Best AMD Cpu vs Intel i7 980

[WaitForMeImOnDialUp]

you guys keep bringing up single core speed..you realize thats one aspect of a cpu right?

 

what about integer math, floating point , prime numbers ,extended instructions , compression, encryption , physics and sorting?

 

show me a core 2 that wont bottleneck 64 player multiplayer bf4,crysis 3, or advanced warfare. even q9650,fx 4300 bottlenecks that.

 

oh but they score good on cinebench the intel sponsered benchmark

[i_build_nanosuits]

you guys keep bringing up single core speed..you realize thats one aspect of a cpu right?

 

what about integer math, floating point , prime numbers ,extended instructions , compression, encryption , physics and sorting?

 

show me a core 2 that wont bottleneck 64 player multiplayer bf4,crysis 3, or advanced warfare. even q9650,fx 4300 bottlenecks that.

 

oh but they score good on cinebench the intel sponsered benchmark

The FX is great at integer instructions because it has 8 integer units (2 per modules), prime numbers is what you have down here (super PI), extended instructions let's not get into that cause AMD FX supports only old sets of instrutctions so it's a fail in that category and the other two (physics and sorting) again depends on the per core performance...by the way i see you've ran passmark cpu benchmark and now you think you have a great chip because it scores high in this test? well it's because this test give the same amount of credit to integer maths then it does to ''single-thread performance'' and in real life it's not how it works, single-thread performance is key and it should be worth 80% of the final score...the FX is doing a phenomenal job at integer maths (even better than my i7) and it's because as i sais it has 8 integer units...but it share only 4 floating points units (FPU) and this really kills the performance and it's the reason why this CPU should be considered a ''quad-core'' because it only has 4 mondules containing only one FPU per modules.

 

you are set to ''ignore''...here's a quick reality dose for you if cinebench is not cutting it pal:

 

 

This is worth consulting...see where the BEST AMD CPU' ranks up:

http://www.cpubenchmark.net/singleThread.html

[WaitForMeImOnDialUp]

 

This is worth consulting...see where the BEST AMD CPU' ranks up:

http://www.cpubenchmark.net/singleThread.html

 

juding by that passmark single core list the 8350 beats

 

the fx4300,the q9650,the e8400,the e2180(lol at the cpu) ,beats the 980 and even the 980x extreme.

 

good post thanks

[i_build_nanosuits]

Look, i don't want to get into a war here, all i'm saying is that the OP's core i7-980 6 core 12 threads nehalem CPU is superior to even an FX-9590 by a very noticeable margin, anybody saying otherwise is obviously clueless and in denial.

[FuzzyYellow]

the 980 for sure.

[WaitForMeImOnDialUp]

 

you are set to ''ignore''...here's a quick reality dose for you if cinebench is not cutting it pal:

 

 

cool.

 

lol at the "pal"

[i_build_nanosuits]

juding by that passmark single core list the 8350 beats

 

the fx4300,the q9650,the e8400,the e2180(lol at the cpu) ,beats the 980 and even the 980x extreme.

 

good post thanks

AGAIN' this is because they based the single-thread score doing irrelevant tests such as the ones you've described before and those intel chips are at stock speed...when you give them a slight overclock like harrynowl did the stronger IPC on these breach the gap and they come out on top of an overclocked FX such as demonstrated in the graphs that' i've shown you earlier.

[WaitForMeImOnDialUp]

okay no more stupid wars, but personal question nano...when you had the 8350, were you happy with it for the time being? we know you are much happier now but when you did have one wouldnt you have said its better than people give it credit for?

[i_build_nanosuits]

cool.

 

lol at the "pal"

how about this, LTT community users submission spread sheet single-threaded performance at rendering:

https://docs.google.com/spreadsheet/ccc?key=0AlC81MjwelBgdEZNV3l6aHl1eUNwSUR4Rml0MXMzN1E&usp=sharing#gid=1

The highlithed results are the ones you should be looking at, notice the clockspeed this is the FASTEST single-threaded core for an FX CPU ever recorded in this forum, hurrah to @Priller for that but as you can see ryan leech's core 2 tops it at much lower clocks...

[i_build_nanosuits]

okay no more stupid wars, but personal question nano...when you had the 8350, were you happy with it for the time being? we know you are much happier now but when you did have one wouldnt you have said its better than people give it credit for?

if i would have had a mid-range GPU in regards to gaming yes it does deliver...when you ramp up your GPU capacity you quickly run into cpu limitations which for me was a let down...

they are still better IMHO then what most people depict them to be...yes.

