Best AMD Cpu vs Intel i7 980

[79wjd]

How... there are both Intel CPU's that have 6 cores 12 threads, I would only upgrade if you want things like M.2 and Sata Express

Because the number of cores doesn't mean anything when comparing different archithectures. 

 

A 5820k is four generations ahead, and is about 40% faster per core at the same clock speeds.

[i_build_nanosuits]

when overclocked yes, at stock..not so much

 

passmark score

 

 

single core speed

 

 

 

once again..when overclocked the 980 will surpass it easily.

i only posted these because not everyone overclocks.

i tought we made clear that passmark results where mostly irrelevant for a multitude of reasons yesterday didnt we?

those benchmarks are flawed in many ways and does not reflect actual in use performance like something like cinebench rendering single thread and multi thread does for example.

[WaitForMeImOnDialUp]

i tought we made clear that passmark results where mostly irrelevant for a multitude of reasons yesterday didnt we?

those benchmarks are flawed in many ways and does not reflect actual in use performance like something like cinebench rendering single thread and multi thread does for example.

thats true, its only passmark scores and in other benchmarks its probably much different.

[Fate]

Thats a realy funny statement.

 

infact, i have a FX8350 and a core 2 quad rig.

And do you realy think that the core 2 quad is better then the FX8350?

 

No fucking way.

 

The FX8350 is better in everything i have trown to it.

its not even close, i´m sorry

"single threaded performance"

Granted FX 8350 has 4x the cores as a c2d, the statement stands. 

[Faa]

Thats a realy funny statement.

 

infact, i have a FX8350 and a core 2 quad rig.

And do you realy think that the core 2 quad is better then the FX8350?

 

No fucking way.

 

The FX8350 is better in everything i have trown to it.

its not even close, i´m sorry

God you're so biased as fuck, don't tell me you haven't seen the benchmarks I included in this post; http://linustechtips.com/main/topic/287831-best-amd-cpu-vs-intel-i7-980/page-4#entry3917977

Now that's proven that Conroe (2005) has more IPC than the 8350 and two of them at 4.2GHz will outperform Prillers suicide 5.4GHz 8350.

 

 

i7-980 better then a FX9590 in streaming?

 

meh idk lol.

 

i would say post some proof for those claims.

A 980 is only 17% slower clock-for-clock than a 3930K, you could have googled this or checked the 2.8GHz benchmark I linked over 15 times so goodjob on claiming the 8350 is only 20-25% slower than the 3930K. There's no point we provide you "some proof" when you go afterwards like "Its bullshit, biased, blablabla".

Here you go;

Tech Yes City shows that 50% boost the i5 has over the 8350 as well -> fact.

 

 

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.

No you can't. With a single thread only one ALU cluster is being driven when the 2nd is idling. When you pump a 2nd thread in a module the singlethreaded performance degrades but windows does a good job moving a 2nd thread to a different module so the 1st thread's performance wouldn't be affected. You'll never gain more performance from disabling CMT, 5 threads and up will perform better on 4m/8c than 4m/4c. Those 2m/4c vs 4m/4c is just showing the difference between a 4300 and a 8350 with CMT disabled -> doesn't equal the 8350 getting more singlethreaded performance when you disable CMT.

[Dabombinable]

 

God you're so biased as fuck, don't tell me you haven't seen the benchmarks I included in this post; http://linustechtips.com/main/topic/287831-best-amd-cpu-vs-intel-i7-980/page-4#entry3917977

Now that's proven that Conroe (2005) has more IPC than the 8350 and two of them at 4.2GHz will outperform Prillers suicide 5.4GHz 8350.

 

 

A 980 is only 17% slower clock-for-clock than a 3930K, you could have googled this or checked the 2.8GHz benchmark I linked over 15 times so goodjob on claiming the 8350 is only 20-25% slower than the 3930K. There's no point we provide you "some proof" when you go afterwards like "Its bullshit, biased, blablabla".

