• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

So ladies and gents, was fermi a fail after all?

deusexlumina

deusexlumina

deusexlumina

Soldato
Joined
12 Jun 2008
Posts
3,011
I remember a year and a bit back when the fermi architecture was getting a right grilling (excuse the pun :p ) due to its large die size, high power consumption and resultedly high heat output. Many people were saying that if AMD/ATI stayed on top of thier game, then fermi would be dead in the water within a few months.

As we all know, this isn't true, and if you take a look at the lineup now there is some amazing value for money when you compare it to the GPU market a few years back. The overclocking potential of the cards is also surprisingly high, with the 460gtx and 560ti getting an honourable mention there.

My personal view is that nvidia really did a good job out of making the best of a bad lot, and thier reputation is certainly growing once again, if at least in the discrete GPU community. But what do you guys think, was the fermi architecture a success after all or have nvidia just swept thier mess under the bed so to speak?

If a mod could add a poll, that would be great :) no flamewars etc, civilised discussion please.
 
Last edited:
Won't say they were a failure by any stretch of the imagination but they're way too hot and loud for my taste.

The MSI 560 today at under £200 is incredibly tempting for someone like me who values low noise.
 
They are flawed in some areas, power consumption, they can run hot/loud, but ive been more than pleased with the two 470's i bought. Especially as they were pretty cheap and they overclock extremely well.
 
It's a mixed barrel. The initial Fermi chips were what you would call fail; extremly demanding in terms of power, extremely expensive, and as I remember not quite as fast as Nvidia wanted us to believe, on top of being late.
It's taken the massive price cuts (read: losing Nvidia cash) to clear the original Fermi chips; so in that sense, Fermi is a bit of a fail.
That said, the revised Fermi chips aren't doing too badly, but it DID take a revision of Fermi to get it where it is now, the chips earning Nvidia some returned credence is these revised chips by the by, rather than the originals.

I'm glad they've managed to return to some degree though; having only one prospective choice in the market really puts a dampener on competition and ups the prices. In terms of business sense, ATI's move still seems to have made more sense as they've likely made much higher overall profit to cost than Nvidia have done this time round, but its not a whitewash as it could have been.
 
I would say yes, mainly because they were very late and gave AMD a big head-start in the DX11 market; additionally, they ran very hot and loud and fell short of the performance many of us expected.

With that said however I don't think they were the disaster that some people painted them to be.
 
AMD couldn't even topple the 570/480 with their fastest single GPU card, fermi a fail be it 400 or the 500 cards? never, complete success.
 
Alexrose1uk said:
It's a mixed barrel. The initial Fermi chips were what you would call fail; extremly demanding in terms of power, extremely expensive, and as I remember not quite as fast as Nvidia wanted us to believe, on top of being late.
It's taken the massive price cuts (read: losing Nvidia cash) to clear the original Fermi chips; so in that sense, Fermi is a bit of a fail.
That said, the revised Fermi chips aren't doing too badly, but it DID take a revision of Fermi to get it where it is now, the chips earning Nvidia some returned credence is these revised chips by the by, rather than the originals.

I'm glad they've managed to return to some degree though; having only one prospective choice in the market really puts a dampener on competition and ups the prices. In terms of business sense, ATI's move still seems to have made more sense as they've likely made much higher overall profit to cost than Nvidia have done this time round, but its not a whitewash as it could have been.
Click to expand...

This, is absolutely spot on.
 
The original fermi was an abject failure, the 470 was more expensive, hotter, louder and slower than 5870.

What has followed though starting with the 460 has been great.
 
Love my 480 to bits.

It has character its a bit noisy when playing games but its hardly deafening. Temperature's are absolutely fine and the performance you can not complain about. Power consumption should not be an issue with an enthusiast forum surely? this not green peace! :D

It was a bit of a fail with original prices and it being late but thats about it.
 
Love my 480 to bits.

It has character its a bit noisy when playing games but its hardly deafening. Temperature's are absolutely fine and the performance you can not complain about. Power consumption should not be an issue with an enthusiast forum surely? this not green peace!

It was a bit of a fail with original prices and it being late but thats about it.
Click to expand...

Power in itself isn't an issue; but it limits you in a lot of ways. Producing a more efficient, less power consuming card is better all round, even for enthusiasts.
The revised Fermi deals with this; cooler, less energy consumption, lower cooling requirements (and thus potentially quieter coolers) and enabling higher clocks and sometimes overclocking potential; so it's benefitted enthusiasts overall.

The revised Fermi has also directly benefitted mobile tech enthusiasts, as they've been able to bring out faster cards based on it than they could within reason [ahem :D] based on the original Fermi. The 470m based on the revised Fermi is roughly the same speed as the original Fermi based 480m, whilst costing less, running cooler and absorbing less juice as I understand. That's a progressions in and of itself. The very most expensive dual 485m SLI sets now available are extremely fast; and whilst massively expensive compared to desktop equivalents are pushing performance in the same region as the desktop 480/570/580. The original Fermi design just wasn't up to this in a DTR/mobile chassis.


The original Fermi GTX480 would seem to parallel a lot with ATIs 2900XT. Not a bad card at all; but not what it should have been.
 
Last edited:
straxusii said:
The original fermi was an abject failure, the 470 was more expensive, hotter, louder and slower than 5870.

What has followed though starting with the 460 has been great.
Click to expand...

It wasn't up against the 5870 tho product catagory wise - but it still gives the 5870 a hard run for its money.

Fermi has its negatives - but its far from the completely broken, unfixable, architecture some want to believe.

For me personally its a sucess, the noise (not really an issue with good case cooling), heat and power consumption don't bother me, I can't fault my 470s on performance or image quality and I got mine at a great price.
 
My 480 is the best card I have ever owned, considering I'm getting 580 performance out of it at 820 core I'm getting even better value for money.
 
The Nvidia cards are a mixed bag. The top end cards are very fast and very competitive with the ATi offerings. However, their low end cards such as the GTX430 have been very underwhelming. There hasn't been an unseating of ATi's dominance in this area, and ultimately this is the area most people will be looking towards as most can't afford enthusiast cards. The noisy and vocal enthusiast community are fairly won over by Nvidia I'd say as regards competitiveness and speed, but Nvidia need to really work on their low end range because that's where the bulk sales will be and their current offerings are somewhat damp.
 
You must log in or register to reply here.
Back
Top Bottom