Got Apex Parhelia review

[dshumake]

Killer G,

You are right. Since triple head (surround gaming) is the way to play games that is how the card should be tested.

Of course that also means that ALL cards need to be tested at that resolution and if they cannot then they should receive a 0 for the benchmark.

dshumake

[KillerG]

Originally posted by funky-d-munky
You couldn't have just switched it to single-head and done the bench. I mean how hard to do is that. Giving it a no score at 1600x1200 just because you couldn't take the time to switch to single head from thriple head. You took months to do a review but ya couln't once slip it in single head?

To anyone not knowing the tech of the Parhelia they are going to think it can't be used at 1600x1200.
It just seems a little half-assed and a little misleading.

But to be fair other than this debate, I again thought the review was solid.
funky

Well, after I give this...

I have chosen to test at resolutions from 800x600 to 1600x1200. But, before we even get to the first round of benchmarks, I should let you know that the Parhelia will be returning a result of 0 (zero) for all 16x12 tests as it will not run games at 16x12 when configured for TripleHead display.

...very succinct explanation for the zeros if anyone is confused it is due to their own inattention or unwillingness to receive the full benefit of the review by reading it all.

As for if this is half assed, perhaps you should also note it as being such because I did not benchmark every single game that anyone might wish to play using the card. The combinations and permutations of tests and results presentations are just to vast (infinite really) to include everything. That said, if you are so very in people having the results for testing with only a singlehead setup then how about you do a little testing yourself and post those results here. Don't forget to do a fresh itstall of the OS and all of the testing applications before you begin. And, don't forget to run every test 5 times just to be sure the results are reliable.

Just in case you are wondering, yes, I am being a bit of a wise ass here. This is due to your insistence upon using wording that one might regard as insulting. Of course I am referring to your inference that I may have conducted this review in "half-assed" fashion.

Tell me, were you happier with reviews that merely sang the priases of the card's features and rehashed the marketing BS without even including SG benchmarks? Did you get on those folks for being "half-assed"? I seriously doubt it. I understand that peeps like to stand up for the products that they believe in. But, if you would like to comment on review methodology and procedures, I would suggest that you have a go at it yourself. All of the necessary tools are easily available to you.

[KillerG]

Originally posted by dshumake
Killer G,

You are right. Since triple head (surround gaming) is the way to play games that is how the card should be tested.

Of course that also means that ALL cards need to be tested at that resolution and if they cannot then they should receive a 0 for the benchmark.

dshumake

I think that is a bit of a stretch. The review makes it abundently clear that Parhelia is the only card that can make use of three monitors and has anything even like SG mode for gaming. I think you are just being argumentative for the sake of arguement at this point.

[KillerG]

Getting back to the DX9 compliance issue as I said I would, here is the reasoning for leaving out some DX features as was originally indicated my Matrox PR and was reiterated by the very same person just yesterday.

If a portion of the die had been used for the implementation of DX9 pixel shaders, some other features would have had to be left out. It was a matter of what could be fitted to the chosen die size.

I realize that this is a choice that Matrox was free to make and that DX9 games will not be available in quantity for some time so I did not go into in a lot of detail. But, being as you asked...

The way I see it, I don't give a monkeys about the problems an architecture team may have as far as space saving measures go. If there is feature that should be there, bloody change the design and include it. If this is somehow beyond the capabilities of either the design team, the manufacturing process, etc... or outside of the R&D budget or the target cost for GPU manufacturing then something is broken and needs to be fixed. I mean come on, look at the Parhelia GPU. It is so far behind the competition that it is absolutely ridiculous. Take for example the operating frequency of 220 MHz. What the hell is up with that? More than 100 MHz behind the current market leader and almost 300 behind what the NV 30 will be!?! Doesn't this make any of you wonder what the heck is up with that? It would seem to me to pretty clear that Matrox is completely unable to compete as far as processing speed is concerned. There is also a signifigant lack of technology along the lines of efficiently performing its tasks as well. Even at 220 MHz, with the very fat data bus on the Parhelia it should do much better than it does. Bottom line here: The Parhelia GPU is a very weak example of graphics technology when it is compared to the cards put out by its main competitors.

I am pretty sure nobody in this forum wants to hear that but, as the nitpicking at what I believe is the most comprehensive and honest review of the card to date continues, I figure we might as well go at it whole hog or not at all.

Anyone want to get into the reason why I smaked it a bit for FAA not working with stencil buffers now or ever? Just in case you do, be forwarned that this is a huge can o' worms as far as I am concerned. I should have smacked it harder for this. I could have simply saved myself a ton of trouble and made the review one simple sentence. Something like this...