Here you go;

Tech Yes City shows that 50% boost the i5 has over the 8350 as well -> fact.

 

 

No you can't. With a single thread only one ALU cluster is being driven when the 2nd is idling. When you pump a 2nd thread in a module the singlethreaded performance degrades but windows does a good job moving a 2nd thread to a different module so the 1st thread's performance wouldn't be affected. You'll never gain more performance from disabling CMT, 5 threads and up will perform better on 4m/8c than 4m/4c. Those 2m/4c vs 4m/4c is just showing the difference between a 4300 and a 8350 with CMT disabled -> doesn't equal the 8350 getting more singlethreaded performance when you disable CMT.

 

Don't forget as well that I got the same results in my test, and found that a 980 at stock speeds flogs an FX 8350 (stock speeds as well).

[Profquast]

Getting more and more excited to start my electrical engineering study and understand how the hell a CPU works, because it's obvious that IPC is about as irrelevant as clockspeed when you look at all the variables at play. Also wow this blew up.

[Faa]

Getting more and more excited to start my electrical engineering study and understand how the hell a CPU works, because it's obvious that IPC is about as irrelevant as clockspeed when you look at all the variables at play. Also wow this blew up.

It blew up because we have had liars like you here.

 

 

An 9xxx series CPU will beat it but not by much and it consumes ungodly amounts of power so not worth it. Sandy Bridge was like a 30% IPC increase so the gap between AMD and intel is smaller here, but it still exists.

This clearly was made up, you had no proof to provide or you would have provided it, 30% is the IPC difference between Nehalem & Vishera. I like to see a 9370 4.3GHz thing which is in terms of IPC a decade old touching a modern CPU that comes with 2 more cores & HT.

The benchmark you have seen over 50 times;

Any reasons you are making numbers by yourself up after you've seen this IPC at 2.8GHz test? That's lying.

[Sintezza]

 

God you're so biased as fuck, don't tell me you haven't seen the benchmarks I included in this post; http://linustechtips.com/main/topic/287831-best-amd-cpu-vs-intel-i7-980/page-4#entry3917977

Now that's proven that Conroe (2005) has more IPC than the 8350 and two of them at 4.2GHz will outperform Prillers suicide 5.4GHz 8350.

 

 

 

 

Core 2 is not superiour to a FX8350.

i have  Core 2 Quad, and the FX8350 is simply better

In fact, the C2Q is slow as fuck in compairisson to the FX8350, so is the Core 2 duo.

i owned them both.

 

And dont come with a statement that a core 2 duo is better then a core 2 quad single threaded, because thats the biggest laught i have ever seen, same cpu, same archtitecture, clockable at the same speed.

You can even disable 2 cores on the Q2Q and make it a C2D

 

Streaming i7-980 better then the FX9590, like i said idk.

I wanne see some real proof about that first, before i can say something usefull about that.

Just "suggestive" numbers dont tell me anything.

It might be true that the i7-980 is better in gaming and streaming, i never claimed that this was false.

THats you turning arround words again.

 

In terms of rendering the i7-980 is better in terms of raw performance.

But it still depends allot on which render application you are using.

 

You know that im not going into any discussions with fanboys.

I just say my opinnion, based on my own experiances, since i owned a FX8350 and a core 2 quad.

And people are free to like it or not.

 

Honnestly i dont care lol

i´m just ROFLMAO

[sgloux3470]

 

Any reasons you are making numbers by yourself up after you've seen this IPC at 2.8GHz test? That's lying.

 

 

Has anyone pointed out to you yet how stupid it is to compare IPC's when AMD CPU's often come clocked significantly faster?  It's two different ways to get to the same place: if AMD had 50 IPC and 4.0 GHz and intel had 100 IPC at 2.0 GHz they'd perform very similarly.  Obviously, Intel has higher IPC and can certainly reach those 3.8-5.0 GHz clockspeeds and it tends to seriously outperform AMD in that range.  IPC * Clock speed = approximate CPU performance.  You can't just look at IPC.

