• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

1Ghz 5870 18% faster than GTX480 in Heaven with Tessellation off!

Rroff said:
I think that GPUZ is a fake btw unless it can't read the right info from the card - 2 bits of info on it relate to the GTX275 and some of the other info doesn't match the A3 cards they've been using.

EDIT: Thie BIOS string, device id and vendor all match a GTX275 and the texture fillrate and memory speeds should be approx. double what they are. The bios string should start with atleast 74.
Click to expand...

Whilst I think you're probably right about it being fake, especially going by the BIOS code, I've checked, you're right, and also the inspecific transistor count, I think the texturing fillrate is right. If they faked it I'm surprised they came across the bit of information that rather than running at the core clock, the texturing units run at half of the shader clock (so 675MHz in this case). Given the GF100 chip has 64 texture filtering units (it has 128 fetch units IIRC), multiply by the clock speed they're running at and we get 43.2GTex/s.

Not sure what you mean about the memory speeds, they seem potentially correct, too (I doubt they're going to have a DDR base clock of 1800, at least, that'd imply 7200MHz effective GDDR5 memory, which we're unlikely to see for a good couple of years. It is much lower than I'd expect it to be, though. )
 
Might be GPUZ not having correct data for the the texture fillrate (and doubling it), the memory bandwidth deffinatly should be well into the 200s of GB/s.
 
nvidia cards , when they finally com e out will be just the same as ati, its just a matter of prefference, good luck losers, im stickin to my super 4890, till i cant play games proper any longer!!
 
Rroff said:
I would still be running my 8800GT SLI if it wasn't for the lack of VRAM at higher res :D tho quite liking my GTX260 SLI.
Click to expand...

I still couldn't believe how good my 8800GT SLI set up was. At the time I'm pretty sure it was the most power efficient combo going too!
 
_TubbZ_ said:
But is it really? not like theres any games making it struggle and justify ripping it out for an extra few percent or 1 or 2 extra effects :rolleyes:
Click to expand...

Yes, I have the 4870X2 and will not be upgrading it any time soon...however, it still becomes last gen because the 5XXX range is the new one.
 
Last edited:
That GPUZ screen has been proved fake on another forum, the device ID was wrong (one of the 200 series cards, a 285 or something).

Can certainly tell Fermi isn't too far from release now, the number of leaks/fake leaks have increased exponentially in the last week...
 
Well nothing can be said for sure yet.

But I haven't really entered into this whole mess so here is one of my thoughts.

If nvidia are boasting their card equals a 5870 where heavy tessellation is not involved, think of this...

ATI didn't know how fast Fermi would be. They set their clock speeds and sold their cards.

Nvidia have set their clock speeds to match the 5870, not beat it. Surely if they could have, they would have? Fermi must have been a flop. The guy is boasting that it nails tessellation really well, but at the same time if you read between the lines he is saying that it real world situations, the two are roughly the same.

Now bear in mind Nvidia knew the competition, and set the 480 speeds to try to make it competitive and the best they can do is match the 5870? If they could have made it faster they would have, no doubt...I don't think there will be much OC headroom on these things.
 
Could it be that gpuz shot might be displaying wrong info as surely gpuz havn't updated to take account of the 4xxseries yet so it just displays random stuff because it doesn't know what it is representing did that when i got my 5870. Not saying it's real just an idea why some of it might be off while some of it might be right. I just want some proper benchmarks not synthetic or optimised to judge things on but either way i have been and suspect will continue to be ery happy with what i have for some time to come :).
 
Rhys you need to calm down and the point is there are now more then a couple of these threads about the same damn thing and as far as i know nvidia are not making threads on here.
 
BTW I just re-ran the benchmark with pure DX10 enabled rather than DX11 and as I surmised if is rendering some/all of the other DX11 effects just not through DX10 RPs. Dropping to the DX10 option I got a ~10% performance increase in that scene bringing my fps upto 90 :D
 
Reading on another forum, that die size of 621mm2 is physically bigger than TSMC are capable of.
 
You must log in or register to reply here.
Back
Top Bottom