Misleading information on the 1080 reviews

http://www.techpowerup.com/223440/msi-and-asus-send-vga-review-samples-with-higher-clocks-than-retail-cards

What is whit that? What is going on with the marketing? Don't Asus and MSI realize, that this will not stay hidden, or they count on the consumer being numb enough from all other shit and just take it?
I don't understand this...

1 Like

Now this is good transparent journalism.

1 Like

Both MSI and ASUS sent modified GPUs to the press. Gigabyte so far did not.

http://videocardz.com/61121/asus-and-msi-accused-of-sending-modified-cards-to-the-press

Nvidia's GPU launch has stunk since the start.

crazy bs overclocks at press demo, inflated benchmark numbers, only showing 1080p performance (when 1440p and 4k as the only reason to buy this GPU)

best part according to an NV engineer it has the same performance clock per clock as maxwell, just OC'd to shit

3 Likes

So it's a rebrand like AMD cards were? It's funny that they tried to benchmark mostly at 1080p, I guess this card be nice for 1080p if your doing 200Hz or 165Hz, but I have no problem with 1440p 50-60 fps on Ultra on my R9 390, and I just play 4K 60 fps on less demanding games like Rocket League.

1 Like

heavily overclocked rebrand...

Yeah, it was pretty good for a rebrand, but I was just calling it for what it was even though it don't perform like a rebrand.

If you ignore the completely reworked core architecture, new memory controller, increased L2 cache and cache-miss prediction, introduction of a more modular core sub-units (SM, TPC, GPC) allowing for better yields and slices of imperfect ASICs, oh and not to mention that it's finFET and not planar.

So much innovation... All is forgiven then...

Really? Many people say there is no big difference at all...

@psycho_666 They're not rebrands, its a die shrink and supposedly new arch, but according to the engineering it essentially has the same clock per clock performance of Maxwell indicating they just die shrunk it and gave a very minor tweak (and the die shrink lets it reach these crazy clock speeds)

but looking at the 4k benchmarks, where the 980TI falls on its arse, so does the 1080. Which is why I'm 100% not impressed with it.

Those people aren't looking at die shots and reading into the specs. They had to rebuild the architecture for finFET, you can't just shrink planar into finFET and have everything work like it did.

4 Likes

Yeah, i dont realy understand why people keep saying that Pascal is a rebrand.
On which its obviously not.

Also about those review samples with higher clock speeds, are we talking about FE cards?
Or are we talking about aftermarket cards from their own?

"MSI and ASUS have been sending us review samples for their graphics cards"

I guess it's for their own aftermarket cards.

2 Likes

Yeah thats exactly my point.

I think we are talking about the aftermarket cards... After all, MSI and ASUS seems to be most aggressive with their GPU review sample marketing...
The sad thing is, the difference is basically 5%... Does that means there is even lower difference in performance?

1 Like

The performance diffrence between the 1080 and 980Ti you mean?

I think the better question that should be asked is, are these review sample clockspeeds achieveable with an overclock? Cause if that's the case TechPowerUp is just grasping for straws.. Yeah it's wrong MSI and ASUS are doing that but come on complaining about a 3% overclock. If you're buying a 1080 you're most likely going to overclock it anyways..

the upside to this, is the reviewers can copy the bios and give it out to the public to use or give to the hands of talented bios editors for some awesome custom bios's

omg the comments wccftech writer got trolled so hard

So let me get this together:
-Nvidia faked temps and overclocks at launch event (or got damn lucky)
-Pascal is Maxwell at 16nmFinFET (with all the benefits 16nm got them)
-MSi and ASUS try some wierd move to get people to buy that gold ingot

Has AMD send them some early Vega card that just destroyed 4 way SLI 1080s or what is going on? Something is of by miles here and I bet there is a good capitalistic reason behind it.

Comparing the 1080 and 980ti, it gets really funny.
The 1080 has only 91% the cores of the 980ti but 161% the clockspeed.
Memory interface is only 67% compared to the 980ti which with the higher memory speed gets the 1080 95% the memory bandwith of the 980ti.
Someone cut some corners and really, the 1080 is optimised for 1440p, not 4k. (Which might be why the FuryX gets so close to it in 4k benchmarks.)

2 Likes