 

Lets see some real world tests, like video rendering.

[Dabombinable]

i dont care about those benchmarks, a core 2 due superiour to a FX8350.

is simply BS, like i said i have  Core 2 Quad, and its not beating the FX8350 in anything.

In fact, the C2Q is slow as fuck in compairisson to the FX8350, so is the Core 2 duo.

i owned them both.

 

like i said i have owned those systems, But yeah intel fanboys just believe every intel favour benchmark chart they can ever find.

do the tests your selfs, and then speak.

 

 

Also streaming i7-980 better then the FX9590, like i said idk.

I wanne see some real proof about that first, before i can say something usefull about that.

 

Just "suggestive" numbers dont tell me anything my friend

*Sigh* If a Core 2 Quad or Core 2 Duo is running at the same frequency as an FX 8350 for example (4GHz), both cpus will be faster in tasks that make use of 4 or 2 threads as each individual core is faster than each individual core of an FX 8350. In tasks that make use of more than 2 or 4 threads the FX 8350 will pull ahead as it has 8 'cores' to play with. Core count is the FX line's only saving grace. And that FX 9590 is a CPU running at a suicide factory overclock with liquid cooling and a high end motherboard required to even use the damn thing. Then there is the need for a beefy PSU to run it as well because of the sheer amount of power it draws. Do not call me a fan boy on this because I have done my own research and comparisons. Also remember that to match an intel CPU at stock AMD needs to have massive overclocks, which are useless as you can overclock with Intel cpus as well on AIR and easily negate the overclocks of AMD cpus, then there is the fact that the 980 has 6 cores, 12 threads instead of 4 cores, 8 modules.

 

 

Has anyone pointed out to you yet how stupid it is to compare IPC's when AMD CPU's often come clocked significantly faster?  It's two different ways to get to the same place: if AMD had 50 IPC and 4.0 GHz and intel had 100 IPC at 2.0 GHz they'd perform very similarly.  Obviously, Intel has higher IPC and can certainly reach those 3.8-5.0 GHz clockspeeds and it tends to seriously outperform AMD in that range.  IPC * Clock speed = approximate CPU performance.  You can't just look at IPC.

 

Lets see some real world tests, like video rendering.

For me I go by the amount of opperations per second a single core in any CPU can do (measured in MOPS) at the same frequency as other CPUs, I've never gone that in-depth with IPCs.

[Sintezza]

*Sigh* If a Core 2 Quad or Core 2 Duo is running at the same frequency as an FX 8350 for example (4GHz), both cpus will be faster in tasks that make use of 4 or 2 threads as each individual core is faster than each individual core of an FX 8350. In tasks that make use of more than 2 or 4 threads the FX 8350 will pull ahead as it has 8 core to play with. Core count is the FX line's only saving grace. And that FX 9590 is a CPU running at a suicide factory overclock with liquid cooling and a high end motherboard required to even use the damn thing. Then there is the need for a beefy PSU to run it as well because of the sheer amount of power it draws. Do not call me a fan boy on this because I have done my own research and comparisons. Also remember that to match an intel CPU at stock AMD needs to have massive overclocks, which are useless as you can overclock with Intel cpus as well on AIR and easily negate the overclocks of AMD cpus, then there is the fact that the 980 has 6 cores, 12 threads instead of 4 cores, 8 modules.

 

 

If that is what you believe, that will happen in the real world...

Then hey thats fine to me.

 

Im not going into argue about that.

I have personaly experianced, that the FX8350 is simply better then a core 2 Duo / Quad.

BUt thats my experiance!

 

-------------------------------------------------------------------------------------------------------------------------------------------

 

That the i7-980 is better then the FX8350, like i said in terms of raw cpu performance for rendering, it will be better yes. 6 cores 12 threads.

In streaming + gaming, till now only one chart of SKyrim shows the i7-980 scores higher then a FX8320.

That still isnt enough evidence for me.

