• 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.

AMD cpu that matches I5 2500k?

shankly1985

shankly1985

shankly1985

Soldato
Joined
25 Nov 2011
Posts
20,679
Location
The KOP
Hello looking to build my brother a Gaming pc. I have old parts from my older pc.
7950, asrock z68 extreme 4 gen3 motherboard, corsair 4gb ram

Now I sold my old I5 2500k and upgrade to i7 2600k, am thinking buying a amd cpu plus will need a motherboard also. Do you think this would be worth it or should I just put the money towards a i5?

Amd cpu plus mb costing? Vs just Intel cpu?
With the pc running just single 7950 and a 1080p display 60hz he not expecting to run games ultra just so he can least play them. So any cpu that won't bottleneck a 7950.

Thanks.
 
joeyjojo said:
6300 or 8320. At normal resolutions (1080p+) with a single GPU there's not much difference between them and the 2500K.
Click to expand...

Depends on the games and GPU, while there's examples of parity FX8320/i5 2500K in X game, there's Y game that's far from parity, and there's X, where the 2500K's a little ahead, I guess we could argue there's Z, a few games that the FX8350 just pokes ahead, stock versus stock (Although somewhat ironically, one of these titles is a Mantle title, so the win becomes a draw)

There is NO AMD CPU that matches an i5 2500K all around, you can either get the FX6300, which is going to be slower most of the time, and parity some of the time, think there's some odd occasions where the FX6300 is faster too.

Or an FX8320 that's slower/parity/faster depending on the work load (Basically, where the FX6300 can gain parity/win, the FX83, if the software's threaded enough, should claim a win and add more onto the FX6300 performance)

Plus, when we're talking overclocking, the 2500K has more in the tank to gain on average, not just in total MHZ gain, but percentage gain too.

Factor in that your brother has a potential upgrade path to the you i7 (Which makes his set up have higher performance potential than the AM3+ will ever get)

That review's running only a GTX 580, that's a far cry (;)) from what's likely to be an overclocked 7950.
 
Last edited:
Probably best off grabbing a pre-owned 2500k and just dropping it in. Will be cheaper and offer equal or better performance (in most scenarios) than switching platforms. Maybe bang in an extra 4gb RAM if budget allows too.
 
Average gaming performance of fx-6350 (an overclocked fx-6300) vs i5-3570k over 8 or 9 games tested by Tom's Hardware using a 7970;
http://www.tomshardware.co.uk/piledriver-k10-cpu-overclocking,review-32759-19.html the i5 is around 15-20% ahead overall

Individual game benchmarks for the AMD CPUs is earlier in that article, the individual scores for Intel are here;
http://www.tomshardware.com/reviews/ivy-bridge-wolfdale-yorkfield-comparison,3487-2.html

Most games the i5 was a little ahead. At playable setting (not highest) the FX kicked the i5's arse on Crysis 3, though.

A 2500k would be around 10% behind the 3570k, so probably evenly matched with the 8320.
 
Last edited:
cheesyboy said:
Average gaming performance of fx-6350 (an overclocked fx-6300) vs i5-3570k over 8 or 9 games tested by Tom's Hardware using a 7970;
http://www.tomshardware.co.uk/piledriver-k10-cpu-overclocking,review-32759-19.html
Click to expand...

It's an overclocked FX6350 versus a stock i5 3570K..........
The 2500K doesn't *have* to lose 10% performance either (Because the Ivy could be at the GPU limit, there's no reason the Sandy won't be too)

Couple in that Ivy's OC'ing tends to be worse than the 2500K's, a 2500K and 3570K end up within literally a couple of percent of each other (Unless we're talking using newer instruction sets, which is what allows the FX83 to 1 up the 2500K, stock versus stock in Crysis)
 
Last edited:
Unless money was a concern, not many people would pick the FX83 over an i5 2500K (Bearing in mind, you'd be looking second hand, and can literally pick up a 2500K for 90 quid) for gaming performance, let alone an FX6300.
EDIT : Those Crysis 3 results confuse me, the FX63 shouldn't lose when the details are turned up, if the prior results are accurate (I mean, we're talking a 20 FPS lead over the i5, something almost unheard of, you don't just suddenly lose that edge when the details are turned up.)

EDIT 2 : http://www.techspot.com/review/642-crysis-3-performance/page6.html, look at the Medium setting, FX8350 versus i5 3470, the FX8350's winning by a frame.


EDIT 3 : This is the most relevant benchmark for Crysis 3 ; http://i.imgur.com/0nIkCAb.jpg

And I believe that's down to the instruction set difference, it's also the result I would rely on.
 
