• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

GTS 450 folding

I don't think it's that close.
http://www.anandtech.com/show/3909/nvidias-geforce-gts-450-pushing-fermi-in-to-the-mainstream/15

Just over 1/2 the SPs @192 of a 460@336SP's for $129.
Ignore the ram because 768 vs 1gb means nothing for CUDA. link

A 460 768 for $150-$160 is a better buy.

The Rage3D charts don't list the 460. HWC used the client for 24 hrs and HFM.net to average. That's not really apples to apples comparison considering the 450 while installed could have gotten better WU than the 460. Second, the only 450 above the 460 is an OC'd one, take that apples to apples (somewhat) and do stock clock versions and the 460 is ahead. If they put an OC'd 460 on that chart, it would still be #1.

Bottom line - less SP's means less PPD. For only $21 more I'd rather buy a 460. ;)
450 will sell to people mainly because of the word Fermi and 1gb of memory but in f@h performance - $21 more for a 460 would be a no brainer. Sucks too because I was saying to myself the other day I was hoping for another 9600GSO. :eek: Cheap card, awesome PPD. Alas...
 
Good stuff but I really wish these reviewers would state which WU they used to get their numbers.
 
The thing is there is diminishing returns on the number of SPs just like there is in the CPU world.

So if you can buy 4x GTS450's for the same price of 3x 460s. The 4x GTS450s are going to outperform the 460s @ stock and they will probably overclock better too.
 
nitrobass24 said:
So if you can buy 4x GTS450's for the same price of 3x 460s. The 4x GTS450s are going to outperform the 460s @ stock and they will probably overclock better too.
Click to expand...

Hmm....[fusses with chin]
I like the way you think. :D
Very good point....

I will say while looking over AT's article, at first I thought the price said $99 (cause their chart lists the 250 specs and price)...I was about to jump for joy, but then I realized I was looking at the wrong column and the 450 was $129.

But nitro's got a very good point. :cool: (and I didn't need to realize that because I have low funds, CC balance available, and an upgrade itch I'm dying to scratch for more D) :D
 
Ill agree with you zero2dash about the $20. Specially if someone has any interest in gaming. but power usage is only 106 watts so i read somewhere. price/points/watts single 6 pin connector.
 
Too bad there are not any single slot 450's.

7 of these on my SR-2 would go nice :D
 
nitrobass24 said:
Too bad there are not any single slot 450's.

7 of these on my SR-2 would go nice :D
Click to expand...

Are you mad? :eek:

Such talk is just too scary to think about.

On a plus note, does cpu utilization increase with GPU folding performance? By this, I mean whether cpu utilization will be higher for a standard GTX 460 versus a GTX 450. CPU utilization would affect how well your system as a whole does if you are also running a CPU client.
 
amdgamer said:
Are you mad? :eek:

Such talk is just too scary to think about.

On a plus note, does cpu utilization increase with GPU folding performance? By this, I mean whether cpu utilization will be higher for a standard GTX 460 versus a GTX 450. CPU utilization would affect how well your system as a whole does if you are also running a CPU client.
Click to expand...

I know my gtx260 uses up to 5% where my gts250 used 3%. doing simler amount of ppd i would guess them to be close in the case of the 460 vs 450
 
An overclocked 450 does about 9k PPD and an overclocked 460 does about 12k. For $30 more, the 460 is definitely worth it, especially if you want to have more PPD in fewer slots. I don't see the 450 as an attractive folder unless you are really on a budget.
 
looks like a great htpc low energy folder to me.

debating.....
 
The numbers being reported here seem out of whack to me.

Based upon hardware canucks numbers of the stock versions:

the 450 gets about 40 ppd per shader
the 460 gets about 26 ppd per shader?!?

my overclocked 470 get 30 ppd per shader

something's not right
 
yeah if those PPD numbers are correct the 450 is only marginally slower @ folding than the 460
 
From an efficiency standpoint:

450 - 9K ppd / 106W = 84.9 ppd/W
460 - 11K ppd / 130W = 84.6 ppd/W

So they are just about the same. Sounds like a single L5640 machine plus a 450 would make one hell of a HTPC...
 
nitrobass24 said:
The thing is there is diminishing returns on the number of SPs just like there is in the CPU world.

So if you can buy 4x GTS450's for the same price of 3x 460s. The 4x GTS450s are going to outperform the 460s @ stock and they will probably overclock better too.
Click to expand...

I was totally thinking the same way, but this is as close as I can come to making them close.
Not quite the 9600GSO we were looking for as of today. Still very nice for a budget card if you don't have the extra $30.
Then again these are the into prices! Give them a couple weeks and the prices should fall like the 460's did.

450 - $130 x 5 = $650 = 45k PPD
460 - $160 x 4 = $640 = 44k PPD

Sweet Spot price

450 - $110 x 3 = $330 = 27k PPD
460 - $160 x 2 = $320 = 22k PPD

lehmann said:
the 450 clocks higher. legitreviews got a asus card over 1015 core 2030 shader. http://www.legitreviews.com/article/1408/17/
Click to expand...