 

But you are free to believe in what you want.

like i said i dont care

 

This is an open cummunity, for everyone to share their thoughts.

[Dabombinable]

If that is what you believe, that will happen in the real world...

Then hey thats fine to me.

 

Im not going into any discussion about that.

Also benchmark numbers like passmark numbers, dont tell me anything.

I have personaly experianced, that the FX8350 is simply better then a core 2 Duo / Quad.

 

But you are free to believe in what you want.

like i said i dont care

I know an FX 8350 will perform better than an Core 2 Quad or a Core 2 Duo in every day situations. Point is with the same amount of cores as an FX 8350 running at the same frequency as an FX 8350, the FX 8350 would get flogged. That can't happen of course because obviously intel has moved on from the Core 2 range and won't be adding more cores, but it is a testement to just how poorly AMD has designed their FX cpus architectures. Of course you are going to notice that an FX 8350 is faster than a core 2 quad because you weren't running it at the same frequency as your FX 8350, and because of the (repeatedly mentioned) lack of cores.

[Sintezza]

I know an FX 8350 will perform better than an Core 2 Quad or a Core 2 Duo in every day situations. Point is with the same amount of cores as an FX 8350 running at the same frequency as an FX 8350, the FX 8350 would get flogged. That can't happen of course because obviously intel has moved on from the Core 2 range and won't be adding more cores, but it is a testement to just how poorly AMD has designed their FX cpus architectures. Of course you are going to notice that an FX 8350 is faster than a core 2 quad because you weren't running it at the same frequency as your FX 8350, and because of the (repeatedly mentioned) lack of cores.

 

Ah then i didnt read your comment correctly sorry.

 

But its basicly not an apples to apples compairisson in the first place.

Because nobody is going to disable modules on a FX8 core lol.

 

But back to the main question, the i7-980 better in gaming + live streaming.

i would realy like to see some more evidence for that.

Till now only have seen skyrim.

I did no nowhere claim that the FX9590 was better then the i7 right?

 

Skyrim in particular is a game that realy likes core´s and cache.

Skyrim also run´s better on a 5820k then on an 4670K for example.

So its not that strange, that Skyrim performs better on the i7-980 6 core 12 threads.

[Faa]

i dont care about those benchmarks,

 

Core 2 is not superiour to a FX8350.

i have  Core 2 Quad, and the FX8350 is simply better

In fact, the C2Q is slow as fuck in compairisson to the FX8350, so is the Core 2 duo.

i owned them both.

 

And dont come with a statement that a core 2 duo is better then a core 2 quad single threaded, because thats the biggest laught i have ever seen, same cpu, same archtitecture, clockable at the same speed.

You can even disable 2 cores on the Q2Q and make it a C2D

Then you're wrong like always. We're saying it has more IPC, not better singlethreaded performance. Clock for clock (which is what we refer to with IPC) Conroe will outperform Vishera, you've gotten the proof you wanted end of it. 

Here;

Thats 17K for a 2500K at stock.

Feel free to check out the video -> youtube.com/watch?v=AlPMxaZS4Wo

At stock it's around 8K (it will perform definitely worse than that i5 661) as far as I can remember, that's just nearly a 60% gain orsomething wtf. Seeing there the difference a 3960x makes over a 2600K and the difference that one makes over the 2500K, it's taking advantage of more than 4 threads. Seeing the 8350 only performing 35% better, it's easily said the singlethreaded performance is miles better and the 4-threaded performance.

 

 

 

I just say my opinnion, based on my own experiances, since i owned a FX8350 and a core 2 quad.

Yeah except they're wrong. If people are allowed bringing subjectivity to a thread where people ask for objectivity then I might as well be like;

"8350 can't stream at 8GHz"

"8350 is not a CPU but a realtek based soundchip"

"8350 requires 2x 1500W PSU and won't work on 110V"

"You need a 8350mm fan to cool it down"

Nobody wants your Narnia fantasized opinions, give them facts they are asking.

 

Streaming i7-980 better then the FX9590, like i said idk.

