New msi 5770 Hawk.. one fast overclocking 5770

[blanketyblank]

not exactly hard to get a 5870 (or even a 5850 with better cooling) to 1GHz+... how would these beat up a single card with the same clocks?

He said crossfired if you missed that part. Actually right now 2 stock 5770s can already beat out a 5870 at least in terms of average fps. So it certainly wouldn't be surprising two of these things OCed can beat out even a maxed OCed 5870.

 

[bunnyfubbles]

He said crossfired if you missed that part. Actually right now 2 stock 5770s can already beat out a 5870 at least in terms of average fps. So it certainly wouldn't be surprising two of these things OCed can beat out even a maxed OCed 5870.

I used the word "these" which is plural and thus inferring I was referring to two 5770s.

while the CF 5700s can sometimes beat out their single card/GPU bigger brother 5800s, that is a best case scenario, the majority of games won't see the same results.

 

[dug777]

I used the word "these" which is plural and thus inferring I was referring to two 5770s.

while the CF 5700s can sometimes beat out their single card/GPU bigger brother 5800s, that is a best case scenario, the majority of games won't see the same results.

http://www.xbitlabs.com/articles/video/display/radeon-hd5770-hd5750-crossfirex.html

This set of benchmarks begs to differ

 

[MrK6]

http://www.xbitlabs.com/articles/video/display/radeon-hd5770-hd5750-crossfirex.html

This set of benchmarks begs to differ

No, they really don't. If you read through the benchmarks, you'll notice that the 5770's start to lose out as you increase the resolution/image quality and they many times have lower minimum FPS. That's just the nature of the beast. While the 5770's scale incredibly well in CF (it's actually ridiculous how well they work, looks like 80-95% in almost every game), they still aren't a 5870. The horsepower of a 5870 at 1GHz+ is a force to be reckoned with .

That said, in response to the article, this MSI 5770 isn't that impressive. In all honestly, most 5770's can do 1GHz+ (some need a little voltage). The fact that MSI crippled the idle power savings of the card would drive me away from this product. I just built a rig with a reference design based HIS 5770 and through 15 minutes of testing got a stable overclock of 960MHz/1350MHz (Overdrive limit on the core), no voltage adjustments. This is with the egg cooler, and the card loads at ~55C. With a little voltage I'd see the card having no problem breaking 1GHz and maybe even 1.05GHz.

 

[bunnyfubbles]

http://www.xbitlabs.com/articles/video/display/radeon-hd5770-hd5750-crossfirex.html

This set of benchmarks begs to differ

on AT's 5700 article the CF 5770s split evenly with the 5870.

When we consider either article uses the most popular games which are thus bound to be specifically targeted and optimized for by AMD, well, good luck getting the same level of scaling across all games.

I confidently stand by my original comment. Besides, if we go back to the root of my inquiry it starts with the OP claiming the OCed 5770s would "beat up on" the 5870. Even in your example of xbit's tests whenever the 5870 loses its almost always by a negligible margin, and even then its usually beating the 5770s in minimum frame rate. Not what I would call being "beat up"

 

[dug777]

No, they really don't. If you read through the benchmarks, you'll notice that the 5770's start to lose out as you increase the resolution/image quality and they many times have lower minimum FPS. That's just the nature of the beast. While the 5770's scale incredibly well in CF (it's actually ridiculous how well they work, looks like 80-95% in almost every game), they still aren't a 5870. The horsepower of a 5870 at 1GHz+ is a force to be reckoned with .

That said, in response to the article, this MSI 5770 isn't that impressive. In all honestly, most 5770's can do 1GHz+ (some need a little voltage). The fact that MSI crippled the idle power savings of the card would drive me away from this product. I just built a rig with a reference design based HIS 5770 and through 15 minutes of testing got a stable overclock of 960MHz/1350MHz (Overdrive limit on the core), no voltage adjustments. This is with the egg cooler, and the card loads at ~55C. With a little voltage I'd see the card having no problem breaking 1GHz and maybe even 1.05GHz.

