News - As PC market shrinks AMD likens Intel/Nvidia to beached whales

[CAT-THE-FIFTH]

Every single CPU that AMD has released is beaten out by the "Intel equivalent" which basically means Intel does just as much or more with half the cores and slower clocks.

The i3 series do as much as the quad cores, and the i5 series (and i7) beat out the AMD 8 cores.

I have a Core i5 myself but the argument that every single Intel CPU beats an AMD price comparable equivalent is just hilarious.

I had a SB Core i3 and my mates FX6300 is faster in plenty of modern games,and will be in the next few years. This is why I ditched it.

The Llano and Trinity CPUs are faster in many non-gaming applications when compared to price comparable Intel CPUs,and the IGP is faster too. I know because I have used both.

I have never understood why people after buying into certain brands for their own personal use and then need shout from the top how they are the best at every level. Its like they need to keep justify every product that brand sells,otherwise it will cheapen their own choice.

I'm a big fan of Intel CPUs.

AMD fans like to throw us the "bang for the buck" argument.

Fans?? Us?? Is this an e-peen war with sides??

Sounds rather dorkish.

Most are worried about value for money. In fact everyone I have ever done a build for in real life has looked at value for money and 95% of the hundreds of builds me and other people on Hexus have helped out with have been about,getting the best value for money components within the budget,irrespective of what parts are used,be it Intel,Nvidia or AMD.

Heard or HUKD?? A very popular UK site centred around bank for buck and we even have a sub-forum dedicated for bang for buck deals.

In the US the equivalent is Slickdeals.

I give you a simple test. One company sells the Core i5 4670K for £400 and every other retailer sells it for around £200. Would you buy the £400 Core i5 4670K or from one of the retailers which sell it for £200?? Assume they are all legit decent retailers.

What about when buying a car?? A house?? Will you pay beyond the odds for it or try and negotiate the best price or best deal??

Answers on a post card to Captain Obvious.....!

[teppic]

This thread's so full of inaccuracies about AMD performance I wouldn't even know where to start in a proper reply.

7970 only 7-10% higher than a 760? Try 20-50% depending on the game. At 1440p it can rise to 60-70% faster.

[Yash047]

One day, I dreamed that Intel bought Nvidia like how AMD bought Ati. Intel started using GeForce as its IGP. GeForce expansion graphics card started using Intel technologies like the 3D transistors. I wonder what would happen to AMD in the semiconductor industry.

[kompukare]

This thread's so full of inaccuracies about AMD performance I wouldn't even know where to start in a proper reply.

7970 only 7-10% higher than a 760? Try 20-50% depending on the game. At 1440p it can rise to 60-70% faster.

Hm, on HUKD there's someone claiming a 760 is faster than a 7950 Boost and that 'A 7970 comes close to a 760 for a cool £100 more'

http://www.hotukdeals.com/deals/xfx-...0#post18769794

I had a look at that: certainly TPU had them close with the 760 faster than the 7950 (non-boost) but other reviews like computerbase

http://www.computerbase.de/artikel/g...760-im-test/4/

had the two even at 1080P and the 7950 around 5% faster at 1600P. HT4U had the biggest difference with the 7950 leading by 11% at 1600P.

http://ht4u.net/reviews/2013/55_dire...st/index30.php

The choice of tests has a huge bearing. Another huge factor which the English language reviews tend to ignore is the thermal throttling which a lot of Kepler cards tend to do once they warm up: Kepler cards are great running 5-10min of benchmark but not so good at running games - this is especially the case for Titan.

The hardware.fr also noticed a difference with the 760 when running with extra case fan:

http://www.hardware.fr/articles/900-...formances.html

And they also had the 7950 4-5% quicker at 1080P.

[teppic]

My post above was about the 7970, not the 7950?

One example, taken with recent drivers at stock clocks - Bioshock Infinite, max settings no AA, at 1440p:

760: 24.6fps
7970: 41.7ps

69.5% better performance.

The increase is more typically 20-50% though.

[chockimon]

The AMD 7970 and Nvidia 760 aren't even in the same price bracket, so why are you even comparing the two. It costs nearly £100 more.

[DanceswithUnix]

The AMD 7970 and Nvidia 760 aren't even in the same price bracket, so why are you even comparing the two. It costs nearly £100 more.

Because this thread has descended into petty fanboy or possibly trolling based arguments. Welcome to the Internet

[teppic]

I was responding to this comment,

"7970 is such a bad example. It costs twice as much as GTX 760 (at least in my country) and is only about 7-10% faster. Ofc., it's a year difference between them, but even so it still considering the price difference the winner is clear."

Which is a load of rubbish/trolling/whatever.

[DanceswithUnix]

I was responding to this comment,

"7970 is such a bad example. It costs twice as much as GTX 760 (at least in my country) and is only about 7-10% faster. Ofc., it's a year difference between them, but even so it still considering the price difference the winner is clear."

Which is a load of rubbish/trolling/whatever.

Yeah there is some pretty wild stuff in here
A thread title like that was always going to attract some indignation though.

The thing that I find odd with the shrinking PC marked is that Samsung are apparently moving Flash production away from mobile phones and into SSDs.

[crossy]

Heard or HUKD?? A very popular UK site centred around bank for buck and we even have a sub-forum dedicated for bang for buck deals.

Didn't know that about HUKD, thought they tended to be more for consumer products - having seen a couple of HUKD recommends on MSE forums.

The choice of tests has a huge bearing. Another huge factor which the English language reviews tend to ignore is the thermal throttling which a lot of Kepler cards tend to do once they warm up: Kepler cards are great running 5-10min of benchmark but not so good at running games - this is especially the case for Titan.

So presumably the higher cards, 770, 780 also suffer from this issue, so perhaps we shouldn't be taking one-shot benchmark "headline" figures as gospel for these cards? PS, thanks for the graph and links - interesting.

My post above was about the 7970, not the 7950?
One example, taken with recent drivers at stock clocks - Bioshock Infinite, max settings no AA, at 1440p:
760: 24.6fps 7970: 41.7ps
69.5% better performance.
The increase is more typically 20-50% though.

Thanks for this - I'm a BioInf player using a 7970, but in my case paired with a 1090T and running at 1080p. I've been suspecting that some of my games don't "like" my comparatively elderly cpu, so it'll be interesting to see what figures my rig can bench at, and compare them to yours.

I was responding to this comment,
"7970 is such a bad example. It costs twice as much as GTX 760 (at least in my country) and is only about 7-10% faster. Ofc., it's a year difference between them, but even so it still considering the price difference the winner is clear."
Which is a load of rubbish/trolling/whatever.

To leap to the posters defence, it might very well be the case in his country. Heck, I've seen some vendors in the UK who seem to have some great deals. So perhaps he's been unlucky enough to find a local shop who managed to get a good deal from NVidia, or alternately is being ripped off by their AMD/ATi supplier?

Because this thread has descended into petty fanboy or possibly trolling based arguments. Welcome to the Internet

Hmm, true. Although Cat (as usual), teppic and kompukare seem to be doing their level best to make this lead balloon thread ascend.

And if you want to see a real fanboy cat fight then I'd suggest trying The Register comments section - real "burn the heretic" stuff!