Performance of the ATI 4830? 3dmark scores?

Hey everyone,

I just bought an Envy 15 with the i7 + 4830 in a 1" chassis. Yeah, I'm pretty excited. But I was wondering... what kind of graphics performance am I looking at? Am I going to be able to play pretty much all games fluidly at 1080p (on medium settings for the harder games...)? Could the 4830 run Crysis on medium/low settings at 1080p? Do we have a 3dmark06 score for the 4830?

Thanks for the input!

 

Only score ive seen is 7200 3DMark06 with an i7-820QM. No details about the res or score breakdown. I can't find the link but it was in the HP Envy thread.

Seems way lower than it should be, but I will wait for more tests before forming conclusions.

 

Yeah, I saw those scores to. It seems a little low. I'm hoping that it performs closer to the 9000+ mark with updated drivers and maybe a drop in resolution if that 7200 was for a 1080p run. **Quadron crosses his fingers and prays to the sweet mother of ....

 

Well, with GDDR5 it should score near the 10k mark, but if it is equipped with GDDR3, then it most probably scores between 7-8k depending on clockspeeds.

Basically, the GDDR3 version would be an HD4670 with twice the shaders, same memory performance, so games will run great on lowish resolutions around the 1366x768-1600x900. Higher resolution will impact greatly.

 

There is no GDDR5 version for the HD 4830.

 

Just the 4860 is GDDR5, it will perform similar to any mid range card with a 128-bit bus, try keeping the resolution lower.

Most of the 3dmark score will be the i7, those things boost 3dmark scores a lot.

 

Sadly, the 4830 is the reason I passed on the envy. It should still do pretty well though on recent games with medium resolution and quality settings. Actual benchmarks are very, very sparse.. So be sure and post yours when it arrives!

 

The 4850 had relatively bad 3dmark06 as well, worse than the gtx260m, but far better 3dvantage (or w/e it's called). It's generally accepted that 3dmark06 takes CPU too much into account, so it may be worth thinking of other benchmarks for this. But if I had to estimate, I'd guess performance between GTS260m and GTX260m level. So you could run all games for next few years at lower res very fluently, just don't expect to be able to do high res + high eye-candy with fluid framerates on graphics-intensive games.

 

Here, look at these tests, then take a little bit off, as you core is running 50MHz slower, and your memory 200MHz slower.

You could make up that ground by overclocking.

It's a very respectable GPU in the mobile world.

 

wait 7200marks @ 1080p? thats good, i manage ~ 9000 with my rig. @ 1080p. 7200marks for 1280x1024 is going to appalling tho.

^completely invalid man....those are desktop gpus.

 

And the Mobility 4830 is taken directly from the desktop version of the 4770. Same core, same everything, just DDR3. So, essentially it is a 40nm 4830.

 

Yeah, those are desktop GPU's, which is going to be a lot different.

I have a feeling that we're going to be getting some reviews of the Envy 15 soon. I mean, the reviews for the Envy 13 have been for so long, and I guess several of the stores actually have the Envy 15 in stock now. **Crossing my fingers for a review... lol

 

Thats the difference, the GDDR3, it puts it at a completely different level than the GDDR5 version, the doubled memory speeds negates the 128bit bus, effectively making it a 256bit card.

 

If the 4830 is taken directly from a 4770, then why didn't you point me towards benchmarks for the 4770... but tell me they would be lower because of the memory... lol

It's cool. I guess I can wait my lappy to come in next week. I'll survive just the same.

 

- I linked to a review of the desktop 4830.

- The Mobility 4830 will be very similar, as the only difference between the 4770 and 4830 are the core size and the GDDR5. So, the Mobility 4830 is a GDDR3 4770, yes? Then direct correlation can and should be be drawn from the desktop 4830.

There is really nothing to disagree with.

Because -50MHz/-200MHz on the core is easier for me to discuss than GDDR5 vs GDDR3.

 

Okay, that makes sense.

 

So how much would you say needs to be taken off for these to correlate to the mobile version? Estimated 3dmark06 score drop? Estimated average FPS drop (I know this will vary a lot)?