He said 5800.

5770 cf Compared to the 5850 at 19 maxed

wic loses on min
crysis win
ut win
lp win
stalker win
fc2 win
cod win
wolf win
bf win
stormrise win
hawx win
cojbib win
res win
borderlands no
l4d2 win

 

[MrK6]

He said 5800.

5770 cf Compared to the 5850 at 19 maxed

wic loses on min
crysis win
ut win
lp win
stalker win
fc2 win
cod win
wolf win
bf win
stormrise win
hawx win
cojbib win
res win
borderlands no
l4d2 win

Ah, so you want to compare the best of the 57xx series to the worst of the 58xx series. Even then, if you overclocked both, the 5850 would pull ahead.

 

[happy medium]

Ah, so you want to compare the best of the 57xx series to the worst of the 58xx series. Even then, if you overclocked both, the 5850 would pull ahead.

But the 5770's cost 80$ less!

 

[dug777]

But the 5770's cost 80$ less!

Good point.

In any event I don't think it is an unreasonable comparison. Quite clearly the cf 5770s are faster than a 5850 at stock speeds, which is anexplicit rebuttal of the original comment that I responded to many many people only run at stock and overclocking is a variable from card to card further muddying those waters

no intention of ruffling feathers or distressing people, peace folks

 

[cbn]

How much more money is this Hawk HD5770 going to cost?

I like these exotic PCB designs...and I am glad companies have the resources to do this, but an ASUS Voltage tweak 5770 (that does 1000 Mhz+) has the same MSRP as any other garden variety 5770.

P.S. I realize I am not taking into account noise levels when I think of this comparison. The Hawk has a nice Twin Frozr Cooler.

 

[MrK6]

But the 5770's cost 80$ less!

$80 less than a 5870 or a 5850? I'm guessing the former, and you get less performance for the price. I'm just saying Crossfiring lower cards gives a false impression that they'll give a better experience than a single, high-end GPU. This is definitely not the case, especially at higher resolutions. Consider it fair warning.

 

[cbn]

$80 less than a 5870 or a 5850? I'm guessing the former, and you get less performance for the price. I'm just saying Crossfiring lower cards gives a false impression that they'll give a better experience than a single, high-end GPU. This is definitely not the case, especially at higher resolutions. Consider it fair warning.

In the past X-firing lower end cards usually meant putting up with 512MB of VRAM.

These hd5770s, however, have 1 GB VRAM each.

 

[RussianSensation]

Review of MSI Radeon 5770 Hawk:
http://www.techpowerup.com/reviews/MSI/HD_5770_HAWK/

Overclocked in COD4 it caught up to 4890:
http://www.techpowerup.com/reviews/MSI/HD_5770_HAWK/33.html

Very quiet fans

 

[edplayer]

When we consider either article uses the most popular games which are thus bound to be specifically targeted and optimized for by AMD, well, good luck getting the same level of scaling across all games.

I don't understand that.

Why would the "most popular" games be optimized for AMD when Intel and Nvidia both have larger market shares (and have had for a long time)?

 

[happy medium]

$80 less than a 5870 or a 5850? I'm guessing the former, and you get less performance for the price. I'm just saying Crossfiring lower cards gives a false impression that they'll give a better experience than a single, high-end GPU. This is definitely not the case, especially at higher resolutions. Consider it fair warning.

Less performance?????? How high resolutions? 1920x1080 isn't enough @ high detail?

12 wins, 1 lose and 1 tie.