But...on topic here i have no choice but to say that the i7-980 is a superior cpu to anything AMD has to offer at the moment specialy considering the OP has the idea of making a switch to it which is nothing but a downgrade honestly.

 

Would i pick an FX over a core 2 Gen CPU...off course i would...i personaly upgraded from an overclocked Q6600@3.6ghz to an FX 8320 and the performance jump was very significant in games...the Q6600 was chocking my HD7950 pretty bad in most games and when i poped in the FX i saw a very nice improvement...and this HD7950 (R9 280) is the sweet spot to pair with an FX CPU for 1080p gaming IMHO.

[sgloux3470]

how about this, LTT community users submission spread sheet single-threaded performance at rendering:

https://docs.google.com/spreadsheet/ccc?key=0AlC81MjwelBgdEZNV3l6aHl1eUNwSUR4Rml0MXMzN1E&usp=sharing#gid=1

The highlithed results are the ones you should be looking at, notice the clockspeed this is the FASTEST single-threaded core for an FX CPU ever recorded in this forum, hurrah to @Priller for that but as you can see ryan leech's core 2 tops it at much lower clocks...

And what about multithread?

[i_build_nanosuits]

And what about multithread?

lets not get into that

multi-threaded performance of the FX vishera line-up is VERY good for the price, to someone looking for a budget workstation the FX8320 is an awesome value...it can compete with an i7-3770 or even with a core i5-4670K

https://docs.google.com/spreadsheet/ccc?key=0AlC81MjwelBgdEZNV3l6aHl1eUNwSUR4Rml0MXMzN1E&usp=sharing#gid=0

Problem is, multi-threaded performance is irrelevant to 90%+ of the users in here (AKA as gamers...)

 

Rank #46 is an intel i7-980X CPU...you can see that this cpu renders faster than even an i7-4790K.

EDIT: prokon ranked #28 with that chip as well...as you can see with a significant overclock this i7-980 can compete with even the 6 cores i7-3930K and 3970K...that's how good this cpu really is.

If op want a significant improvement to his rendering time, i'm sorry to say but he has no choice but to look at X99 and core i5-5960X CPU's...or 8 core xeons.

[sgloux3470]

lets not get into that

multi-threaded performance of the FX vishera line-up is VERY good for the price, to someone looking for a budget workstation the FX8320 is an awesome value...it can compete with an i7-3770 or even with a core i5-4670K

https://docs.google.com/spreadsheet/ccc?key=0AlC81MjwelBgdEZNV3l6aHl1eUNwSUR4Rml0MXMzN1E&usp=sharing#gid=0

Problem is, multi-threaded performance is irrelevant to 90%+ of the users in here (AKA as gamers...)

 

Rank #46 is an intel i7-980X CPU...you can see that this cpu renders faster than even an i7-4790K.

EDIT: prokon ranked #28 with that chip as well...as you can see with a significant overclock this i7-980 can compete with even the 6 cores i7-3930K and 3970K...that's how good this cpu really is.

If op want a significant improvement to his rendering time, i'm sorry to say but he has no choice but to look at X99 and core i5-5960X CPU's...or 8 core xeons.

Well it's also a 980X at 5.3 Ghz which is a pretty insane over clock

Multithreading is certainly applicable to video rendering etc that OP seems to want to use it for.

That said I do agree I doubt it'd be significantly better. It's still a $1000 CPU vs a $200 one after all. For major improvements he'd need to spend a crap ton of money.

[i_build_nanosuits]

Well it's also a 980X at 5.3 Ghz which is a pretty insane over clock

Multithreading is certainly applicable to video rendering etc that OP seems to want to use it for.

That said I do agree I doubt it'd be significantly better. It's still a $1000 CPU vs a $200 one after all. For major improvements he'd need to spend a crap ton of money.

here's an agregate comparison between the i7-980 and FX-8350 both at stock, results provided by tom's hardware...this is as good as it gets to compare both chips:

Tom's Hardware - Benchmark [19] SiSoftware Sandra 2012 SP4c  Pro

 

As you can see both at stock they compare fairly well to one another, the i7 being slightly faster in most tests...but they are certainly not a worthy upgrade to one another in either ways.

 

Toms hardware have ran 35 tests within different softwares to rank the CPU in regards to their ''workstation'' and ''gaming'' performance and the total time results for each CPU's is shown in this page here:

http://www.tomshardware.com/charts/cpu-charts-2013/-36-Total-Time,3179.html

1374 seconds was the total time for the i7-980, and 1577 seconds was the time total for the vishera FX-8350...the i7 is therefore 14.77% faster overall at stock.

