Instead of these not well cooled cards, many silencers pay more attentions to PowerColor's Green Series in recent months. Shall we have a chance to verify them?

http://www.powercolor.com/Global/products_features.asp?id=214
http://www.powercolor.com/Global/products_features.asp?id=236
http://www.powercolor.com/global/products_features.asp?id=237

Please correct the title typo. It's not another passive 5750 (unfortunately), but just a 5570.
I think HIS went cheap and they'll lose on these cards. IMO an S2 style cooler (with two heatpipes) would give much lower temps.

Yes. And the fins need to run left to right just like on a S2 or S1 as well so that it sits along the airflow path through the case.

I don't get, why they recommend a 400W power supply for a 40W card? Sure the rest of the system won't need so much power if it matches these "lowly" graphic cards.

Are we going to get a 5770 review or maybe even a shootout between different card coolers anytime soon?

Keep up the good work!

I think I recall reading a footnote somewhere mentioning the 400W PSU recommendation is based off of a system with a high TDP quad-core CPU + 4 HDD's. They also need to account for low-quality power supplies (whose power ratings are outright lies).

Manufacturers have no idea what components buyers will be using with their graphics cards so they need to cover their bases and recommend higher than necessary.

Thanks for the review. Big delta in power use between the ATI reference 5570 and this HIS version. I wonder why?

Maybe the GDDR5 on the HIS

One wonders if a 120mm Nexus/Shiruken/Noctua fan run at 7v or so would drop the temps down significantly? Wouldn't be that hard to jerry-rig a bracket to mount a fan close to the HS on it. I thought the main problem with fans on GPUs was that they were typically 50mm screamers...

Putting a fan on a passive card works wonders.

I put a 70mm 2000RPM fan on my old Geforce 6200 a couple years back, and the temps dropped 20c compared to passive

WTF! I use a ZM80D-HP on my 7900GS and it works fine with just a vent over the PCI slots. I just played on a 30C day and it went up to 73C during games. These heatsinks aren't so great in comparison. That being said the HIS 5770 cooling looks better than usual. I wish some of the different 5770 or 50s were reviewed.

And when the maker's legal dept intervenes it seems that a safe margin is calculated as "real power draw in full stress x 8, but no less than 450W for a single card w/o PCI-E 6pin and 600W for a card with 1 PCI-E 6pin".

This is fun, ok, but it gets serious when a review site (say, guru3d) starts recommending 400W PSUs for such cards. And sites like that tend to use massively OCed quad cores... That's when you know that SPCR is THE site.

so the big gain is 40nm...

all other measurements have been around for 5 years (since the prescott)

..and that is a good point about power needed.
altho 400 is realistic thinking of a modern system, 400 was once recommend for a voodoo and a p2.

today alot more is getting done.



I still want to see one more AGP round go about... in the 45-40nm range.

Why? What's the point of supporting a badly designed hack of a bus when PCI-E is available?

How will they do that?

Low quality power supplies have variances beyond belief.
Some can deliver 80% of their label wattage, while some can only do.....this.
http://www.jonnyguru.com/modules.php?na ... ry&reid=71

IMHO, it is up to the customer to do some research of their own.
If they get burned because they thought a $15 PSU can do 700W, then they have no one to blame but themselves.

I think you two are in agreement and just don't realize it

The argument goes like this:
NVidia and AMD know that when the user buys a new card, plugs it in, and then the PC goes into a reboot cycle, BSOD, or some other classic PSU load tailspin, it's 50%/50% whether the user blames the card or the PSU. So, that is a HUGE impact on return rates and hence, profitability. By over-rating the heck out of PSU wattage req., they are lessening the chance that a customer (with the most shitty PSU) won't have a PSU problem. Then, they won't get hit by false returns.

So, it's a CYA to lessen their false bad product returns.

I've got the PowerColor Green 5570 (#237 in the above list; got it from NewEgg), and I'm happy with it. It generally idles below 50 degrees, and high activity may push it into the 60's, but I've never seen it go higher. My PC has steady low to moderate airflow (built to be quiet with components that shouldn't require excessive cooling). I had expected higher temperature numbers (this is cooler than the previous card, an ATI 4350). So either this is truly cool, or the airflow is working optimally (which I wouldn't expect, given the down-facing heat sink which probably isn't in an airflow path), or the temperature sensor is low. Yeah, I'd like to see a review with some hard numbers. Does this "green" 5570 consume fewer watts, and thus dissipate less heat, than the HIS 5570?

By the way, on the PowerColor "green" line...in the 5570, the heat sink fins run along the height of the card, but in the 5670 and 5750, they run along the length of the card, and the heat sink is heftier. So if airflow is from the front of the PC, it seems unlikely to flow through the fins of the 5570 and more likely on the higher models; instead it will bump into the end fin of the 5570. I don't know how much this matters. As I said, the heat numbers look good, so I can't say it's not working.