Lucid Hydra 200: Vendor Agnostic Multi-GPU, Available in 30 Days

[annihilat0r]

Shifting the baseline just to accomodate neo-math doesn't work either.

actually it works, according to the mathematical definition of linearity.

y = ax is linear but it's not the only case of linearity, y = ax + b is also linear, which means a linear increase by adding a second card doesn't have to be %100. in fact, each card can add %1 more performance (100%, 101, 102...) and it would still be linear.

[m0da]

actually it works, according to the mathematical definition of linearity.

y = ax is linear but it's not the only case of linearity, y = ax + b is also linear...

f(x) = mx + b satisfies the definition of linearity if and only if b = 0. Your equation is linear, but, a system with that response would not be considered linear. In a linear system, an input of 0 MUST yield an output of 0.

As always with claims of this magnitude, let's see if they get anywhere near even 50%.


Also, I love that quote in your sig. You should've seen my face the first time I saw it

[trinibwoy]

actually it works, according to the mathematical definition of linearity.

y = ax is linear but it's not the only case of linearity, y = ax + b is also linear, which means a linear increase by adding a second card doesn't have to be %100. in fact, each card can add %1 more performance (100%, 101, 102...) and it would still be linear.

Yes, if you ignore the first data point going from 0 to 100. Which makes absolutely no sense when talking about GPUs because it's the performance of the first GPU that defines your starting point. If you do that then your second data point (101) would make no sense at all because you would have no baseline upon which to calculate a ratio.

[adpr_02]

Why are we even arguing this? Linear basically means that to make this graph you take a ruler and draw one line on that ruler (w/e slope, without moving it) from -infinity to +infinity or w/e you define the limits. If you have y values equaling 0, 1, 1.5 which are equally spaced on the x-axis, I'm sorry but that's not linear.

That's not what we should be arguing. I think that where the confusion comes in is what are the limits? 0 --> a few cards(slope of 1) , or 1 --> a few cards (slope anything > 0)? If the latter, a lot of people will be very disappointed when this thing comes out. I think you could go as far as say it's "false advertising" as the majority of people are interested in 2 cards - where sli and crossfire are already adequate.

[Decami]

Why are we even arguing this? Linear basically means that to make this graph you take a ruler and draw one line on that ruler (w/e slope, without moving it) from -infinity to +infinity or w/e you define the limits. If you have y values equaling 0, 1, 1.5 which are equally spaced on the x-axis, I'm sorry but that's not linear.

That's not what we should be arguing. I think that where the confusion comes in is what are the limits? 0 --> a few cards(slope of 1) , or 1 --> a few cards (slope anything > 0)? If the latter, a lot of people will be very disappointed when this thing comes out. I think you could go as far as say it's "false advertising" as the majority of people are interested in 2 cards - where sli and crossfire are already adequate.

Adequate? near 100% scaling is beyond adequate. I love how people say a company getting rather large attention and funded by companies that could take over the world in a secret underground armed facility if they chose to spend their cash that way..... is just straight up lying, when there is doubt, and then ignore the technology as if it means nothing when theres less doubt about its claims. Your seriously mentioning the phrase "false advertising" and at the same time making up numbers speaking like its truth.

To some of us, who refuse to use the multi card solution known as SLI/Xfire because of the terrible scaling and micro stutter, it is far from "adequate." The people in this world, who push its technology forward, don't think in "adequate" terms. Who the hell wants something thats...."Adequate." Thats my newly most hated word now.

This is Xtreme Systems forums. Not Adequate Systems forums!

[villa1n]

Maybe its me... but Do we have any benches...or are we just using three pages to discuss mathematical semantics?

[adpr_02]

Adequate? near 100% scaling is beyond adequate. I love how people say a company getting rather large attention and funded by companies that could take over the world in a secret underground armed facility if they chose to spend their cash that way..... is just straight up lying, when there is doubt, and then ignore the technology as if it means nothing when theres less doubt about its claims. Your seriously mentioning the phrase "false advertising" and at the same time making up numbers speaking like its truth.

To some of us, who refuse to use the multi card solution known as SLI/Xfire because of the terrible scaling and micro stutter, it is far from "adequate." The people in this world, who push its technology forward, don't think in "adequate" terms. Who the hell wants something thats...."Adequate." Thats my newly most hated word now.

This is Xtreme Systems forums. Not Adequate Systems forums!

Chill out, dude.

I never said they're lying. I re-read my statement, and I did seem rather anti-Lucid - which i am not. I do beleive they have something revolutionary cooking up for us. My bad with that.

However, all I said was IF the scaling was not linear, it would be bad. I don't know where you see these "false numbers" you accuse me of giving out, or me concluding that it's flat out a lie. I just gave two possible scenarios and recapitulated what was being argued in the previous few posts.

We don't know any numbers and all we can do is hypothesize at this point. Besides it's good to be critical of what companies tell us some times.

just my thoughts.

[madcho]

if they re not scalar as claimed they'll not be buyed.

And 30 days after no motherboard is on market.

Maybe they'll offer on socket G too lucid ^^

[Decami]

Chill out, dude.

