Pelly Opens his Site: PC Perspective!!!

[pelly]

you can clearly identify that there is a proprietary bridge design in there as well?

It is clearly evident that you did NOT even read the full article...nor the full responses I have made here in this very thread...

As a result, our conversation is over...

Ragejg...thanks for the feedback! ;D

 

[Dave Baumann]

...but (as the article suggests)...there are no current applications (aside from HDTV editing) that take advantage of the full bandwidth of PCI-Express. By the time there are, you will certainly see native solutions by both vendors. Until then, it is doubtful we'll see any applications that can exploit the benefits a native solution has...

Sean, a word of caution here. What you are seeing right now is a big marketing tug of war with one vendor obviously playing up (possibly overly so) the benefits of PCI-Express and the other has to suggest that you don’t really need anything more right now. That message is very short term – its going to die as soon as all the products have transitioned to PCI-Express natively, which is probably going to be in a relatively short space of time. So, yes, applications that take advantage of PCI-Express’s main benefits are not likely to appear before all vendors have native solutions, but that message is only of benefit to those that are selling hybrid systems. The question is – are there going to be applications that make use of PCI-Express in the period that may affect users of hybrid systems?

 

[pelly]

Good afternoon Dave,

Thanks for the response...I completely understand and agree with your statement. I am certain that there are upcoming titles and applications which will be able to take advantage of all the features of PCI-Express...The examples listed within the article such as VPU/GPU-controlled AI and such are especially interesting...The main question to ask here is whether the ucpoming architectures which provide native PCI-Express are able to fully take advantage of the features. If I see the ATI card blow the NVIDIA card out of the water...you better believe I will recommend the ATI card...However, those judgements will be reserved until testing is complete...

The whole issue here is...when to opt for a native card...This all seems to depend upon when the applications are coming out. If my contacts have served me well, I believe this will all be a non-issue once the applications are available as every vendor will be using a native solution...

If I were to recommend a card to anyone looking for something new...my choice would be the following:

-For the person looking to retain their current rig, stay with AGP, etc...Go buy either vendor's AGP solution.

-For the person looking to upgrade to PCI-Express that always has to have the latest and greatest (ie: does not hold on to hardware for years) I would say that you should opt for either vendor's PCI-Express solution.

-For the person looking to spend $500 on a card to have for an extended amount of time...I would say WAIT! Wait until the apps/games in question are out and then buy the latest and greatest card which captures your fancy. They will all be PCI-Express...native...fast.....with any/all bugs worked out on the cards and platform...

Regardless of whether it has been positive or negative, I appreciate the feedback...thanks guys...have a good one...

 

[Joe DeFuria]

-For the person looking to spend $500 on a card to have for an extended amount of time...I would say WAIT! Wait until the apps/games in question are out and then buy the latest and greatest card which captures your fancy. They will all be PCI- express...native...fast.....with any/all bugs worked out on the cards and platform...

But that's a cop-out.

What about the person who wants bleeding edge at whatever cost, and who wants it now? (You know..."enthusiasts" )

In any case, my main issue (again) with your article is the implication that ATI's implementation isn't actually "native", but an "on-chip bridge."

 

[digitalwanderer]

What about the person who wants bleeding edge at whatever cost, and who wants it now? (You know..."enthusiasts" )

Go buy an Alienware!

 

[pelly]

In any case, my main issue (again) with your article is the implication that ATI's implementation isn't actually "native", but an "on-chip bridge."

Hold on! Let me make this crystal clear...

"I HAVE NOT MADE ANY DECISIONS OR JUDGEMENTS (public or not) WHETHER ATI's SOLTUION IS TRULY NATIVE."

At this time, I am working with the appropraite channels to get more information before making a decision. As a result, all details regarding the images in question were clearly labeled as being speculation with the possibility clearly outlined that the VPU could very well be truly native...

With regards to your "enthusiast" comment...it would be foolish for me to pick a card without having tested either. Picking card A sight unseen would make me a Fan-Boy of company A. Likewise, picking card B sight unseen would make me a Fan-Boy of company B. Deciding to test the cards and then recommend the fastest/best solution is what makes me an Editor and not a Fan-boy...

 

[Joe DeFuria]

Hold on! Let me make this crystal clear...

"I HAVE NOT MADE ANY DECISIONS OR JUDGEMENTS (public or not) WHETHER ATI's SOLTUION IS TRULY NATIVE."

Hold on!

"I HAVE NOT SAID THAT YOU MADE A DECISION OR JUDGEMENT ON WHETHER ATI's SOLUTION IS TRULY NATIVE."

You put the implication out there. Hey..."what if"....

On what basis? Pics of graphics cores?

With regards to your "enthusiast" comment...it would be foolish for me to pick a card without having tested either.