[Priller]

how about this, LTT community users submission spread sheet single-threaded performance at rendering:

https://docs.google.com/spreadsheet/ccc?key=0AlC81MjwelBgdEZNV3l6aHl1eUNwSUR4Rml0MXMzN1E&usp=sharing#gid=1

The highlithed results are the ones you should be looking at, notice the clockspeed this is the FASTEST single-threaded core for an FX CPU ever recorded in this forum, hurrah to @Priller for that but as you can see ryan leech's core 2 tops it at much lower clocks...

 

I would just like to point out that was with all 8 cores running. You can increase your single threaded performance but shutting off 1 core in each module. It pulled up single thread performance by about 9%. Not to mention you can go higher in the overclock with 4 cores running instead of 8.

 

http://linustechtips.com/main/topic/62476-post-your-cinebench-r15r1152003-scores-over-250-submissions/page-176#entry3811733

And what about multithread?

[i_build_nanosuits]

I would just like to point out that was with all 8 cores running. You can increase your single threaded performance but shutting off 1 core in each module. It pulled up single thread performance by about 9%. Not to mention you can go higher in the overclock with 4 cores running instead of 8.

 

http://linustechtips.com/main/topic/62476-post-your-cinebench-r15r1152003-scores-over-250-submissions/page-176#entry3811733

 

http://en.wikipedia.org/wiki/Bulldozer_%28microarchitecture%29

The Tom's Hardware website commented that the lower-than-expected performance in multi-threaded workloads may be because of the way Windows 7 currently schedules threads to the cores. They point out that "if Windows were able to utilize an FX-8150's four modules first, and then backfill each module's second core, it'd maximize performance with up to four threads running concurrently." This is similar to what happens on Intel CPUs with HyperThreading – Windows 7 "schedules to physical cores before utilizing logical (HyperThreaded) cores."^[33]

 

In January 2012, Microsoft released two hotfixes for Windows 7 and Server 2008 R2 that significantly improve the performance of Bulldozer CPUs by addressing the thread scheduling concerns raised after the release of Bulldozer.^[35][36]

On 6 March 2012, AMD posted a knowledge base article stating that there was a compatibility problem with FX processors, and certain games on the widely used digital game distribution platform, Steam. AMD stated that they had provided a BIOS update to several motherboard manufacturers (namely: Asus, Gigabyte Technology, MSI, and ASRock) that would fix the problem.^[37]

 

^{TLDR: there shoudnt be any performance difference nowadays if you have an up to date version of windows 7 or any versions of windows 8...}

[Priller]

http://en.wikipedia.org/wiki/Bulldozer_%28microarchitecture%29

The Tom's Hardware website commented that the lower-than-expected performance in multi-threaded workloads may be because of the way Windows 7 currently schedules threads to the cores. They point out that "if Windows were able to utilize an FX-8150's four modules first, and then backfill each module's second core, it'd maximize performance with up to four threads running concurrently." This is similar to what happens on Intel CPUs with HyperThreading – Windows 7 "schedules to physical cores before utilizing logical (HyperThreaded) cores."^[33]

 

In January 2012, Microsoft released two hotfixes for Windows 7 and Server 2008 R2 that significantly improve the performance of Bulldozer CPUs by addressing the thread scheduling concerns raised after the release of Bulldozer.^[35][36]

On 6 March 2012, AMD posted a knowledge base article stating that there was a compatibility problem with FX processors, and certain games on the widely used digital game distribution platform, Steam. AMD stated that they had provided a BIOS update to several motherboard manufacturers (namely: Asus, Gigabyte Technology, MSI, and ASRock) that would fix the problem.^[37]

 

^{TLDR: there shoudnt be any performance difference nowadays if you have an up to date version of windows 7 or any versions of windows 8...}

 

That was with an 8150. Those hotfixes were integrated into their CPU scheduling. 

 

Windows cant control the resource sharing between cores in a module. The performance difference is only seen when 1 core in the modular is turned off allowing the core left to use all of the resources availible to it.

 

I also showed you the difference in the link. One with 8 active cores running 4 threads vs 1 core per module active. 

 

Yes my stuff is always up to date.

[i_build_nanosuits]

That was with an 8150. Those hotfixes were integrated into their CPU scheduling. 