5770 in crossfire @ 1920x1080 8x aa
Beating 5850's in crossfire in Far Cry 2..
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/5

Tied a 5870 in Anno 1404
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/6

Beating a gtx 295 and wooping a 5870 in Call of duty 5 WAW @ 1920x1080 4x aa
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/7

Beating a gtx 295 and 5870 easily in Hawx @ max settings.
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/8


Tied a 5850 and losing to a 5870 in Crysis Warhead @1900x1080
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/9

Woopen a 5870 and beating a gtx295 in Fallout 3 @ 1920x1080 ultra detail.
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/10

Beating both the 5870 and gtx 295 in Brothers in Arms @ 1920x1080 High settings.
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/11

Beating a 5870 in Resident Evil 5 @ 1920x1080 and 2560 x1600..
http://www.guru3d.com/article/radeon-hd-5770-in-3way-crossfirex-review-test/12

Beating a 5870 in Mass Effect @ 1920x1080 max detail
http://www.guru3d.com/article/radeon-hd-5750-review-crossfirex/17

Wins in modern warefare 2... @1920x1080
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_8.html#sect1

Wins in Left for Dead 2..@1920x1080
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_9.html#sect0

Wins in WOlfinstein...@1920x1080
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_9.html#sect0

Wins in Street Fighter 4 @1920x1080
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_10.html#sect1

Wins in World in Conflict @ 1920x1080
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_13.html#sect1

DID I MISS ANY GAMES!

All that (stock) performance for under 310$ shipped (under 300$ with rebate) at the Egg when you buy 2.
http://www.newegg.com/Product/Produc...82E16814161317

Oh you wanna talk minimum framerates??? The 5770 crossfired is better in most cases!!! In fact a 5770 and 5750 is equal to a 5870.
Take a look at this review of a frankenfired 5770 with a 5750.
Some of the results also gives minimum framerates.
http://www.xbitlabs.com/articles/video/display/asus_eah5770-powercolor_pcs_hd5750_8.html#sect1


5750 in crossfire?

Beating a 5870 in 3 tests and 1 tie out of 8 games for 260$ shipped.
http://www.guru3d.com/article/radeon-hd-5750-review-crossfirex/13

Conclusion written by XBIT labs in there review..

Quote:
Considering the results of this and earlier test sessions with Radeon HD 5770 and HD 5750, including those at overclocked frequencies, purchasing two inexpensive Radeon HD 5750 cards and overclocking them to 850/5000MHz seems to be the most profitable option.
At such frequencies, the cards totaling $250-260 will be faster than a single Radeon HD 5870 which currently costs $400 or more.



Crossfire scales very good now (cheaper) and with most.., if not all popular games.

 

[blanketyblank]

if you haven't kept up with prices a 5870 is around 400
5770 crossfired is around 300
5850 is around 300

So considering the 5770 (not OCed yet) crossfired can usually match or beat the 5870 at least in average fps they make a very decent deal since they are about $100 cheaper.
Their price is about the same as a 5850 (depends on brand and sale) and in which case I've never seen them losing. I agree there are issues when using a multi gpu setup, but we're just talking speed here.

In terms of comparing these hawk cards though it's probably more fair to put them up against the 5870 since they seem to cost closer to $200 each.

 

[MrK6]

Less performance?????? How high resolutions? 1920x1080 isn't enough @ high detail?

*snippage*

Crossfire scales very good now (cheaper) and with most.., if not all popular games.

Have you used Crossfire? Or a 5850 or 5870 for that matter? There's a reason for the premium a high-end single GPU carries - you get what you pay for. Crossfire isn't the same as a single GPU nor does it perform like one, no matter how much some people would like to believe it is. Canned benchmarks and timedemos don't really show the differences, but try playing an actual game and they're blatant. The sad thing is people read these reviews and run out and buy two 5770's instead of a 5850 and don't realize what they're missing. I suppose I could also throw in that since these cards all overclock to the 1000-1050MHz range on air and then crap out, it makes the 5850 an even better value.

 

[happy medium]

Have you used Crossfire? Or a 5850 or 5870 for that matter? There's a reason for the premium a high-end single GPU carries - you get what you pay for. Crossfire isn't the same as a single GPU nor does it perform like one, no matter how much some people would like to believe it is. Canned benchmarks and timedemos don't really show the differences, but try playing an actual game and they're blatant. The sad thing is people read these reviews and run out and buy two 5770's instead of a 5850 and don't realize what they're missing. I suppose I could also throw in that since these cards all overclock to the 1000-1050MHz range on air and then crap out, it makes the 5850 an even better value.

