RV790 Samples Carry Faster Memory

[btarunr]

Having taped out late last year, samples based on the new RV790 graphics processor have been doing rounds in the industry. A few more details have surfaced about it. Earlier noted to have identical clock speeds to that of the RV770XT (Radeon HD 4870), the RV790 samples are now known to have higher memory clock speeds. While the Radeon HD 4870 has its memory frequency at 900 MHz GDDR5 (effective 3.60 GHz), the samples carry memory clocked at 975 MHz (effective 3.90 GHz). Interestingly the memory chips on the sample, labeled IDGV1G-05A1F1C-40X, made by Qimonda, are specified to run at 1.00 GHz, reaching the 4 GHz effective memory speed mark. The samples feature 1 GB of memory. The RV790 is AMD's new current-generation graphics processor built on the newer 40nm silicon fabrication process. The new process is expected to reduce the GPU's power consumption and thermal footprint. The RV790 is conceived to be an immediate successor to the RV770 GPU.

View at TechPowerUp Main Site

 

[ShadowFold]

Whoa. So they have 40nm samples going around now?! I can't wait to see these in action!

 

[Odin Eidolon]

4870 with 40nm, same clock speeds? oc'ing monster coming guys!

 

[jbunch07]

NICE! I want to see some of these in action!

 

[OnBoard]

So this would be a countersrike to 55nm GTX 260? Hope it's slower than GTX 280 or at least costs an arm

 

[Flyordie]

flamebaiter ITT...
Anywho-
It will be the counterstrike to the GTX260 216.

 

[OnBoard]

flamebaiter ITT...

If that was meant for me, my comment was only about my freshly bought GTX 280 that is still on the way. You know you don't want something new to be already slow and outdated.

This will be good for ATI, if 65nm to 55nm is anything like 55nm to 40nm it will be a great overclocker.

 

[mikek75]

Bugger, so I'm obselete already, LOL. Oh well, hopefully the 4870 1GB will come down in price and I can go crossfire soon.

 

[Deleted member 3]

So this would be a countersrike to 55nm GTX 260? Hope it's slower than GTX 280 or at least costs an arm

It's just a die shrink, I'm actually hoping they make those power usage claims they made before true.

 

[slyfox2151]

If that was meant for me, my comment was only about my freshly bought GTX 280 that is still on the way. You know you don't want something new to be already slow and outdated.

This will be good for ATI, if 65nm to 55nm is anything like 55nm to 40nm it will be a great overclocker.

the GTX200 has been out for quite a wile now so its not really new in those terms.

 

[DonInKansas]

Bugger, so I'm obselete already, LOL.

This is usually the case with tech stuff. Even when you buy on the bleeding edge you're out of date a week later.

 

[slyfox2151]

This is usually the case with tech stuff. Even when you buy on the bleeding edge you're out of date a week later.

buy an I7 ^^ nothings going to beat those within the next 3 months

 

[ShogoXT]

I heard this has more shaders too, so not just a die shrink?

 

[oli_ramsay]

I heard this has more shaders too, so not just a die shrink?

Mee too, I hope the clock speeds aren't 750MHz though, I can't see it performing much better than a RV770 if it's just a die shrink with slightly faster memory.

 

[OnBoard]

the GTX200 has been out for quite a wile now so its not really new in those terms.

New for me, half year old for rest of the world

 

[WarEagleAU]

It is possible it could clock faster and show more improvement over the 4870 with the die shrink. More shaders are fine, but Nvidia is getting the trick done in shader heavy games with unlocking the shader clock. AMD does something like that and they should have a true winner.

 

[PCpraiser100]

Thoase are great speed improvements, but I think ANMD should really start expanding the PCB's length to get to their overclocking potenital that NVIDIA has right now. I'm waiting for Shader 5.0 cards so I don't really care.

 

[soldier242]

the only bad thing is that qimonda is bankrupt, so i guess there won't be anymore RAM from them

 

[Salsoolo]

buy an I7 ^^ nothings going to beat those within the next 3 months

3 months!? you should be safe with i7 untill westmere man

 

[Haytch]

the only bad thing is that qimonda is bankrupt, so i guess there won't be anymore RAM from them

Depends how this all goes i guess.
Maybe now would be a good time to invest in them

 

[Bjorn_Of_Iceland]

I just noticed that ATi would result to brute force like pumping up clockrates, squeezing more shaders, using smaller fab process and using faster mems. They should once again try to think out of the box and think of unorthodox hardware architecture to jumpstart architectural trend and claim the performance crown. Im thinking that there will come a time where we will reach the smallest fab limit and nVidia having more headroom for improvement and ATi having none.

 

[DarkMatter]

I just noticed that ATi would result to brute force like pumping up clockrates, squeezing more shaders, using smaller fab process and using faster mems. They should once again try to think out of the box and think of unorthodox hardware architecture to jumpstart architectural trend and claim the performance crown. Im thinking that there will come a time where we will reach the smallest fab limit and nVidia having more headroom for improvement and ATi having none.

Yeah that's been true since G80, but IMHO that time will "never" come. I remember the days when they said 32 nm would be the limit, later 22nm and some are saying now we could reach 11nm. We are still at 40nm and IMO by the day 32nm is of wide use the panorama will change again. And I'm just talking about "standard" silicon, that is, the same in use today, with the required changes. Many other things can come and change it more.

Anyway 40nm will probably be a deciding factor, because both will have their next high-end cards around the same time.

 

[Hayder_Master]

more clocks + 40 nm = great card and great overclock
and by the way , can do crossfire with old 4870

 

[Mega-Japan]

So this is expected to be out about when? =O

 

[R_1]

There is a great potential in memory department of HD4870. It's teraflop GPU will preform equally better when core or memory clocks are increased.