Ahhh it wasnt a rumour after all. Good news on the AMD front.

nv has a real habit of "releasing" "new" SKU's... the 250... and 240 and 9800 and 8800... now the 780 and the 980. Pretty soon their model numbers are going to be completely meaningless (if they arent already).

Rebanding again? I can understand the video card rebranding to an extent, but rebranding the chipset doesn't make any sense.

Has nVidia completely given up on releasing new products that are actually new.

So it's just a rebadged 780a. :rolleyes:

maybe this will drop the price of the "OLD" 780a so i can finally get an AMD setup :laugh:

There's a shock, another re-branding. Next they should re-brand themselves "Nvidia - We confuse people".

erockerSo it's just a rebadged 780a. :rolleyes:

and whats wrong with 780a?

phanbueymaybe this will drop the price of the "OLD" 780a so i can finally get an AMD setup :laugh:

i'll sell you mine when i get 980a

I almost wonder if there is at least something changing about this, perhaps a die shrink? I hope the addition of ACC support isn't the only difference.

cdawalland whats wrong with 780a?

Nothing? :confused:

Why would you buy a 980a when you already have a 780a other than the change in name? It's rather ridiculous.

I wish there was an nForce AMD chipset that handled 3-way SLI at 3(x16) speeds.

btarunrI wish there was an nForce AMD chipset that handled 3-way SLI at 3(x16) speeds.

Apparently Nvidia puts their time and efforts into renaming stuff instead of producing new stuff.

erockerApparently Nvidia puts their time and efforts into renaming stuff instead of producing new stuff.

Research and Development costs money, of which there is a shortage worldwide. There has to be a serious business value in what NVIDIA is doing. Apparently it doesn't feel the need to spend gobs of money making a chipset that does what the nForce 790i Ultra does: to provide three graphics cards with full electrical x16 bandwidth, not to mention the increase in platform costs. To do a 790i Ultra with the AMD platform, it needs the SPP to hold 32 lanes, 16 directly routed to a graphics card, 16 routed to a BR-03 that gives another 32 lanes. To do that, the MCP part of the chipset has to be separated to form a southbridge, a distinct MCP chip.

With 780a SLI, what NV did was it made an SPP + MCP chip that gives out only 16 lanes in all to graphics, to which the BR-03 is connected, this BR-03 gives out 32 lanes in all: either x16, x16, NC or x16, x8, x8. The "southbridge" you see on 780a SLI motherboards is actually the BR-03 chip, not an MCP. Now just remove the BR-03 chip from the equation, and you have yourself the nForce 750a SLI, where the SPP itself splits lanes into x8, x8 for two cards.

For the level of performance Phenom X4 processors ended up with, NV felt the 780a was enough. If you're buying three GTX 285 or even three GTX 260, you probably have enough dough for an Intel platform, thinks NVIDIA.

btarunrI wish there was an nForce AMD chipset that handled 3-way SLI at 3(x16) speeds.

I think the difference is marginal.

When you look at the 790i/780i vs. the 780a in terms of bandwidth available to the cards:

790i/780 = PCI-E 2.0 x16/x8/x16 = PCI-E 1.0 x32/x16/x32
780a = PCI-E 2.0 x16/x8/x8 = PCI-E 1.1 x32/x16/x16

I don't think there is any graphics card that is Tri-SLI capable, that also benefits from anything higher than PCI-E 2.0 x8.

DarknovaThere's a shock, another re-branding. Next they should re-brand themselves "Nvidia - We confuse people".

more like "Nvidia - The Way It's Meant to Be Confused" :laugh:

newtekie1I think the difference is marginal.

When you look at the 790i/780i vs. the 780a in terms of bandwidth available to the cards:

790i/780 = PCI-E 2.0 x16/x8/x16 = PCI-E 1.0 x32/x16/x32
780a = PCI-E 2.0 x16/x8/x8 = PCI-E 1.1 x32/x16/x16

I don't think there is any graphics card that is Tri-SLI capable, that also benefits from anything higher than PCI-E 2.0 x8.