I never said they're lying. I re-read my statement, and I did seem rather anti-Lucid - which i am not. I do beleive they have something revolutionary cooking up for us. My bad with that.

However, all I said was IF the scaling was not linear, it would be bad. I don't know where you see these "false numbers" you accuse me of giving out, or me concluding that it's flat out a lie. I just gave two possible scenarios and recapitulated what was being argued in the previous few posts.

We don't know any numbers and all we can do is hypothesize at this point. Besides it's good to be critical of what companies tell us some times.

just my thoughts.

yeah totally agree, your post just seemed, idk, as you said anti lucid for no particular reason, but i see your point now.

It just drives me insane, some people on here, that are immediately against something for absolutely no reason. Like they enjoy putting things down, are watching things fail.
Especially something, as revolutionary as this will be if, if everything turns out as expected.

There should be a positive attitude for industry changing techniques that hugely benefit the future of its particular industry. Progress is good, and should be hoped for.
Not saying you shouldn't ask about it though, or doubt it, just don't hope for things like such to fail.

[naokaji]

Maybe its me... but Do we have any benches...or are we just using three pages to discuss mathematical semantics?

No products and no benches, just some folks in some lab that made a promise and too many people falling for the hype.

[trinibwoy]

There should be a positive attitude for industry changing techniques that hugely benefit the future of its particular industry. Progress is good, and should be hoped for.
Not saying you shouldn't ask about it though, or doubt it, just don't hope for things like such to fail.

What techniques? They've given no details whatsoever on the technology and there has been no independent verification that it even works. You're asking people to blindly swallow empty promises from an unknown and unproven company.

[BenchZowner]

I happen to know someone who has seen it working ( I can't post any comments regarding its efficiency, etc ).
Seeing is believing, using needs public launch, and TBH I can't see that happening anytime sooner than.. hmm... March ? Hopefully ( for their sake )

[annihilat0r]

I can't believe still we have no performance / frametime data at all

[trinibwoy]

I happen to know someone who has seen it working ( I can't post any comments regarding its efficiency, etc ).

Heh, well now we can all claim we know someone who happens to know someone who has seen it working

[AndrewZorn]

I can't believe still we have no performance / frametime data at all

thanks for saving me from reading pages of speculation, came in basically to see if anything has happened yet (no sarcasm or anything)

[Manicdan]

if they already have a working product and havent released, the one of two things is going on. either they are struggling to get it to work, or they are making sure it comes out with every expectation met or exceeded. too bad those are two extremes and theres no way to tell which one is happening.

[Aberration]

Or they are making sure they have enough supply to meet demand when released.

Worst thing you can do is release a product without enough supply, and then the price goes through the roof.

[Decami]

What techniques? They've given no details whatsoever on the technology and there has been no independent verification that it even works. You're asking people to blindly swallow empty promises from an unknown and unproven company.

Actually im not, and its clearly in the quote you posted from me. Funny thing is no information has surfaced on whether its working or not working as it should be, yet everything is already an empty promise to you.

[Manicdan]

Or they are making sure they have enough supply to meet demand when released.

Worst thing you can do is release a product without enough supply, and then the price goes through the roof.

if supply was an issue, then i think we would see it on more boards than just the MSI big bang

[Aberration]

???

I would think that if supply wasn't an issue, you would see it on more boards.

It also could have nothing to do with Lucid, could be a production problem with MSI.

[trinibwoy]

Funny thing is no information has surfaced on whether its working or not working as it should be, yet everything is already an empty promise to you.

Well what's the appropriate response to having no information and only promises?

[Manicdan]

???

I would think that if supply wasn't an issue, you would see it on more boards.

It also could have nothing to do with Lucid, could be a production problem with MSI.

it really could be viewed both ways, hard to correlate exact reasons between lack of production chips and number of board partners. if there are no chips then they can keep it discrete, (lack of supply due to fab issues) or if they have millions of chips then it would have to be on many motherboards before they sell out (lack of supply to due very high sales expectations) since i doubted they are having trouble building the chip (they showed it off like a year ago and yields shouldnt be bad on a chip that uses 5w) i cant imagine that they cannot produce enough to keep up with supply

i hope the delays are due to trying to give us every feature possible, and not because they are struggling to give us the minimum features required.

[Decami]

Well what's the appropriate response to having no information and only promises?

what promises do you expect to have the answers for before product release. No promises were made about releasing information about thier product before release...so....still not understanding what "promises" you are refering to.

The only promise i can think of is the cards performance. But hows that an empty promise when we know nothing either way. and its before release.

We dont know anything yet. Hell not having information could have nothing to do with Lucid. Maybe MSIs baords are failtown, we dont know.

[Sh1tyMcGee]

Looks like i was right, this was too good to be true.

[trinibwoy]

what promises do you expect to have the answers for before product release. No promises were made about releasing information about thier product before release...so....still not understanding what "promises" you are refering to.

One easy example is DX11. We had details on the new pipeline months (years?) before we had hardware so we know how stuff works before we know how fast it runs. So far Lucid is just dropping vague nonsense like "intercepting DirectX calls". Even the journalists who supposedly saw it in action we given zero technical information at even a high level.