R520 benchmarks - Hardware Analysis

Status
Not open for further replies.
7up1n3 said:
popblood.gif
I see you registerred in april too, fellow lurker ;)
Got some popcorn for me?
 
geo said:
Is someone making you be here against your will? Who is this mean person?


Obviously no one is making me stay here. I do however think its fair to chime in with my subjective thoughts on the matter. I think it's disapointing people are basing their expectations on the posts from these two individuals. As they really havent shown us anything. While everyone is getting involved in the drama we have lost focus on whats really important. The R520.

*Edit*

P.S. I dont blame ATI PR for doing their job. I mean. He's doing a good job at what hes supposed to be doing. And it shows by the threads I have been viewing. I certainly think this has been a PR disaster either way and he's doing an incredible job of dealing with it.
 
Last edited by a moderator:
Firefox said:
...would however have liked you or any other ATi rep to come out with this before everything went utterly out of control like it has...

Sorry for the delay - but this kind of 'exchanging' is not really what we do

Although we were prepared for the situation - the timing was 'excellent' as we were away at an annual conference

We saw the numbers - contacted our benchmarking facility (4 top guys + hundreds of boxes) and asked them if they could 'place' the results

i.e. it is impossible to know exactly what was tested and in what environment (Sander 'sent away' for numbers - and 'numbers' is what he got back - so he says) but sometimes a 'pattern' can be recognizable

The one thing we knew 'the card' wasn't going to be, was a final version of the R520XT for reasons that I am not going to go into here - but which we will share with Baumann et al when they have the real thing in their hands ;)

Once our guys confirmed categorically that it did not match the profile of an R520 series card - they then looked to see if the scores matched anything else

They don't

Anyway, the point is that we do not 'monitor & control' web forums - these things are fun because they are organic and grow in so many strange and unusual ways with little input from the vendors

However, continuing to see the same 'rubbish' being posted everywhere was something that I, personally, was not happy with

The same goes for Sander's desire to position himself as a martyr

He is a sharp-thinking, solid-built guy who certainly does not need 'looking after' - he knows exactly what he is doing and why

I think that the discussion with Sander should take place in front of a 'live studio audience' because then we can move the conversation around to any topic that presents itself - and people can see his face when he reacts to various questions

I am not talking about a 'popcorn opportunity' - but rather a chance to see a 'quick thinker' trying to work his way out of 'tight situations'

Anyway, I suspect that this will not happen

The bottom line is that much of what I have read was rubbish

<engages Super-Geek mode>

A famous TV character once said "It is a lie, and lies must be challenged" and that is the way I feel when I see this stuff

</end Super-Geek mode>

We are making final arrangements for the R520 Tech Day and we will be giving 100 of Europe's top publications access not only to the technology itself, as well as various whitepapers, presentations etc...

...but - most importantly - the people who actually design these chips

As far as possible, this will be a 'PR-free zone' because we want expert press to engage directly with out scientists to discover the truth

In this market, you either believe you have the roadmap or you don't - everything else is irrelevant


Despite all of the public rivalries between board partners, chip vendors and web sites...

...overall, this is a really friendly industry

CeBIT, earlier this year, was typical of the spirit within this industry, in that we had our nVidia counterparts over to our party and we joined their's the night after

Sure we 'go at it' by day - nVidia is a tough competitor - but after hours we get on fine with each other


The situation that Sander has created here with his 'I feel the need to write a column to drop your share price and I do not care if I have never seen the card' approach is not typical at all of the relationships that we have with the press or our competitors


As we move on to our Tech Day and new battles, he will not be able to let this go for several months - possibly years to come

In life, I would say that it is better to have a more positive purpose


Roll on R520, roll on Tech Day, roll on true numbers :D
 
ChrisRay said:
. In the end it really doesnt matter what either says. Because there is nothing new that we have learned.

Yes, Like Ratchet's 7800GTX review, you will write a b*ll licking X1800XT review ;) .. unless.. in andrzej's view, you're not the biggest nv fansite :D

btw, maybe sanders source was the same as rage3d's sherman, claiming that his friend over at asus told him how bad the r520 was and all the "render errors" it had..
 
Hmmm.... That is what was said isnt it.

So the board Sander Supposedly got his numbers isnt the final, but Xbit Labs is a final?
 
Last edited by a moderator:
weeds said:
Hmmm.... That is what was said isnt it.

So the board Sander Supposedly got his numbers isnt the final, but Xbit Labs is a final?


Most benchmark numbers everyone claims to have seen are from Rev2 boards, the board going into production should be rev5. (at least, that's what I can make up from all the pictures and fud at hkepc, l'inq etc.)
The last hkepc article says that have a rev2 board from somewhere in the beginning of this year, with the nv silencer cooler installed.

http://www.hkepc.com/hwdb/r520firstlook.htm
ATi Radeon R520的REV 02
The rev05 should also be quite a bit bigger than the rev2's
 
Last edited by a moderator:
Tom,

I must say you've made a total joke out of yourself with your latest 'update.'

Quite frankly, I have had no problem whatsoever with the 1.13GHz Pentium III and I'm using the exact same motherboards (Asus, Via and i820 chipset as well as the old P2B) as I presume you're using.

If you cannot refrain from posting erroneous and obviously biased reviews or updates or whatever you like to call 'em, then at least make sure your lab is correctly setup and all benchmarking and system assembly operations are performed by a person capable of doing so.

If I had a faulty CPU I would check back with the manufacturer in the first place and have it replaced BEFORE telling the world of your 'mishaps' or posting any benchmarks.

You're mis-informing your audience as well as making your own testing procedures as well as your ethics a joke!