Yes as a matter of fact I do use crossfire and it runs just fine.
Much better then my 4830 crossfire setup I had. The 5xxx series seems to have solved those old problems.

Do you use crossfire ?
No ,you just happen to have a 5870 card, figures.

edit what happen to your minimum fps argument? figures

 

[soccerballtux]

Have you used Crossfire? Or a 5850 or 5870 for that matter? There's a reason for the premium a high-end single GPU carries - you get what you pay for. Crossfire isn't the same as a single GPU nor does it perform like one, no matter how much some people would like to believe it is. Canned benchmarks and timedemos don't really show the differences, but try playing an actual game and they're blatant. The sad thing is people read these reviews and run out and buy two 5770's instead of a 5850 and don't realize what they're missing. I suppose I could also throw in that since these cards all overclock to the 1000-1050MHz range on air and then crap out, it makes the 5850 an even better value.

Have you used crossfire?
In my opinion Crossfire/SLI are best for games that already can get ~40FPS normally on one card-- that way when you dip down to minimum FPS, you're still getting 30fps/card. This minimizes the latency.

 

[cbn]

One thing that interests me about Crossfire/SLI is being able to run the "first issue" cards on the newest process together.

We saw that when HD4770 when it debuted. If it wasn't for the 512 MB VRAM it would have done even better against HD4890.

Maybe ATI will do a repeat of the HD4770 phenomenon with a 28nm-32nm card?

 

[3DVagabond]

I've seen too many times where Xfire and/or SLI don't work with a particular game until the drivers are optimized for it. Latest example Dirt 2 with the GTX295. I'd rather have a single GPU that meets a certain performance level than rely on 2x GPU and hope all my favorite games gets optimized for.

 

[MrK6]

Yes as a matter of fact I do use crossfire and it runs just fine.
Much better then my 4830 crossfire setup I had. The 5xxx series seems to have solved those old problems.
Do you use crossfire ?
No ,you just happen to have a 5870 card, figures.
edit what happen to your minimum fps argument? figures

Not any more, wouldn't touch it unless I absolutely had to. I've tried it in each of the last three AMD generations as well as SLI in two of the last three: multi-GPU is no replacement for single GPUs. Sorry to rain on your parade, but that's the state of things. And the minimum FPS argument still stands - single GPUs give a much smoother experience. Now if you can find some actual gameplay that backs up your opinion, it will be worth something.

Have you used crossfire?
In my opinion Crossfire/SLI are best for games that already can get ~40FPS normally on one card-- that way when you dip down to minimum FPS, you're still getting 30fps/card. This minimizes the latency.

See above. Also, if you had gotten the more powerful single card, you'd have 60FPS constant and wouldn't have to worry about drivers, frames dropping, etc. I'm not knocking multi-GPU, it's currently the best solution we have to get maximum performance. Considering the high-end, it's the only way to add more performance once you've gotten the fastest single GPU card. I am, however, saying that going out and buying two 5750's or two 5770's instead of a 5850 or a 5870 is a poor decision. Don't like it? Don't think I'm correct? It's a free world, do what you want, I could care; just hope you never see a high-end single GPU in action and realize what you're missing.

 

[happy medium]

:biggrin:Only thing I'm not missing is 100$.:biggrin:

It's all good MRk6.:thumbsup:

 

[MrK6]

:biggrin:Only thing I'm not missing is 100$.:biggrin:

It's all good MRk6.:thumbsup:

No worries . I don't mean to sound like a dick, just trying to relate my experience :thumbsup:.

 

[dug777]

No worries . I don't mean to sound like a dick, just trying to relate my experience :thumbsup:.

Personally I would take a 5850 in heartbeat over cf 5770s

cf 5770s does remain an excellent value and absolute performance solution, but I am happy to pay a premium to enjoy any relative 'hassle free' benefits of a single GPU option. That says more about my wish for a simple life than anything else I suppose