Kyro beating geforce2 pro and Ultra in new review

[PotNoodle]

RobsTV,

?I think nam ng sums up these reviews nicely in the other thread:?

Errr ? I think he was using them as an illustration that he thinks something is wrong with th reviewers scores?

SlimHarpo,

? Non-tilers (GeForces) don't show anywhere near that kind of cpu dependency (except, of course, at low resolutions, which is not an issue with the video card).?

the difference from a 700Mhz Duron to the high end machine they were using is fairly substantial; I don?t know what the difference is between the titles tested with a GTS, but it wouldn?t surprise me to hear that it would be fairly similar. Theing become different when T&L is in operation though as GTS?s have it, and KYRO?s don?t ? this would be quite a leveler.

? The way it looks to me right now, the Kyro might be a good buy for someone with a 1.4 GHz Athlon or P4, but is not a good buy at all for someone with an 800 MHz P3 or Duron (they should just buy an MX for $65 and get similar frame rates, or spend $140 on a GF GTS/PRO, and get much faster frame rates).?

Ummm ? aren?t Duron?s more comparable to Celerons?

I?m running a PIII 733 and a KYROII and things are fine.

 

[Teasy]

<<<Why is the Kyro so slow on the 700 MHz system in the hardware pub review, then?

Non-tilers (GeForces) don't show anywhere near that kind of cpu dependency (except, of course, at low resolutions, which is not an issue with the video card).

The way it looks to me right now, the Kyro might be a good buy for someone with a 1.4 GHz Athlon or P4, but is not a good buy at all for someone with an 800 MHz P3 or Duron (they should just buy an MX for $65 and get similar frame rates, or spend $140 on a GF GTS/PRO, and get much faster frame rates).>>>

Is the Kyro II slow on the celeron system compared to a Geforce MX, GTS. Pro?, how do you know because I don't see the comparisons of those cards in that review. Also you seem to be doing the P3 and Duron a diservice by thinking there as slow as a Celeron, a P3 800 is a long way better then a Celeron 700 and a Duron 700 is also faster then a Celeron 700 not to mention that a Duron 700 will goto 900mhz minimum easily. In Q3 800x600x16 normal settings with my Duron 750 overclocked to 975mhz and Kyro II at defualt speed I get 111fps in demo 1 and thats with HQ sound, with LQ sound I get 120fps in demo 1 and in that review with the same settings on a Celeron 700 in demo 1 he's getting 62fps, my system is a budget system and look how great the Kyro II performs, infact his scores look wrong since his T-Bird 1.33ghz is getting a lower score with the Kyro II then I get on a Duron 975mhz.

 

[Mingon]

Teasy you are comparing an overclocked Duron to a standard duron, the review featuring cpu-speeds was at aceshardware and showed that a 667 celeron blows

 

[Ironduke]

<< Funny how powervr2 counters Kyro2 UT scores with a UT score using 640x480x32.
Well if 640x480x32 is what he wants, then here you go, (from that same review:
Kyro2 loses everyone.
Even GF2 MX is twice as fast as Kyro 2, yes &quot;MX&quot; BEATS Kyro 2
Nope, in Quake3, GTS beats Kyro2
Serious Sam, same results, GTS wins.
I see alot about how great multitexture of Kyro2 is, but:
kyro 2 scores 3rd from bottom, only beating V5 and MX.
MBTR, 2nd from the bottom.

Of course not many run at 640x480, but since powervr2 wants numbers >>




ROFLMAO 640*480

your picking at staws boy and u know it!

 

[Teasy]

<<<Teasy you are comparing an overclocked Duron to a standard duron, the review featuring cpu-speeds was at aceshardware and showed that a 667 celeron blows>>>

The review SlimHarpo was asking about was the hardware pub review not the aceshardware review, I beleive the aceshardware review used MDK2 in there CPU comparison test and its fairly obvious that a Kyro II is going to be slower especially at low res on a lower end system because MDK2 is very HW T&amp;L optimised. None of the benches in the hardware pub review were HW T&amp;L games and so there's no reason for them to be more CPU dependent on a Kyro II then any other graphics card (appart from the DX8 bug which makes the Kyro II more CPU dependent in D3D games that use rendering into textures, but that is now fixed in the new beta DX8.1 620).

 

[SlimHarpo]

<< <<<Why is the Kyro so slow on the 700 MHz system in the hardware pub review, then?

Non-tilers (GeForces) don't show anywhere near that kind of cpu dependency (except, of course, at low resolutions, which is not an issue with the video card).

The way it looks to me right now, the Kyro might be a good buy for someone with a 1.4 GHz Athlon or P4, but is not a good buy at all for someone with an 800 MHz P3 or Duron (they should just buy an MX for $65 and get similar frame rates, or spend $140 on a GF GTS/PRO, and get much faster frame rates).>>>

Is the Kyro II slow on the celeron system compared to a Geforce MX, GTS. Pro?, how do you know because I don't see the comparisons of those cards in that review. Also you seem to be doing the P3 and Duron a diservice by thinking there as slow as a Celeron, a P3 800 is a long way better then a Celeron 700 and a Duron 700 is also faster then a Celeron 700 not to mention that a Duron 700 will goto 900mhz minimum easily. >>



I know from experience that GeForces don't behave like that. At 1024x768 and 1280x1024 in Q3 HQ, the Kyro scores double when going from the Celeron 700 to the Tbird 1.33. You'd be lucky to get a 25% improvement with a GeForce or Radeon in those systems, since either of those cards are mostly fill rate limited in those situations, even with the wimpy Celeron pushing them.

Maybe the Kyro drivers are just unoptimized, which could in fact be a positive thing, given the card's current performance on higher end systems (assuming that Hercules continues to work on them).

