Sign in with
Sign up | Sign in
Your question

AMD vs Intel Processors!

Last response: in Components
Share

Which gives the best performance?

Total: 0 vote

  • AMD .
  • 0 %
  • Intel.
  • 0 %
February 20, 2013 10:44:24 AM

Which offers the best performance on desktops/Gaming Rigs, AMD or Intel?

More about : amd intel processors

a b ) Power supply
February 20, 2013 10:49:13 AM

It's not always about performance - Price/Performance can be important as well - AMD competes well in that light. That being said - Intel.
-Bruce
a b ) Power supply
February 20, 2013 10:50:07 AM

For gaming, Intel's i5s are generally the best option.
Related resources
February 20, 2013 12:28:43 PM

intel>
i5 - great performance for a decent price
i7 - excellent performance all over the board that costs arm and leg
amd>
X4/X6 - good performance for a great price...
8350 - great performance for a good price, but only for specific tasks...

on a tight budget I'd go for an i5, otherwise an i7 :) 
a b ) Power supply
February 20, 2013 12:51:37 PM

i7 has little to no realistic advantage over i5 for gaming rig. Unless the budget is something like $2000 or more, there isn't good reason to get an i7 for a gaming-oriented system.
a b ) Power supply
a b À AMD
February 20, 2013 4:53:31 PM

Intel inside, idiot outside. AMD all the way.
February 20, 2013 5:01:55 PM

"on a tight budget I'd go for an i5, otherwise an i7"

What?

I've said a bazillion times, and it's not getting any less false.

The i5 2500k has 120% of the performance of a 965 BE. Surprise surprise! It has 200% of the price of a 965 BE!

AMD is king. Intel is only really useful for workstations where that 20% performance increase can turn a 10 day 3D render to an 8 day 3D render.
a b ) Power supply
a b À AMD
February 20, 2013 5:15:31 PM

dish_moose said:
It's not always about performance - Price/Performance can be important as well - AMD competes well in that light. That being said - Intel.
-Bruce

^ Pretty much sums it up ^
a b ) Power supply
February 20, 2013 5:25:55 PM

Intel offers the best performance, AMD is nowhere close to touch Intel. Intel has a bigger lineup and does also have CPUs for the biggest enthusiast, though those CPUs come at a great price.

AMD CPUs are right now lacking behind in every aspect.

Example the 8-core AMD FX 8350 should have great multitasking, right? Well guess again:

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - An i5-2400 beats an AMD FX 8350 overclocked to 4.8 GHz.

Gaming is the same story, so it is with editing and video encoding as well:

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - Encoding and editing

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - Gaming

Now that we have the numbers in place, AMD seems to offer great value, they also kinda do, the specifications they sell is awesome, the real world performance is something else.

If AMD had better CPUs I would switch my Intel based system though, I've always loved AMD since I build my very first desktop. I love their low prices and decent products, but if their low prices on the CPU market would come with great performance, then AMD would be more attractive.
a b ) Power supply
February 20, 2013 5:26:56 PM

MajinCry said:
"on a tight budget I'd go for an i5, otherwise an i7"

What?

I've said a bazillion times, and it's not getting any less false.

The i5 2500k has 120% of the performance of a 965 BE. Surprise surprise! It has 200% of the price of a 965 BE!

AMD is king. Intel is only really useful for workstations where that 20% performance increase can turn a 10 day 3D render to an 8 day 3D render.


Hehe, funny. You ignored the fact that the i5-2500K has much more overclocking headroom than the Phenom II x4 965 BE and that if you want to do stock comparisons, the i5-2500K is not worth considering when the i5-3470 has similar if not superior stock performance for significantly less money.

Furthermore, you aren't being truthful about the performance difference between a Phenom II x4 965 and an LGA 1155 i5 such as the i5-2500K even at stock either. The i5-2500K may only have a roughly 20% advantage in some games that aren't very CPU bottle-necked, but in any title where the CPU gets hit hard, that difference will easily move up towards ~40% and that's ignoring Turbo which can stretch it a little further.

There's also the considerable power consumption difference, the fact that an i3 can easily meet or beat the Phenom II x4s almost every time for about the same amount of money and even lower power consumption, and I could go on.

Hell, I'm not even an Intel fan. I have several AMD systems including the laptop that I'm using right now and one of my main machines that has a Phenom II x6 1090T BE and more. I'm just stating facts that you so callously ignored with your apples to oranges comparison.
a b ) Power supply
February 20, 2013 5:29:37 PM

lostgamer_03 said:
Intel offers the best performance, AMD is nowhere close to touch Intel. Intel has a bigger lineup and does also have CPUs for the biggest enthusiast, though those CPUs come at a great price.

AMD CPUs are right now lacking behind in every aspect.

