Wednesday, November 11th 2009
AMD Radeon HD 5970 Specs Surface
In a few days from now, AMD will unveil its new flagship graphics accelerator, the ATI Radeon HD 5970, which will intends to cement the brand's performance leadership over every product from rival NVIDIA. The HD 5970, codenamed "Hemlock", is a dual-GPU accelerator, with two codenamed "Cypress" GPUs in an internal CrossfireX configuration.
Built on the 40 nm process, these GPUs will feature 1600 stream processors each, and will each have a 256-bit wide GDDR5 memory interface to connect to 2 GB of memory (4 GB total on card). The clock speeds are where the specifications of these GPUs differ from their single-GPU avatar, the Radeon HD 5870. The core is clocked at 725 MHz, while the memory runs at 1000 MHz (4000 MHz effective).
The accelerator will not have a rear panel identical to those of other Radeon HD 5000 series accelerators. It has the usual broad air vent occupying one slot, while the other has two DVI-D and one mini DisplayPort (DP) connector. The mini DP connector can give out DVI output using a dongle, and in this way, support for ATI Eyefinity technology remains intact. The NDA covering this accelerator is said to expire on the 19th of November, not very far away.
Source:
TechConnect Magazine
Built on the 40 nm process, these GPUs will feature 1600 stream processors each, and will each have a 256-bit wide GDDR5 memory interface to connect to 2 GB of memory (4 GB total on card). The clock speeds are where the specifications of these GPUs differ from their single-GPU avatar, the Radeon HD 5870. The core is clocked at 725 MHz, while the memory runs at 1000 MHz (4000 MHz effective).
The accelerator will not have a rear panel identical to those of other Radeon HD 5000 series accelerators. It has the usual broad air vent occupying one slot, while the other has two DVI-D and one mini DisplayPort (DP) connector. The mini DP connector can give out DVI output using a dongle, and in this way, support for ATI Eyefinity technology remains intact. The NDA covering this accelerator is said to expire on the 19th of November, not very far away.
147 Comments on AMD Radeon HD 5970 Specs Surface
64 bit OS is just required to get 64 bit extension, and getting over 3.37 gb of system memory running Microsoft windows operating systems
Linux can do 4gb in 32 bit just fine.
This 4GB card will be seen as a 2GB card if it were to memory map the entire addressable memory as the card is basicly crossfire on a stick. So, no you do not haev 4GB of memory to use, you have 2GB per core.
Any further disagreements from this will result in a cheese grater and salt vinegar treatment to your ass.
Was there a 4970?
hint: That's not a fixed number.PAE PAE PAE PAE PAE PAE PAE PAE
www.microsoft.com/whdc/system/platform/server/PAE/PAEdrv.mspx
more useful info
en.wikipedia.org/wiki/Physical_Address_Extension
anyways im going back to the discussion of the video cards
Here's a link to an old but execllent article on Dan's Data, which is where I learn about this issue long ago.
its not as simple as allowing more than 4GB in a 32 bit operating system, and its nowhere near as good as a true x64 environment
on the GTX 295 the GPUs are on their seperate PCBs and each GPUs has it's own ram:
3/4/5 generation
5/6/7/8/9 type of card, 5= htpc class, 8 = enthusiast class, 9 = dual GPU
30/50/70/90, card rankings for type of card.
Simples : ]
so you can get a 5850, 5870 and 5890 all in normal (3/4 monitors) or eyefinity (6 monitors)
first number
second number
3rd and 4th numbers.
4
8 <- see single GPU
90 <- how pimp the card is in its class.
i misread that.
you mean the 9 as in x9xx
say... the 4970 that never existed
the second digit, not the 3rd : ]
For once I didn't make a typo either :p
I would like that. It would make the HD5890 a high-end but very balanced GPU like the HD4890 was already.
worked well on games like stalker, supreme commander (original/early patches) and even the sims 3, preventing eventual crashes as they hit the 2GB barrier.
I'm not entirely sure that all 4GB is addressable, i think its another case of "in a perfect world"
a large address space aware app can use whatever is available, if windows sees only 3.25GB of ram, i get the feeling the app may find a similar limit.
EDIT: from what I know, there are registers (or something) that take the ~0.8GB chunk out of the 4GB, making it 3.2GB addressable, and out of that, some goes to the kernel, etc leaving the lower (numbered) address spaces to apps. Don't quote me though, I'm not spot on... but that's the idea as I know it.