I wanne see some real proof about that first, before i can say something usefull about that.

You've gotten your proof and apparently you aren't even aware of the i7 980 being a 6 core/12 thread CPU. 3930K with 17% less performance is alot faster than your 10 years old quad core with CMT.

 

 

I wanne see some real proof about that first, before i can say something usefull about that.

 

What about you make some benchmarks, run benchmarks that are singlethreaded and clock both CPU's at the same clock? If you don't post them it will automatically mean you found out that you were wrong. People who believe in conspiracy theories like crazy, what kinda things do they believe these days in? Illuminati orsomething, they're getting retarded and retarded over time while losing the track on reality. AMD fanboys are on such a path as well. Feel free to follow them, not going to make any difference anyways.

 

 

You know that im not going into any discussions with fanboys.

People who throw evidence out that always show Intel > AMD are AMD fanboys.

 

 

Has anyone pointed out to you yet how stupid it is to compare IPC's when AMD CPU's often come clocked significantly faster?  It's two different ways to get to the same place: if AMD had 50 IPC and 4.0 GHz and intel had 100 IPC at 2.0 GHz they'd perform very similarly.  Obviously, Intel has higher IPC and can certainly reach those 3.8-5.0 GHz clockspeeds and it tends to seriously outperform AMD in that range.  IPC * Clock speed = approximate CPU performance.  You can't just look at IPC.

You're basing it off the theory IPC x clock = CPU performance and mainly refering to instructions per cycle. You'll never know how much exactly the difference is, you can only compare the end results and you do that with the same clock speeds for both -> you refer with IPC to clock-for-clock performance. Since each core has its own clock speed, you refer to one core.

 

Lets see some real world tests, like video rendering.

 

You haven seen them, and you saw it getting massively outperformed (another guy who thinks a 3930K-17% is as good as a 8350).

[Sintezza]

 

Then you're wrong like always. We're saying it has more IPC, not better singlethreaded performance. Clock for clock (which is what we refer to with IPC) Conroe will outperform Vishera, you've gotten the proof you wanted end of it. 

Here;

Thats 17K for a 2500K at stock.

 

 

Where is the core 2 quad in this chart?

 

Just a benchmark list without any specs on which software version and what not is used.

And a just some printscreen of a video, is not evidence.

 

Show me better

[Faa]

Why are you changing subject to a 2500K?

please stay to the subject core 2 quad please.

Where is the core 2 quad in this chart?

 

And dont rotate my words thank you

 

Lets say, we end this discussion, because you chance from subjects.

I´m sorry you blown it.

I'm not changing the subject to a 2500K.

Here's the core 2 Quad, can't even believe you're asking for it when you removed that picture from the quote - how hilarious;

CPU score of 16K - 1K below a 2500K just 5% difference. Ask Harrynowls QX9650 cinebench score, nearly matching a 2500K stock as well.

Ending the discussion is on your side, you're just beating a dead horse that's all.

[Sintezza]

I'm not changing the subject to a 2500K.

Here's the core 2 Quad, can't even believe you're asking for it when you removed that picture from the quote - how hilarious;

CPU score of 16K - 1K below a 2500K just 5% difference. Ask Harrynowls QX9650 cinebench score, nearly matching a 2500K stock as well.

Ending the discussion is on your side, you're just beating a dead horse that's all.

 

Still the FX8350 beats it with 22k cpu score

It also beats the i7-965.

 

im sorry .

 

I think you beating a bunge of BS lol.

Please show me better evidence, of your claim

Because that Q2Q looks poor with 15k score lol.

 

Till then im out lol.

 

Like i said, i have owned them both so yeah..

[Sintezza]

I dont get your logic lol.

 

Your claim that Conroe out performs Vishera.

And you show benchmark in which the FX8350 scores 22k and the Core 2 Quad only 15K

 

lol

[Faa]

Still the FX8350 beats it with 22k cpu score

It also beats the i7-965.

 

im sorry .