Last edited:
Martini1991 said:
Unless money was a concern, not many people would pick the FX83 over an i5 2500K (Bearing in mind, you'd be looking second hand, and can literally pick up a 2500K for 90 quid) for gaming performance, let alone an FX6300.
EDIT : Those Crysis 3 results confuse me, the FX63 shouldn't lose when the details are turned up, if the prior results are accurate (I mean, we're talking a 20 FPS lead over the i5, something almost unheard of, you don't just suddenly lose that edge when the details are turned up.)
Click to expand...

Probably something hitting the lower single core limit of the FX. Or AMD's poorer memory performance.

Martini1991 said:
EDIT 2 : http://www.techspot.com/review/642-crysis-3-performance/page6.html, look at the Medium setting, FX8350 versus i5 3470, the FX8350's winning by a frame.


EDIT 3 : This is the most relevant benchmark for Crysis 3 ; http://i.imgur.com/0nIkCAb.jpg

And I believe that's down to the instruction set difference, it's also the result I would rely on.
Click to expand...
Tom's uses what they say is a "brutal Crysis 3 sequence" and tested at 1080p, whereas the techspot one is a pretty tame follow the leader section at higher res.
 
Last edited:
cheesyboy said:
Probably something hitting the lower single core limit of the FX. Or AMD's poorer memory performance.
Click to expand...

I don't particularly buy it, (And Tom's gets slaughtered everytime it's ever used to show an Intel result positively)

I'd put bets on ; http://i.imgur.com/0nIkCAb.jpg being the most accurate (Which, as I've mentioned, the 2500K's instruction set difference versus the FX's, Ivy has that instruction set, so would probably gain parity, as we see from the techspot results, but OP's looking at Sandy)
 
Martini1991 said:
Unless money was a concern, not many people would pick the FX83 over an i5 2500K (Bearing in mind, you'd be looking second hand, and can literally pick up a 2500K for 90 quid) for gaming performance, let alone an FX6300.
EDIT : Those Crysis 3 results confuse me, the FX63 shouldn't lose when the details are turned up, if the prior results are accurate (I mean, we're talking a 20 FPS lead over the i5, something almost unheard of, you don't just suddenly lose that edge when the details are turned up.)

EDIT 2 : http://www.techspot.com/review/642-crysis-3-performance/page6.html, look at the Medium setting, FX8350 versus i5 3470, the FX8350's winning by a frame.


EDIT 3 : This is the most relevant benchmark for Crysis 3 ; http://i.imgur.com/0nIkCAb.jpg

And I believe that's down to the instruction set difference, it's also the result I would rely on.
Click to expand...

It depends where you test in Crysis3. The most CPU intensive part is "Welcome to the Jungle" AFAIK,where a more multi-threaded load is pushed due to the grass animations. If you test a corridor scene,you could see better relative framerates for the Intel CPUs though.

Also,the FX6200 tested in the Techspot review tends to be slightly slower than a FX6300 AFAIK.

Edit!!

The same goes with BF4 MP. Some sites test 64 player large conquest maps and others 32 player domination maps. The smaller maps tend to make Intel look a bit better.
 
Last edited:
It makes no sense to change their benchmarking location, so it can't be the location that's made them go from a massive gain to a lose.
I'd pick the results that actually show the sandy, and as said, it gets beaten by the fx83 in crysis 3.

I wasn't looking at the 6200 results, rather the fx83 ones.

And bf 4 results become somewhat irrelevant due to mantle,
 
Last edited:
Martini1991 said:
It makes no sense to change their benchmarking location, so it can't be the location that's made them go from a massive gain to a lose.
I'd pick the results that actually show the sandy, and as said, it gets beaten by the fx83 in crysis 3.

I wasn't looking at the 6200 results, rather the fx83 ones.
Click to expand...

Theoretically: At lower settings, there will lots of smaller processing threads to deal with in a particularly demanding section, which the FX-63's higher thread count will deal with better than the i5.
Once you turn the detail up, one or more of those processing threads (higher textures or whatever) is demanding enough that the FX simply can't chew through it as quickly, and it hold up performance.
 
Martini1991 said:
Sure, if you ignore the techspot results on medium with the fx83 only just by a frame edging the i5.
They contradict the results.
Click to expand...
Yes, but....
cheesyboy said:
Tom's uses what they say is a "brutal Crysis 3 sequence" and tested at 1080p, whereas the techspot one is a pretty tame follow the leader section at higher res.
Click to expand...
So the FX's extra threads aren't put to any use: the per-core grunt of the i5 wins.
 
None of this is relevant to the OP :D

He should just go and buy a compatible i5 or i7 and drop it into the mobo he already has. No-one should buy an AM3+ setup to replace a functioning 1155 motherboard.
 
You must log in or register to reply here.
Back
Top Bottom