I think however its all down to the settings used which gave them good results and NV's card sub par results from what it would normally get. Typical propaganda and both NV and ATI do it all the time.

Though I've been playing with my X1950 XTX today and I am quite impressed. Yes its not much quicker than an X1900 XTX, but it is quicker, quieter and cooler. Plus it cost the same as what the XTX was.

I will post some benchmarks tomorrow but in my Conroe rig it scored:-
3D Mark 2006 - 7200 Marks
3D Mark 2005 - 14001 Marks
3D Mark 2003 - 24500 Marks
3D Maek 2001 - 57000 Marks

So the results are very good for a single card. I overclocked it too 680MHz core, 2.1GHz memory and the Conroe CPU was running at 3.75GHz to obtain those results.

For a single card solution I think the results are quite impressive and even though its only a little quicker than the X1900 XTX it is a damn good card at a good price. Don't forget the X1900 XTX is now going EOL and its only so cheap now because were clearing it out.

Still I won't deny that an X1900 XT 512MB is still a fantastic buy, especially the HIS models due to the fact they already have a cooler that is practically identical to the one of the X1950, albeit the one on the X1950 is slightly better. :)

Thanks for the reply
 
Gibbo said:
GDDR4 consumes less power and runs far cooler. That is power savings, also the board is ever so very slightly differently laid out, so possibly a few tweaks here and there. But my sample is definetely cooler than what an X1900 XT runs out, for example I can hold onto the X1950 XTX whilst running at not get burnt even though its still hot. :)

It does use a bit less power, yes. But still, perhaps the most power hungry memory on any card(with gddr3) is samsung 1.1ns memory(7800gtx 512mb had them). They used about 1,6w per memory(1,6x8=12,8). Therefore, for the memory itself, 30 to 40, even 50% wouldn't make much difference! So they must have done some more changes to make it use 20-30w less than x1900xtx
 
Back
Top Bottom