And you're right about the Celeron != P3 or Duron thing, of course. I just wasn't paying attention. Your Q3 scores don't sound all that spectacular, though, compared to my Celeron and GF 2 MX system. I still think the Kyro looks to be a lot more cpu dependent than other cards out there, at least at the moment, though it may well just be a driver issue.

[edit]
Actually, I think it's just that those hardware pub benchmarks are f*cked up. The hardware analysis review shows some much more sane looking cpu scaling comparisons here: http://www.hardwareanalysis.com/content/article/1291.14/

 

[uknemesis]

Well I'm using the latest drivers for my gf2 mx and my kyro 2 and here's what I get with quake 3 -

640x480x32bit

K2=98fps
GF2=80fps

800x600x32bit

K2=92fps
GF2=54fps

1024x768x32bit

K2=78fps
GF2=33fps

Going from 640x480 to 1024x768 will cost you 20fps with the Kyro 2 and you'll lose 47fps with the Geforce 2. I think that's pretty clear evidence that I'm cpu limited isn't it?

These are scores with MAX settings for Quake 3 and no I didn't get the results mixed around. These are actualy scores for my Duron 750, even I was surprised esp after reading the early reviews. Maybe my GF2 is really slow, I don't know. Maybe a problem with Quake 3 so I ran another test!

Stars Opengl benchmark

640x480x32bit

K2=141fps
GF2=88fps

800x600x32bit

K2=96fps
GF2=57fps

1024x768x32bit

K2=60fps
GF2=34fps

UNREAL (Direct3D)

640x480x32bit

K2=68fps
GF2=64fps

800x600x32bit

K2=94fps
GF2=55fps

1024x768x32bit

K2=55fps
GF2=43fps

Well anyway, regardless of the above my games look much smoother than with my gf2 and that's the main thing!

Nemesis

 

[Teasy]

<<<I know from experience that GeForces don't behave like that. At 1024x768 and 1280x1024 in Q3 HQ, the Kyro scores double when going from the Celeron 700 to the Tbird 1.33. You'd be lucky to get a 25% improvement with a GeForce or Radeon in those systems, since either of those cards are mostly fill rate limited in those situations, even with the wimpy Celeron pushing them.

Maybe the Kyro drivers are just unoptimized, which could in fact be a positive thing, given the card's current performance on higher end systems (assuming that Hercules continues to work on them).

And you're right about the Celeron != P3 or Duron thing, of course. I just wasn't paying attention. Your Q3 scores don't sound all that spectacular, though, compared to my Celeron and GF 2 MX system. I still think the Kyro looks to be a lot more cpu dependent than other cards out there, at least at the moment, though it may well just be a driver issue.>>>

Well my scores look spectacular compared to the scores in the review were talking about, my Duron 975mhz and Kyro II gets 20fps more then his T-bird 1.33ghz and Kyro II, therefore his benches look far more CPU limited then they should be which I can only assume would also carry over to the Celeron system. It is true that at lower res the Kyro II sometimes isn't as fast as other cards and your right in thinking that thats partly driver related, but not totally, the reason Kyro II excells at high res is because its a TBR and therefore is much more efficient with its available bandwidth, so its not really so much that the Kyro II is slow at low res, its more that it doesn't loose as much when going from low res to high res.

 

[bevancoleman]

Ok, having a look at 3dmark (and making sure I an't look at OC CPUs) I still see the geforce 2 MX scoring a lot higher than his was

1024x768 16bit is ranging from 2346 to 3089 3dmark 2001.
this is after removing any which have high bus speeds (I only looking at 100Mhz) and one really slow one on the last page

1027x768 32bit is ranging from low 2000 (I couldn't be bothered scrolling through all the pages) to 2988


I still saying that his geforce 2 scores are WAY low.

 

[Archknight]

The Exorcist : I hope I haven't insulted you to call you a advocate, and Kyro2 is a great card for its price. $105 Inno3d retail version at pricewatch.

And I done a search on the MX 32 and MX 64 MB (the MX400). The MX400 with 32MB cost as low as $75 for Inno3d and the $79 Inno3d again. I guess the RAMs don't really cost huh. About $26 difference of Kyro2 and MX400(64MB). Sitting bewtween the MX and GTS/Pro, this is quite a competitor.

This turns out to be another typical Kyro2 Vs Geforce2 thread. From ever since the First Kyro2 review thread which passed 300+posts, the subsequence thread doing near to 100 or even more. Come on now if the Kyro2 doesn't have the goods, there would be so much discussion about it.

 

[dfloyd]

Well I disagree, from current benchamrks I have seen put the Kyro II somewhere between the Geforce II GTS and the Geforce III. www.planethardware.com has a good review of the card and it romps on the Ultra in quite a few tests.

 

[PointlesS]

lol....they say no one plays in 640x480....well I do on my crappy k6-2 @ 450 mhz w/128 megs of ram, savage 4...

 

[ThExorcist]

ArchKnight

No offense taken. I believe everybody here is just expressing a little bit of an opinion, along with as I mentioned what seems to be an excellent value as well as a great product.

Actually we are providing a lot of first hand experience and feedback for those that just read these topics and I am sure that is helping many people come to a decision.

 

[bananaboy]

ive played q3a in 640x480x16 on a pentium 133 with a voodoo1... i'd like to see an old junk nvidia card like the riva128 do that. 3dfx was right all along about TnL and FSAA, and looking at where we're at right now I can't see how anyone can argue.

At least we have this to hold the flame, the rest of you better be thankful because its been the cause of some great pricedrops im sure.

nvidias shaking in their boots, it was odvious from that adobe document we downloaded a while back where they dissed on kyro and get this, their 'unproven commitee'...thats the most hilarious thing i'd read in a long long time. the only committee NV has thats proven itself is their marketing.