Then again if you can take the shaders over 1900MHz???? Maybe the 9600 GSO is back!
 
that's an insane shader clock.

what's the highest somebody's gotten a 460?
 
musky said:
1780 here. It wasn't quite stable at 1800.
Click to expand...
I would rather have the 590ghz (lol) total shader power in your GTX460, than the 390ghz total in a GTS450 clocked at 2ghz.
 
lehmann said:
what is your ppd on that thing? and which card did you get?
Click to expand...

It is the eVGA 768Mb one that TigerDirect sells. HFM says 10,998 ppd, and it is extremely consistant. Actual stats have shown on the few days that the card was the only thing I had returning work somewhere around 10300. My guess is that every third day or so you would get an extra unit turned in.
 
musky said:
From an efficiency standpoint:

450 - 9K ppd / 106W = 84.9 ppd/W
460 - 11K ppd / 130W = 84.6 ppd/W

So they are just about the same. Sounds like a single L5640 machine plus a 450 would make one hell of a HTPC...
Click to expand...
Both the 460 and the 450 produce those PPD numbers while overclocked, and you're using their stock TDPs for comparison. That doesn't give you an accurate representation. I suspect that when overclocked, a 460 would produce better PPD/W.
 
Zero82z said:
Both the 460 and the 450 produce those PPD numbers while overclocked, and you're using their stock TDPs for comparison. That doesn't give you an accurate representation. I suspect that when overclocked, a 460 would produce better PPD/W.
Click to expand...

The 460 numbers are actual, from a kill-o-watt and HFM on one of my machines with the card was running 1800 shaders. The 450 numbers are from this thread. I don't have a 450, so I am using what I am told.
 
That's really cool that a GTS450 can do 9K per day. How does it compare to a GTX260 on PPD?
 
The shaders in my Sparkle GTX 460 are OC'd to 1680 and I get 11-14K PPD depending on the wu.
 
DooKey said:
The shaders in my Sparkle GTX 460 are OC'd to 1680 and I get 11-14K PPD depending on the wu.
Click to expand...

Really? Mine never budges off of 10998 ppd, ever. I have only run P10632's, though. Is there something I should be setting on the GPU client to get other units?
 
I'm running -advmethods. For example I'm currently running a 10933 and I'm pulling 13.7K ppd on this one.
 
DooKey said:
I'm running -advmethods. For example I'm currently running a 10933 and I'm pulling 13.7K ppd on this one.
Click to expand...

Interesting...I am such a n00b... :)

Adding that flag right now.
 
musky said:
The 460 numbers are actual, from a kill-o-watt and HFM on one of my machines with the card was running 1800 shaders. The 450 numbers are from this thread. I don't have a 450, so I am using what I am told.
Click to expand...
How did you use a Kill-A-Watt to measure the consumption of the 460? You can't measure individual components with it.
 
Zero82z said:
How did you use a Kill-A-Watt to measure the consumption of the 460? You can't measure individual components with it.
Click to expand...

Take a reading without the video card, then take a reading of the same system with the video card running. I actually got the 130 from not running a GPU client and then running a GPU client. It is more of an effective power draw, since you would need a video card in a system anyway.
 
musky said:
Take a reading without the video card, then take a reading of the same system with the video card running. I actually got the 130 from not running a GPU client and then running a GPU client. It is more of an effective power draw, since you would need a video card in a system anyway.
Click to expand...
Well, that doesn't tell you how much the card actually consumes since idle power isn't taken into consideration. You definitely can't compare it to the 450's TDP. Also keep in mind that your numbers are measured from the wall with a somewhat inaccurate measuring device, so the actual GPU power consumption will be lower due to PSU inefficiency.
 
musky said:
The 460 numbers are actual, from a kill-o-watt and HFM on one of my machines with the card was running 1800 shaders. The 450 numbers are from this thread. I don't have a 450, so I am using what I am told.
Click to expand...

106 watts is stock. nvidia reference.
 
Zero82z said:
Well, that doesn't tell you how much the card actually consumes since idle power isn't taken into consideration. You definitely can't compare it to the 450's TDP. Also keep in mind that your numbers are measured from the wall with a somewhat inaccurate measuring device, so the actual GPU power consumption will be lower due to PSU inefficiency.
Click to expand...

What is the TDP on this card? I just look at it from a practical perspective - how many watts are required to fold with this card....and that answer is 130. I would need a video card running anyway, so idle power draw is pretty much irrelevant. So how would I measure actual power draw? Every review I have seen take a full system power draw reading.
 
musky said:
What is the TDP on this card? I just look at it from a practical perspective - how many watts are required to fold with this card....and that answer is 130. I would need a video card running anyway, so idle power draw is pretty much irrelevant. So how would I measure actual power draw? Every review I have seen take a full system power draw reading.
Click to expand...
I don't remember what the 460 TDP is. Some reviewers have equipment that allows them to measure the power consumption of the video card directly. Anyway, my point is that to compare power consumption, you need to measure both cards the same way. You can't make assumptions like you were doing.
 
You must log in or register to reply here.
Back
Top