Example the 8-core AMD FX 8350 should have great multitasking, right? Well guess again:

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - An i5-2400 beats an AMD FX 8350 overclocked to 4.8 GHz.

Gaming is the same story, so it is with editing and video encoding as well:

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - Encoding and editing

http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350... - Gaming

Now that we have the numbers in place, AMD seems to offer great value, they also kinda do, the specifications they sell is awesome, the real world performance is something else.


That link seems very screwed up. The FX-8350 has greater average highly threaded performance than any i5 even at stock. It's almost always between the LGA 1155 i5s and i7s when it comes to highly threaded performance and multi-tasking. Have a look at Tom's reviews and Anand's reviews of it for better accuracy.
a b ) Power supply
February 20, 2013 5:32:13 PM

blazorthon said:
That link seems very screwed up. The FX-8350 has greater average highly threaded performance than any i5 even at stock. It's almost always between the LGA 1155 i5s and i7s when it comes to highly threaded performance. Have a look at Tom's reviews and Anand's reviews of it for better accuracy.


According to the numbers, it's another story. I only rely my opinion here on the numbers, nothing else. According to Tomshardware the FX 8350 also bottlenecks the Radeon 7970, for some reason I can't seem to find that review. It was posted not too long ago. :) 
a b ) Power supply
February 20, 2013 5:35:41 PM

lostgamer_03 said:
According to the numbers, it's another story. I only rely my opinion here on the numbers, nothing else. According to Tomshardware the FX 8350 also bottlenecks the Radeon 7970, for some reason I can't seem to find that review. It was posted not too long ago.


When the numbers are not representative of the numbers of the other reviews, there's obviously something wrong.
a b ) Power supply
February 20, 2013 5:51:28 PM

blazorthon said:
When the numbers are not representative of the numbers of the other reviews, there's obviously something wrong.


How can you know that those numbers are representative? I haven't come across other benchmarks that strictly goes against those numbers. Teksyndicate is an exception. But I seem to not believe them, because I've looked at so many other benchmarks that supports the exact opposite, just as you said you did. It's hard getting a clear view on what is real these days.

a b ) Power supply
February 20, 2013 6:01:18 PM

blazorthon said:
http://www.tomshardware.com/reviews/fx-8350-vishera-rev...

You've seriously never seen the review done by the very site that you're talking on? The only times that i5s won in highly-threaded workloads was when the workload was done on a RAM drive and that's obviously because of AMD's inferior memory performance, not CPU performance.


Didn't I just mention that I did?

It's also kinda hard comparing results from different benchmarks.

Here is the review that I was talking of before.

http://www.tomshardware.com/reviews/fx-8350-core-i7-377...
a b ) Power supply
February 20, 2013 6:10:11 PM

lostgamer_03 said:
Didn't I just mention that I did?

It's also kinda hard comparing results from different benchmarks.

Here is the review that I was talking of before.

http://www.tomshardware.com/reviews/fx-8350-core-i7-377...


No, you didn't mention that and you specifically said that AMD was also losing in multi-tasking when that's actually their strength.

Whats the point of posting that review when it has nothing to do with highly threaded performance? We already know that AMD loses in lightly/moderately threaded performance such as most gaming situations compared to Intel's quad-core LGA 1155 CPUs.

It's actually not too difficult to compare benchmarks if you know what to look for. For example, looking at those multi-tasking tests from your link, it's easily proven that those tests aren't even using all eight threads of the AMD FX-8350, so of course it doesn't do well. This is made obvious by the i5s performing almost exactly identical to their i7 brothers in performance per Hz despite not having Hyper-Threading like the i7s have. Hyper-Threading and the six-core i7s should keep the i7s and i5s separated in six or more threaded work, yet they aren't in those tests.
a b ) Power supply
February 20, 2013 6:13:35 PM

blazorthon said:
No, you didn't mention that and you specifically said that AMD was also losing in multi-tasking when that's actually their strength.

Whats the point of posting that review when it has nothing to do with highly threaded performance? We already know that AMD loses in lightly/moderately threaded performance such as most gaming situations compared to Intel's quad-core LGA 1155 CPUs.

It's actually not too difficult to compare benchmarks if you know what to look for. For example, looking at those multi-tasking tests from your link, it's easily proven that those tests aren't even using all eight threads of the AMD FX-8350, so of course it doesn't do well. This is made obvious by the i5s performing almost exactly identical to their i7 brothers in performance per Hz despite not having Hyper-Threading like the i7s have. Hyper-Threading and the six-core i7s should keep the i7s and i5s separated in six or more threaded work, yet they aren't in those tests.


In the specific review it also does. As I said before, I only rely on the numbers.

