[TPU] Sapphire HD 4730 512 MB Review - Page 2 - Overclock.net

Kuntz · 07-01-09

Quote:

Originally Posted by el gappo

but the higher card downclocks to match the lower one so it is pretty pointless. just get 2 of these at 1ghz cant go wrong

That is not true at all. When you run two different cards, they maintain their proper clock rates, and you retain all the performance of the fastest card.

skugpezz · 07-01-09

Quote:

Originally Posted by videoman5

So, can one crossfire a 4730 with a 4830?

yes they can be crossfired, the card with the faster ram clcok will donclock the ram only

Kamikaze127 · 07-01-09

I think... this card is fail...

skugpezz · 07-01-09

Quote:

Originally Posted by Kamikaze127

I think... this card is fail...

we have same thoughts it fails, but I like the price though

brettjv · 07-01-09

Quote:

Originally Posted by Kuntz

That is not true at all. When you run two different cards, they maintain their proper clock rates, and you retain all the performance of the fastest card.

All cards maintain their clocks in xfire (unless you manually change them), but there is no possible way that one fast card (A) + one slower card (B) in xfire = the same performance as A + A (i.e. the same as two of the faster cards).

What you net by doing A + B is something that's more like an average between the two cards performance. Card A makes a frame as fast as it can, then card B makes a frame as fast as IT can.

To put it another way, if you had two of the same card, one of them at 1000MHz, and the other one at 600MHz, you would get approx. the same xfire performance as you'd see with both cards at 800MHz.

At least, this would be the case with AFR. Now, some of the other modes, you may well see the same perf as you'd see with both cards being at 600MHz. But you'd never get perf like both cards being at 1000MHz, that's for certain.

Kuntz · 07-01-09

Quote:

Originally Posted by brettjv

All cards maintain their clocks in xfire (unless you manually change them), but there is no possible way that one fast card (A) + one slower card (B) in xfire = the same performance as A + A (i.e. the same as two of the faster cards).

What you net by doing A + B is something that's more like an average between the two cards performance. Card A makes a frame as fast as it can, then card B makes a frame as fast as IT can.

To put it another way, if you had two of the same card, one of them at 1000MHz, and the other one at 600MHz, you would get approx. the same xfire performance as you'd see with both cards at 800MHz.

At least, this would be the case with AFR. Now, some of the other modes, you may well see the same perf as you'd see with both cards being at 600MHz. But you'd never get perf like both cards being at 1000MHz, that's for certain.

You are correct. Most people think that the faster cards are "nerfed" down to the slowest card, and that performance is "wasted". This is not the case.

By OCing my "fastest" card (check sig rig), my frame rates and 3dmark scores increase accordingly. This proves, beyond any doubt, that your fastest cards clock speeds do matter, and aren't nerfed or limited in any way.

vicious_fishes · 07-01-09

ati,

please, PLEASE make a 1gb version of something in the 4700 series

sincerely,

guy with a psu that couldn't power a lightbulb.

Ihatethedukes · 07-01-09

Quote:

Originally Posted by d3v0

It's kind of a mess haha. Its a power guzzling loud ....9800GT??????

Did you even read the review? It's trading blows with the 9800GTX most of the time, under load, uses less power than both. It averages comparably. Idle looks a little less than stellar I agree. The 9800GT is no quiet experience either. I still don't understand why they'd say 'get a 9600GT first'. Writers on crack?

Gr3m1in · 07-01-09

Quote:

Originally Posted by videoman5

So, can one crossfire a 4730 with a 4830?

Did you even read the article?

Quote:

You may combine up to four HD 4730 cards in CrossFire for increased performance or higher image quality settings. There is no official statement whether you can mix the HD 4730 with other RV770 based cards. I tried with the HD 4830, HD 4850, HD 4870 and HD 4890 using Catalyst 9.6 under Windows Vista and could not get a CrossFire setup established.

sLowEnd · 07-01-09

Quote:

Originally Posted by Ihatethedukes

Did you even read the review? It's trading blows with the 9800GTX most of the time, under load, uses less power than both. It averages comparably. Idle looks a little less than stellar I agree. The 9800GT is no quiet experience either. I still don't understand why they'd say 'get a 9600GT first'. Writers on crack?

It uses more power idle than a 4890 and 10 more watts than the 9800GT at load

The 9800GT is practically inaudible at idle whereas the 4730 is loud at both idle and load (And runs hot too despite that)

The 4730 loses to the 9800GT in
-Call of Duty 4
-Company of Heroes
-Enemy Territory: Quake Wars
-Far Cry
-FEAR
-Quake 4
-Team Fortress 2
-Unreal Tournament 3
-World in Conflict

Average performance is worse
http://www.techpowerup.com/reviews/S...D_4730/28.html

Quite a big list there

EDIT: And 2x 6pin connectors? What's up with that?

		Overclock.net - Overclocking.net > Industry News > Hardware News
[TPU] Sapphire HD 4730 512 MB Review

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)