Indeed. Which is, um, exactly why I DIDN'T say you should pick a card before testing either.

 

[nelg]

Hold on! Let me make this crystal clear...

"I HAVE NOT MADE ANY DECISIONS OR JUDGEMENTS (public or not) WHETHER ATI's SOLTUION IS TRULY NATIVE."

What Joe and whql are saying is if you really are refraining from making any judgement you should also refrain from casting doubts.

However, to make things a bit more complicated and dramatic a very reliable source has provided us with information which questions whether ATI's current solution has brought bridged complexity within the core.

 

[digitalwanderer]

 

[pelly]

As an Editor and reporter...it is my job to share information and keep our readers informed. If we have information from a credible source which pertains to a controversial or popular subject, we must share that with the disclaimer that it may/may not prove true. Again, this will all be put to rest when we have actual cards and actual benchmarks...

Dig...you can come out now....I've got to get going for now and get ready for 12" of snow coming tonight...

 

[digitalwanderer]

Dig...you can come out now....I've got to get going for now and get ready for 12" of snow coming tonight...

Thanks, I get scared when the big brains fight.

G'luck with the snow, I got about 2" on the ground last night as a surprise and I'm expecting more....HOPING for more actually, I got a nice snow thrower this year and I've only gotten to use it 3 times. (But I did my house and a couple of neighbors on either side, the thing FLIES!!! )

 

[Joe DeFuria]

As an Editor and reporter...it is my job to share information and keep our readers informed.

It is also your job as a reporter to tell us who your sources are, (why they are credible,) and explain the logic for the basis you draw implications. And in this, you failed completely IMO. As it stands, I see no reasonable basis to suggest that ATI's implementation is anything other than they say it is: native PCI-E support, such that it supports full bandwidth full duplex operation.

Anything else is just FUD.

 

[Richard]

I was also surprised at that part of the article. Everything else (and of the site in general) looked very good.

If you'll excuse the comparison, that part of the article looked like something I'd read on the Inq (or rather, copy/pasted in these forums). You clearly say you're not sure, but the take away message is that there's something dirty on ATi's PCI-express solution.

It's like the INQ recently saying things look 'cloudy' for nVidia but quickly saying they're not sure that's correct.

It might be true, it might not. Until you're sure, I'd expect you to hold on to that information and keep researching. That's what I expect from general reporters. They don't print/open news with "we're not entirely sure but we think there are WMD, look at these two pictures".

If you have the smoking gun WRT ATi's hypothethical deceit, fine, print it, I'll be very happy to know that. If you're not sure, don't cast doubts because that will be the take away message. People will say "watch out for ATi's new cards" instead of "there's a chance ATi's new cards won't be native PCI-express but the jury is still out until they are benchmarked."

The details of your doubts will be lost and soon enough fanATIcs will begin to label your site nvidia-biased.

 

[Randell]

It is also your job as a reporter to tell us who your sources are

rubbish - it's important for some sources to stay protected - you decide on the credibility of the reporter/editor then by their track record.

 

[Joe DeFuria]

rubbish - it's important for some sources to stay protected - you decide on the credibility of the reporter/editor then by their track record.

Well, since we have no prior track record to go on here, what are we to do?

It may be important to protect "sources", but you also have to give a credible explanation for the assertions you make given whatever data the source hands you.

Someone handing you two snapshots of two cores, even if we assume the snapshots are legitimate is one thing. Using this as evidence to make suggestions / about detailed bus implementation is entirely different.

This immediately discredits the editor/author IMO. the author is either passing on rumor / fud without any legitimate basis, or came up with the rumor / fud himself. In either case, credibility takes a hit. At least in my eyes.

 

[Evildeus]

rubbish - it's important for some sources to stay protected - you decide on the credibility of the reporter/editor then by their track record.

Well, since we have no prior track record to go on here, what are we to do?

Wait? Time is the sole base to credibility.

 

[Randell]

rubbish - it's important for some sources to stay protected - you decide on the credibility of the reporter/editor then by their track record.

Well, since we have no prior track record to go on here, what are we to do?

Yes we do - Pelly has been around a long time. Now base your judgement on wehat you've seen before.

 

[Mark]

Ask yourself who has the resources to get detailed pics of a graphics core like that, who has the most incentive to help spread rumours that ATI's PCIe isn't native, and then finally ask yourself who Pelly - ex-nvnews Pelly - might have for a contact.

Smells fishy to me, like that time someone else here posted a few select slides from an ATI presentation provided to him by a source he should have known better to trust.

 

[digitalwanderer]

Smells fishy to me, like that time someone else here posted a few select slides from an ATI presentation to make them look bad.

Hey! I resemble that remark! :?

 

[Mark]

bah, you beat me to my edit