That benchmark I just posted, was just to show you, that I ofcourse read the benchmarks and reviews on the site. I'm sorry if I seemed as if I didn't. I just don't see why synthethic tests means all that much. In the review I posted, they are not just running 1 program, but several and that does look more realistic to me. Because in the end, what do we want? Real-world performance or peformance in some synthethic test?
February 20, 2013 6:13:47 PM

lostgamer_03 said:


AMD CPUs are right now lacking behind in every aspect.



Incorrect.
a b ) Power supply
February 20, 2013 6:21:46 PM

smokeybravo said:
Incorrect.


So you're saying that an i7-3970x is on level with an AMD FX 8350? I would like you to elaborate.
February 20, 2013 6:51:28 PM

lostgamer_03 said:
So you're saying that an i7-3970x is on level with an AMD FX 8350? I would like you to elaborate.


He said every aspect, which is not true.
a b ) Power supply
February 20, 2013 6:58:35 PM

smokeybravo said:
He said every aspect, which is not true.


I said every aspect? And I don't see where you are right? If we're talking performance, and only performance, then there simply isn't a CPU by AMD that can beat the top of the line of Intel. Which is my point.
a b ) Power supply
February 20, 2013 6:59:24 PM

lostgamer_03 said:
In the specific review it also does. As I said before, I only rely on the numbers.

That benchmark I just posted, was just to show you, that I ofcourse read the benchmarks and reviews on the site. I'm sorry if I seemed as if I didn't. I just don't see why synthethic tests means all that much. In the review I posted, they are not just running 1 program, but several and that does look more realistic to me. Because in the end, what do we want? Real-world performance or peformance in some synthethic test?


Tom's ran dozens of real-world tests that were more applicable than any of the real but abnormal tests that were in the link that you posted. Besides that, your link had a far higher synthetics to real world ratio in tests than Tom's.

Running some GIMP, handbrake, Mplayer, and 7-Zip tests at once is not real-world. They don't even give info on what the tests are like Tom's did, so their legitimacy and what they even are is questionable, let alone what they prove.
a b ) Power supply
February 20, 2013 7:11:45 PM

blazorthon said:
Tom's ran dozens of real-world tests that were more applicable than any of the real but abnormal tests that were in the link that you posted. Besides that, your link had a far higher synthetics to real world ratio in tests than Tom's.

Running some GIMP, handbrake, Mplayer, and 7-Zip tests at once is not real-world. They don't even give info on what the tests are like Tom's did, so their legitimacy and what they even are is questionable, let alone what they prove.


Running 1 program isn't multi-tasking. The point of the numbers in that review is to show actual multi-tasking, which is in my book real-world performance.
a b ) Power supply
February 20, 2013 7:13:14 PM

lostgamer_03 said:
Running 1 program isn't multi-tasking. The point of the numbers in that review is to show actual multi-tasking, which is in my book real-world performance.


Your link failed to demonstrate what you want since it didn't even stress the CPUs properly. I'll take a test that is done properly over any skewed test that hides important context, especially since it makes no sense to run some of those programs together anyway. Even if they don't screw with each other's results, they'll still slow each other down beyond how it'd work if you did them individually. That'd be like trying to improve throughput by multi-tasking on a hard drive IE it will make the whole thing slower. Those programs aren't coded to play together nicely and the tests were obviously not adjusted to take that into account since they failed to properly use the resources of any of the tested CPUs.
February 20, 2013 7:55:27 PM

lostgamer_03 said:
I said every aspect? And I don't see where you are right? If we're talking performance, and only performance, then there simply isn't a CPU by AMD that can beat the top of the line of Intel. Which is my point.


Which aspects of performance are you referring to?
a b ) Power supply
February 20, 2013 8:07:49 PM

blazorthon said:
Your link failed to demonstrate what you want since it didn't even stress the CPUs properly. I'll take a test that is done properly over any skewed test that hides important context, especially since it makes no sense to run some of those programs together anyway. Even if they don't screw with each other's results, they'll still slow each other down beyond how it'd work if you did them individually. That'd be like trying to improve throughput by multi-tasking on a hard drive IE it will make the whole thing slower. Those programs aren't coded to play together nicely and the tests were obviously not adjusted to take that into account since they failed to properly use the resources of any of the tested CPUs.


"they'll still slow each other down beyond how it'd work if you did them individually". You just said it yourself, and AMD FX 8350 is slower during that process.
a b ) Power supply
February 20, 2013 8:52:58 PM

lostgamer_03 said:
"they'll still slow each other down beyond how it'd work if you did them individually". You just said it yourself, and AMD FX 8350 is slower during that process.


You still don't understand. Like I said before, you don't even know what tests they ran because they don't tell you. They just say application names. They don't tell you how they were run or even version number or any other such information. What is obvious looking at the results is that the tests did not run properly because they did not properly stress any of the eight nor twelve threaded CPUs. That includes Intel's. Those tests prove nothing because they don't tell you anything about them.