If you wish to have image quality in games equal to that of the comptetion in many future titles (and many present titles too) do not even conside the Parhelia as it will not ever be able to provide it due to its inferior and poorly implemented antialiasing capabilities.

Gees, maybe I should rewrite the review and really get down and dirty on the cards failings while I try to accomodate the concerns mentioned in this forum.

As to not having tested with the 103 drivers as peeps have mentioned, the review was held up at the request of Matrox. I will be damned if I will competely redo the hundred or so hours of testing that it took to put this review together just because they released a new driver between when I sent the first copy of the review for them to have a peek at and when we were able to get some issues settled. If of course someone out there would like to pay me to do this then I would be more than happy to do so. About 25 bucks an hour USD is all it would take.

Also on the driver version front, if someone out there would like to post some v.103 results for, say, MS FlightSim 2k2 in surround mode that show any MAJOR increase to the 16 FPS that SimHQ showed in their review with older drivers or perhaps that show that Nascar 2k2 is actually playable (you know, something over 13 FPS, like maybe, dare I say it 30) then do it up. Andplease, lets not have these tests run at 640x480 or equivilent and lets not turn all the graphics settings down as low as they can go either. Try to remember while setting your testing parameters up that the competition does not have to do those things at all to play those games quickly and beautifully.

Next...???

[Greebe]

I will be damned if I will competely redo the hundred or so hours of testing that it took to put this review together just because they released a new driver between when I sent the first copy of the review for them to have a peek at and when we were able to get some issues settled

Exactly how many months passed by before PR responded? Did you bother to check to see if there were newer drivers released prior to your initial testing of Parhelia?

and while you are at it how about posting up the SH results and when comparing others cards put that FAT 0 next to their SG scores

also if you want to include the ugly then I think you will find that most of us here are ready for any comment/insight on that front.

[William]

Originally posted by KillerG
Gees, maybe I should rewrite the review and really get down and dirty on the cards failings while I try to accomodate the concerns mentioned in this forum.

So you're saying that you've done a review where you've deliberatley brushed over failings that you see in a product?

[KillerG]

Hmmm, apparently I ruffled a feather or two this morning.

Greebe

"Exactly how many months passed by before PR responded?"

KillerG

I neither implied nor insinuated that the delay was caused by Matrox being tardy in their reply. As a matter of fact, Matrox PR did an exceptional job of offering everything that a reviewer could want and did so in a very timely fashion every single time. The fact of the matter is that they wished to discuss the review via phone conversation but a work contract kept me away from telecommunications during business hours for almost two whole months. This was certainly not what either party wished but, it was unavoidable. As it was a request from Matrox to discuss some issues before publication that initiated the dely, I did not see it as my responsibility to redo eveything as the comparitive cards had new drivers issued in the mean time as well. Especially since it was of my own free will and generosity of spirit that I sent them an advance copy and agreed to delay publication until their concerns were addressed.

Greebe

"Did you bother to check to see if there were newer drivers released prior to your initial testing of Parhelia?"

KillerG

Did I bother? OK, yeah, whatever. Is that a rhetorical question or do really think one would go to the effort I did for this review without taking two minutes to grab the latest drivers?

Got an intelligent question? Got some results comparing the two drivers that will thrill and amaze me and make me want to use the card for gaming?

BTW, I did run several tests with the new drivers that showed no differences worth mentioning let alone being worth redoing all the testing for.

Greebe

"and while you are at it how about posting up the SH results and when comparing others cards put that FAT 0 next to their SG scores"

KillerG

How about reading the whole thread before jumping in. I believe this has been addressed.

But, being as we are back to that... Did anyone happen to notice the NVIDIA cards receiving some zeros too for a missing feature? Has anyone checked the web for reviews I have done on previous products to see if this is something that might be well considered and fairly standard methodology for me rather than a simpe attack on Parhelia? I guess some may have but apparently it is not procedures that folks truly wish to discuss. Rather it would seem that Parhelia fans are simply grasping at whatever straws they can find to try to save face for a card that simply does not cut it when gaming is the question at hand.

However, because I am a fairly patient man, I will try to explain this once again woth a "down to brass tacks" bent this time. The only compelling reason to buy the Parhelia is to have the triple desktop. Period. Therefore, it does not make sense from a gaming point of view to benchmark it in any other configuration.


Greebe

also if you want to include the ugly then I think you will find that most of us here are ready for any comment/insight on that front.

KillerG

Really, you mean you are prepared to admit that some aspects of the card are ugly? If not, I wonder where your rebuttal is to the comments in my last post (which you have apparently read). Unless of course you have none because what is said there is undeniable.

[KillerG]

Originally posted by William
So you're saying that you've done a review where you've deliberatley brushed over failings that you see in a product?