From my point of view (just finished the 10th redundant SYSMark 2000 run on a i440BX):

There is NOTHING wrong with Intel's new Pentium III 1.13 GHz processor.

Do you really need all of this 'bashing' and controversy to be able to reach those 20 million hits you're so openly bragging about, if I were you I'd rather have half that and still have my journalistic integrity.

Hope you take this to hart this time, as you really should!

Sander Sassen


Publicly available letter, sent by the self-proclaimed 'Lab Meister' himself to Tom's Hardware - and available at this link...

...oh dear...

:devilish: :devilish: :devilish:
 
Andrzej said:
Tom,

I must say you've made a total joke out of yourself with your latest 'update.'

Quite frankly, I have had no problem whatsoever with the 1.13GHz Pentium III and I'm using the exact same motherboards (Asus, Via and i820 chipset as well as the old P2B) as I presume you're using.

If you cannot refrain from posting erroneous and obviously biased reviews or updates or whatever you like to call 'em, then at least make sure your lab is correctly setup and all benchmarking and system assembly operations are performed by a person capable of doing so.

If I had a faulty CPU I would check back with the manufacturer in the first place and have it replaced BEFORE telling the world of your 'mishaps' or posting any benchmarks.

You're mis-informing your audience as well as making your own testing procedures as well as your ethics a joke!

From my point of view (just finished the 10th redundant SYSMark 2000 run on a i440BX):

There is NOTHING wrong with Intel's new Pentium III 1.13 GHz processor.

Do you really need all of this 'bashing' and controversy to be able to reach those 20 million hits you're so openly bragging about, if I were you I'd rather have half that and still have my journalistic integrity.

Hope you take this to hart this time, as you really should!

Sander Sassen


Publicly available letter, sent by the self-proclaimed 'Lab Meister' himself to Tom's Hardware - and available at this link...

...oh dear...

:devilish: :devilish: :devilish:

Well looks like 2 kids are playing in the mud now, Andrzej you just made yourself look as bad as SSassen
 
neliz said:
Most benchmark numbers everyone claims to have seen are from Rev2 boards, the board going into production should be rev5. (at least, that's what I can make up from all the pictures and fud at hkepc, l'inq etc.)
The last hkepc article says that have a rev2 board from somewhere in the beginning of this year, with the nv silencer cooler installed.

http://www.hkepc.com/hwdb/r520firstlook.htm

The rev05 should also be quite a bit bigger than the rev2's

People already know that the R520 had several revisions, until it reached its final one. People already know the reason behind those revisions, as Dave Orton said so publically. People though haven't put 1+1 together yet, and haven't realised what differenciates those early boards from the final ones, as far as performance is concerned. When they do, they will see why early rev. cards can't be representative of final silicon performance.
 
Kombatant said:
People already know that the R520 had several revisions, until it reached its final one. People already know the reason behind those revisions, as Dave Orton said so publically. People though haven't put 1+1 together yet, and haven't realised what differenciates those early boards from the final ones, as far as performance is concerned. When they do, they will see why early rev. cards can't be representative of final silicon performance.

At this point, more speculation just makes things worse. I don't think anyone will know much more until just before the NDA expires.
 
Andrzej said:
Tom,

I must say you've made a total joke out of yourself with your latest 'update.'

Quite frankly, I have had no problem whatsoever with the 1.13GHz Pentium III and I'm using the exact same motherboards (Asus, Via and i820 chipset as well as the old P2B) as I presume you're using.

If you cannot refrain from posting erroneous and obviously biased reviews or updates or whatever you like to call 'em, then at least make sure your lab is correctly setup and all benchmarking and system assembly operations are performed by a person capable of doing so.

If I had a faulty CPU I would check back with the manufacturer in the first place and have it replaced BEFORE telling the world of your 'mishaps' or posting any benchmarks.

You're mis-informing your audience as well as making your own testing procedures as well as your ethics a joke!

From my point of view (just finished the 10th redundant SYSMark 2000 run on a i440BX):

There is NOTHING wrong with Intel's new Pentium III 1.13 GHz processor.

Do you really need all of this 'bashing' and controversy to be able to reach those 20 million hits you're so openly bragging about, if I were you I'd rather have half that and still have my journalistic integrity.

Hope you take this to hart this time, as you really should!

Sander Sassen


Publicly available letter, sent by the self-proclaimed 'Lab Meister' himself to Tom's Hardware - and available at this link...

...oh dear...

:devilish: :devilish: :devilish:

That topic has already be discussed:

ssassen said:
Ryano,

Indeed, I sent out a mailer to 700 members of the press, making a public apology for my rant against Tom, that I thought was the most respectful thing to do. So I guess that chapter is closed, I ate humble pie, so please move on.

Regards,

Sander Sassen
http://www.hardwareanalysis.com
 
Karma Police said:
At this point, more speculation just makes things worse. I don't think anyone will know much more until just before the NDA expires.

Puting one and one together does not speculation make. I didn't promise magical performance or anything, I am just trying to show why early revision boards (and they way they benchmark) should not be taken into consideration - I thought I was crystal clear.
 
Razor1 said:
...made yourself look as bad...

Really ?

Apologies if you think that

This letter is in the public domain and seems to clearly state the standard by which he believes sites should work when running numbers

My interpretation is that he is saying "you should have the product in your lab - with your regular test environment - have one of the people you trust run the numbers - and then check with the manufacturer to see if there seem to be any issues (i.e. performance way below what you might expect) before going live"

i.e. he has already come out in public to state clearly what he thinks is reasonable/ethical behaviour in terms of testing, results and relationships with vendors

On that basis, it seemed very relevant

Please let me know if you think I have the 'wrong end of the stick' !
 
Status
Not open for further replies.
Back
Top