PCI-E 2.0 GPUs that actually demand lots of bandwidth, like G92 or RV770, don't exactly perform the same when run on PCI-E 1.1 x16 and PCI-E 2.0 x8, though theoretically both should offer the same bandwidth. The RV770 for example takes quite a hit at PCI-E 2.0 x8 in comparison to PCI-E 1.1 or 2.0 x16. PCI-E 2.0 x8 may scrape some performance off a G92, but I'm sure G200 based cards will see a performance loss.

only issue with this convo was that it was not the same design for intel's side i had one it was PCI-e 2.0 16/16 the middle slot was pci-e 1.0 16x

i have had 780a and 780i and IMO 780a is a million times more refined than 780i

DarknovaThere's a shock, another re-branding. Next they should re-brand themselves "Nvidia - We confuse people".

Lmao i can just see it, but anther re-branding is nvidia sandbagging :confused: or just out of ideas atm ? Surly AMD has enough heat on them make them sweat a little, Im not trashing nvidia there products are good just im wtf:confused: atm by it.

Finally a Nvidia board!

newtekie1Has nVidia completely given up on releasing new products that are actually new.

It would seem so. This is very disappointing.


I'm waiting for NVIDIA to announce a name change to VIDIAN just because NVIDIA was around too long.

FordGT90ConceptIt would seem so. This is very disappointing.


I'm waiting for NVIDIA to announce a name change to VIDIAN just because NVIDIA was around too long.

Yes, I understand that there is a need to keep market freshness, and there really isn't a reason in the current economy and market to do a lot of product R+D for new products. However, there really isn't a reason they can't just keep the current products name, people already know them, and know what they are worth. There is nothing really wrong with just saying "hey, our products are already good enough, so we aren't going to waste money developing and releasing new ones".

In all honesty, I think it is actually a smart move to just keep the top and mid-range products of the previous generation around to fill the mid-range and low-end of the current generation. Why spend money developing and releasing products that replace existing products with the same performance? But they don't have to rename them.

NOT Rebranding

Good God people! Doesn't anyone check the specification?

I realize that the article was written a month before the release of the 980a chipset, but do reporters ever check the fact before posting roomers as fact? People go check the facts before you bellyache on the forum. The 980a is NOT a rebranding of 780a.

980a provide support for DDR3, DDR2, DDR1, AM3, AM2+, AM2, PCIE 35-lanes.
780a provides support for DDR2, AM2+ PCIE 35-lanes

AM3 processors supports DDR3 - AM2 and AM2+ do not!

Granted - Nvidia did not up the PCIE lanes to 3-x16 (or more) which some hard core gamers might want to complain about, but most of us can do just fine with dual geforce 295 (@2-x16) or less.

Why ASUS made a 980a mother board (M4N82 Deluxe) using DDR2 is such a waste. There is no reason to put a AM3 processor in it. And if you have an AM2 or AM2+ processor, you might as well use it on a 780a motherboard.

Running DDR2 at 1066 is only supported with 1 memory stick per channel (I.E 64-bit vs 128-bit).

I'll bet the motherboard will perform better running memory at 800 in dual channel mode vs 1066 in single channel mode.

walybareGood God people! Doesn't anyone check the specification?

I realize that the article was written a month before the release of the 980a chipset, but do reporters ever check the fact before posting roomers as fact? People go check the facts before you bellyache on the forum. The 980a is NOT a rebranding of 780a.

980a provide support for DDR3, DDR2, DDR1, AM3, AM2+, AM2, PCIE 35-lanes.
780a provides support for DDR2, AM2+ PCIE 35-lanes

AM3 processors supports DDR3 - AM2 and AM2+ do not!

Granted - Nvidia did not up the PCIE lanes to 3-x16 (or more) which some hard core gamers might want to complain about, but most of us can do just fine with dual geforce 295 (@2-x16) or less.

Why ASUS made a 980a mother board (M4N82 Deluxe) using DDR2 is such a waste. There is no reason to put a AM3 processor in it. And if you have an AM2 or AM2+ processor, you might as well use it on a 780a motherboard.

Running DDR2 at 1066 is only supported with 1 memory stick per channel (I.E 64-bit vs 128-bit).

I'll bet the motherboard will perform better running memory at 800 in dual channel mode vs 1066 in single channel mode.

You do actually know that even nForce 3 can support Phenom II? The motherboard chipset has no technical role to play with memory support in AMD platforms since K8. Several sources have confirmed this is merely a rebranded 780a that is "qualified" for AM3.