I was waiting for that one.

Absolutely not.

I just tried to say what was necessary to get the point across without being nasty about it. This is something that I did very intentionally. The reason being that I would not want readers to think that I was just hell-bent anti Matrox and therefore possibly somehow biased. If they should sense something along those lines, whthere it be real or imagined, they may give no weight to the review at all and I have failed. (Sorry, was distracted and posted without the last line here initially) KG

I think I was pretty clear in letting people know that I would not buy the card for gaming purposes. Besides, I don't think the average Joesphine/Joe gives a monkeys about what will or will not fit on a die and so on. They just want to know if they should buy or not.

[Greebe]

Gawd.. bit defensive today aren't we?!.. and I wasn't even dissing your review, just pointing out what we see as obvious problems...

Shame you refuse to address with professional tact and instead are quite overly defensive for a reviewer imho


Whatever... it's just another so so review


BTW if you really want insight as to P's problems hang here for a bit and read up. I'm sure there are many things that you think we are so clueless of will be on the other's foot.

[K6-III]

There are more reasons than just triplehead to go for Parhelia.

The PrecisionCAD drivers for one.

Certified drivers for professional apps another.

Seeing how things appear to be headed, this card is for people that want to do some of everything....

[KillerG]

Hmmm, not particularily. I really don't like it when peeps ask questions that have already been answered. But, that is not a biggee really.

Of import here is that there seem to be those who think I have written a review that falsely represents the Parhelia and/or its capabilities/worth/value through my expressed opinions or my testing methodology and the results thereof. Especially where gaming is concerned.

What I am attempting to incite is the bringing forth of some solid arguements which include some numbers and so on. If I am wrong, and someone can show me exactly how, I will change the review and post an addendum on my errors to it as well. That is the beauty of HTML. It is dynamic. I have no dillusions of infalability. I am never opposed to discussion, criticism or arguement (provided the passion is real that is). On the contrary, it keeps us all honest and helps ensure that we leave neither misrepresentations nor inaccuracies unaddressed.

[dZeus]

The problem is not that we haven't already shown you what we think is wrong with your review.

[Greebe]

Your review disclaimer should then read "from a typical hard core gamers perspective" , not responding to me with those Q's have already been answered... did you stop to think I (among others) thought they were inadequite or faulted regardless?

[KillerG]

Nope, it is not a card for anyone who includes any sort of real gaming capabilities on their list of things it must do. Not if they have reasonable expectations of performance.

As to whether the drivers are extremely good for pro apps I could not comment. This is why I have no comments in the review that address the pro app drivers. If a reader wishes to know about those aspects of the Parhelia, they are most likely pros who would know where to go to find that sort of review. I did give a couple of pointers towards pro app speed testing in order to further illustrate the performance issues I have with the GPU.

That said, when Matrox saw that I had pointed to some particular test pro app results they expressed some trepidation as to the validity of those results. I was immediatelty concerned as I know next to nothing about these tests. I has simply pointed to results that I had seen on a site I know is reliable for 3D reviews. They did not request that I remove the link or anything like that, they just brought it up. As I am certainly not a pro when it comes to pro apps, I immediately responded that I would look at those results more closely. I would also compare them to others before I published the review. Further, I suggested that if I found obvious fault I would remove the link entirely. And still further, that I would link to other results regardless of my findings on the initially linked ones.

This should cover my angle on pro app related stuff.

Thjere is more about the review process that I could share if you like...

To be brutally honest, I really hope that there is a giant market for Parhelia in the pro workstation sector. As I have noted many a time, I am a Canadian as is Matrox. We like to stick together. More than that, where this review is in particular is concerned, shortly after I began to review the card, the owner of Matrox (yes, if ya don't know Matrox is privately held) gave a giant chunk of change (yep millions) to one of our major Canadian universities. This just about made my heart break as I was reading the test results for Surround Gaming. All I could think of was that wonderful acts like that gift being perpetrated by someone as concerned and dedictaed as that man may be lesser somewhat due to my actions. Shit, I felt like not posting the review. But, I had professional responsiblities. And, more importantly, I maintian that I do take the consideration of the reader as the number one factor in every facet of every review that I write. I could not not post it after reading so many that did not say things correctly or clearly.

As I said to Matrox PR yesterday, it is simply impossible to get everything that one thinks about, mulls over, etc during the review process into the review.

[KillerG]

Originally posted by Greebe
Your review disclaimer should then read "from a typical hard core gamers perspective" , not responding to me with those Q's have already been answered... did you stop to think I (among others) thought they were inadequite or faulted regardless?

they were inadequite or faulted regardless

What were "inadequate or faulted regardless"? That makes no sense at all.