Recent content by ManicOne

Its wot I got. And it fukken rokks.

Translation for the slow: ATI implemented DX11 by addition to its already existing architecture. Tesselation was already there but not DX "visible". NV implemented a new architecture which, apart from being hot and thirsty (in its current implementation, get it?) shows a lot of potential...

Some nvidia fan boy hey... Look at my sig and pull your head in.

Fun idea but not gonna happen on 40nm. 28nm is the next node, which in transistor "length" is a huge shrink. 28 will make these currently ridiculous sized chips more than viable. Some points though; - I don't think Fermi needs a ground up redesign; IMO the arc is healthy, and with a few tweaks...

Umm why? I can't see XFX dropping NV for good.

What respin? Kyle's already said A3 is it for this chip. NV's next high end silicon will most likely be something along the lines of GT580.

This :D:D:D

Trimlock FX5 series were bad architecturally. I honestly don't think Fermi is (in potential anyway). Maybe 3 billion trannies (no not that kind) were just a bit too many on a leaky 40nm process. Bring a few revisions and a shrink and lets see what happens. We're not here for the short term are we.

Didn't you here, NV shot JFK..... Seriously m8, conspiracy theories only work if the truth never comes out. NV is a hardware manufacturer, their performance (and hopefully to a lesser degree mindshare) are what sell hardware. This is just a product that's late to market, not a predetermined failure.

First things first kill Turbo boost, and all power saving measures. Run on an uneven multi at all times. 21 is your friend. Kill VDroop with LLC if you can. Oh fukk it this guide is da bomb (for 920s, but perfect for 930s, no EIST problems for anyone with a 930)...

Wonder if the calls are only in "extreme" tessellation cases. Honestly I doubt Unigine would go out of their way to break DX11 spec. I think the timing of this release is more appropriate than suspicious.

Yeah 5850s have regs. There are two reasons why Cypress chips are binned as 5850s rather than 5870s: Either they can't hit the required clockspeeds for 5870 performance, or part of the shader silicon is broken. If A, then they won't overclock like 5870s will, if B then they might overclock like...

Vcore @ 1.22V here, 4.01GHz. More stable then a stable table. Will check my uncore etc and get back. There are definitely ways to sweetspot this cpu.

QuadMe, the first pic is tesselation off, not normal. Check out the dragon's neck spikes, its obvious. Also, anyone set it to extreme and pushed all the tesselation values to 2? Its INSANE...

Chuck on the 10.3a drivers. Could be that?