 

I think you beating a bunge of BS lol.

Please show me better evidence, of your wrong claim

Because that Q2Q looks poor with 15k score lol.

 

Till then im out lol

Except that you're not realizing 3DMark vantage is heavily multithreaded and we both agreed a 8350 will outperform a Q9550 with 8 threads. Thought we were discussing singlethreaded performance or just 4 threaded performance? Point is proven. Poor C2Q yeah, what about matching the corecount? Two QX9775's that were released in 2007 based on the Conroe architecture from 2005 on a dual socket lga771 board clocked 1.2GHz lower than a 8350 at 5.4GHz;

Still faster. So 8 threaded performance between two QX9775's > 8350 at 5.4GHz

Single threaded performance between two Q9775's or just a single QX9550 at 4GHz > 8350 at 4GHz.

Q9550's are better for gaming.

[Sintezza]

Except that you're not realizing 3DMark vantage is heavily multithreaded and we both agreed a 8350 will outperform a Q9550 with 8 threads. Thought we were discussing singlethreaded performance or just 4 threaded performance? Point is proven. Poor C2Q yeah, what about matching the corecount? Two QX9775's that were released in 2007 based on the Conroe architecture from 2005 on a dual socket lga771 board clocked 1.2GHz lower than a 8350 at 5.4GHz;

Still faster. So 8 threaded performance between two QX9775's > 8350 at 5.4GHz

Single threaded performance between two Q9775's or just a single QX9550 at 4GHz > 8350 at 4GHz.

Q9550's are better for gaming.

 

i´m sorry you just talking BS.

 

2 Quadcore cpu´s VS  a FX8350, is completely a nonsense compairisson.

You do realise that the FX8350 does not work like a real 8 core right?

BEcause of the shared fpu and cache per module

 

It the same story as if you compair a i5-4440 to a 5960X right?

 

Im sorry, you just failed this time.

[Faa]

i´m sorry you just talking BS.

 

2 Quadcore cpu´s VS  a FX8350, is completely a nonsense compairisson.

You do realise that the FX8350 does not work like a real 8 core right?

BEcause of the shared fpu and cache per module

 

It the same story as if you compair a i5-4440 to a 5960X right?

 

Im sorry, you just failed this time.

Look look look who's being denial again. Is it so hard to accept the truth?

It's not the same story as a i5 vs 5960x. We are comparing 8 conroe cores vs AMD's what you call "8 cores".

Lets go for another round;

2 GHz higher overclocked and still getting smashed.

 

[Dabombinable]

Look look look who's being denial again. Is it so hard to accept the truth?

It's not the same story as a i5 vs 5960x. We are comparing 8 conroe cores vs AMD's what you call "8 cores".

Lets go for another round;

-snip-

2 GHz higher overclocked and still getting smashed.

 

I really need to get a dual socket mobo like yours, and another Xeon. That is insane performance for an old rig, and looks like it would smash my main rig too (btw, how did you manage to get the QX9775s working in tandem? I thought only Xeons could do that).

[The Crazed Child]

Equivalent according to the Tom's Hardware chart.

[Faa]

I really need to get a dual socket mobo like yours, and another Xeon. That is insane performance for an old rig, and looks like it would smash my main rig too (btw, how did you manage to get the QX9775s working in tandem? I thought only Xeons could do that).

They're not my screenshots obviously and I don't own a QX9775. QX9775's are LGA771 CPU's not LGA775 CPU's (however they work in lga775 boards). They are expensive on Ebay now, but there are some lga771 xeons that cost a lot less -> http://www.ebay.com/itm/Free-shipping-Intel-Xeon-X5450-LGA-771-Socket-J-SLBBE-CPU-Processor-3-GHz-/281450676054?pt=LH_DefaultDomain_0&hash=item4187c47f56

Motherboard 50$, so for 100$ you got something that's outperforming a 8350 and a 9590 when overclocked. I agree that performance is just insane for its age, so insane that Sintezza didn't want to believe her own eyes.