 

Windows cant control the resource sharing between cores in a module. The performance difference is only seen when 1 core in the modular is turned off allowing the core left to use all of the resources availible to it.

 

I also showed you the difference in the link. One with 8 active cores running 4 threads vs 1 core per module active. 

 

Yes my stuff is always up to date.

fair enough, you know this stuff better than i do it's just that's ive heard about this fix long ago and at first when i had the FX-8320 i was on a pirated copy of windows 7 and then i upgraded that to a legit windows 8 and the performance was a lot better after that so i thought this ''hotfix'' thingy has that covered but i guess i was wrong and it's alright

so now, do you think/plan on raising this single-threaded score higher in the spreadsheet cause i refer to your score quite often these days man

[Dabombinable]

okay no more stupid wars, but personal question nano...when you had the 8350, were you happy with it for the time being? we know you are much happier now but when you did have one wouldnt you have said its better than people give it credit for?

Look, I even did comparisons myself when I was trying to see how much my E8500 at the same clock speed as an FX 8350 would bottleneck a GTX 970. And after I had compared the graphics score, I noticed that for some reason my Core 2 had a higher IPC than the FX 8350. And actually looking at my charts, I was able to get my Core 2's IPC quite close to high end i7 CPUs.

 AMD made the wrong choice choosing an architecture that is similar to Intel's defunct Netburst. It overclocked like hell, but it was slow and saw minimal benefits when overclocked.

[Priller]

fair enough, you know this stuff better than i do it's just that's ive heard about this fix long ago and at first when i had the FX-8320 i was on a pirated copy of windows 7 and then i upgraded that to a legit windows 8 and the performance was a lot better after that so i thought this ''hotfix'' thingy has that covered but i guess i was wrong and it's alright

so now, do you think/plan on raising this single-threaded score higher in the spreadsheet cause i refer to your score quite often these days man

 

I'm aware of that. lol

 

I wasnt really aiming for single threaded performance. I've looked into some other scores and there is some room to lift that score up. 

 

I'm trying to go for the Phenom II X2 B55 world record first then I'll move onto my FXs as I should have my parts by then.

[Dabombinable]

I'm aware of that. lol

 

I wasnt really aiming for single threaded performance. I've looked into some other scores and there is some room to lift that score up. 

 

I'm trying to go for the Phenom II X2 B55 world record first then I'll move onto my FXs as I should have my parts by then.

Good luck with that, I was going to do the same with my Xeon before replaceing my i5 4440, but I killed the motherboard Now I need to spend at least another $100 on my OC rig just to get it working.

[i_build_nanosuits]

Look, I even did comparisons myself when I was trying to see how much my E8500 at the same clock speed as an FX 8350 would bottleneck a GTX 970. And after I had compared the graphics score, I noticed that for some reason my Core 2 had a higher IPC than the FX 8350. And actually looking at my charts, I was able to get my Core 2's IPC quite close to high end i7 CPUs.

 AMD made the wrong choice choosing an architecture that is similar to Intel's defunct Netburst. It overclocked like hell, but it was slow and saw minimal benefits when overclocked.

Lol, IPC is a fixed and undeterminated number it can't be verified or mesured accurately outside of benchmarks results...IPC means Instructions Per Clock and it won't increase or change it's a fixed number for a certain architecture..haswell i5's for example all have the same IPC...what you are refering to is the PER CORE PERFORMANCE or SINGLE-THREADED performance.

[Dabombinable]

Lol, IPC is a fixed and undeterminated number it can't be verified or mesured accurately outside of benchmarks results...IPC means Instructions Per Clock and it won't increase or change it's a fixed number for a certain architecture..haswell i5's for example all have the same IPC...what you are refering to is the PER CORE PERFORMANCE or SINGLE-THREADED performance.

Either way, with more cores an E8500 would flog an FX 8350.

[i_build_nanosuits]

Either way, with more cores an E8500 would flog an FX 8350.

not really because theres a lot more than that to it...amount of cache, cache speeds, newer supported instructions sets on the FX platform etc.

No intel cpu's from that era can compete in real world applications with the FX vishera cpu's.

[Dabombinable]

not really because theres a lot more than that to it...amount of cache, cache speeds, newer supported instructions sets on the FX platform etc.

No intel cpu's from that era can compete in real world applications with the FX vishera cpu's.

Granted the Core 2 line does lack some instructions and has a slower cache (my only real issues is the lack of instructions), but as for the amount of L2, with Wolfdale and Yorkfield they had 3MB per core, while the FX line has 2MB.