They're like me telling you that I tested my 5400RPM hard drive running at pretty much the same speed as my otherwise identical 7200RPM hard drive just because the test was running a huge copy/paste operation between two partitions on the drive and I never even tell you the test parameters. Anyone who knows much about hard drives know that if all else is equal, a 7200RPM drive is faster than a 5400RPM drive regardless of what the test said.

They may not know for sure what's wrong with the test, but that something is wrong is pretty obvious. Perhaps whatever messed up tests they were running disagreed with each other so much that they fought over the CPU cores. AMD having more cores and a modular design wouldn't help that sort of screwed up conflict, especially in an OS without the proper patches, which they clearly lacked in that review.
a b ) Power supply
February 20, 2013 8:54:27 PM

Intel is obviously more advanced and better for gaming, but I always buy AMD just because their stuff works good enough for every game I play and they need support.
a b ) Power supply
February 20, 2013 8:59:13 PM

swilczak said:
Intel is obviously more advanced and better for gaming, but I always buy AMD just because their stuff works good enough for every game I play and they need support.


Intel most certainly is better at gaming right now, but who is more advanced seems rather subjective.
a b ) Power supply
February 21, 2013 6:23:49 AM

blazorthon said:
You still don't understand. Like I said before, you don't even know what tests they ran because they don't tell you. They just say application names. They don't tell you how they were run or even version number or any other such information. What is obvious looking at the results is that the tests did not run properly because they did not properly stress any of the eight nor twelve threaded CPUs. That includes Intel's. Those tests prove nothing because they don't tell you anything about them.

They're like me telling you that I tested my 5400RPM hard drive running at pretty much the same speed as my otherwise identical 7200RPM hard drive just because the test was running a huge copy/paste operation between two partitions on the drive and I never even tell you the test parameters. Anyone who knows much about hard drives know that if all else is equal, a 7200RPM drive is faster than a 5400RPM drive regardless of what the test said.

They may not know for sure what's wrong with the test, but that something is wrong is pretty obvious. Perhaps whatever messed up tests they were running disagreed with each other so much that they fought over the CPU cores. AMD having more cores and a modular design wouldn't help that sort of screwed up conflict, especially in an OS without the proper patches, which they clearly lacked in that review.


I still don't get why you keep on trying. Just because the specifications on the FX 8350 seems better, then it doesn't mean it's faster. The architecture means a lot and just because the FX 8350 has more cores, higher cache and higher clock speeds, then it still is slower. What about the old FX 8150? It also had mind blowing specifications when it launched, still it was pretty bad because of the flaws of the Bulldozer architecture. SO that was a bad comparison with the 2 HDDs.

I neither still don't know why you keep on arguing about the tests not being right? They make several tests running at the same time and then take all the scores of the tests and average them.

IF you take a look at general performance, then they do all the tests individually and take all the scores and average them. They ran the same tests on every CPU of course, which brings me to the conclusion. FX 8350 simply isn't faster in those tests.
a b ) Power supply
February 21, 2013 1:52:24 PM

lostgamer_03 said:
I still don't get why you keep on trying. Just because the specifications on the FX 8350 seems better, then it doesn't mean it's faster. The architecture means a lot and just because the FX 8350 has more cores, higher cache and higher clock speeds, then it still is slower. What about the old FX 8150? It also had mind blowing specifications when it launched, still it was pretty bad because of the flaws of the Bulldozer architecture. SO that was a bad comparison with the 2 HDDs.

I neither still don't know why you keep on arguing about the tests not being right? They make several tests running at the same time and then take all the scores of the tests and average them.

IF you take a look at general performance, then they do all the tests individually and take all the scores and average them. They ran the same tests on every CPU of course, which brings me to the conclusion. FX 8350 simply isn't faster in those tests.

I did many tests on the FX 8320 which is almost as good as the 8350, and it only beats high end i7 processors in one area, integer math calculations. I'm not sure what integer math is and if it's used to play games though.
a b ) Power supply
February 21, 2013 2:01:07 PM

blazorthon said:
Intel most certainly is better at gaming right now, but who is more advanced seems rather subjective.

There is no doubt they are most advanced. They are anywhere from 12 to 18 months ahead of AMD in their release of new technology. Intel has been selling cpu's with 22nm technology for almost a year now, while AMD is still selling 32nm chips. Lithography is one of the main reason chips become faster and more efficient over time.
a b ) Power supply
February 21, 2013 8:00:19 PM

lostgamer_03 said:
I still don't get why you keep on trying. Just because the specifications on the FX 8350 seems better, then it doesn't mean it's faster. The architecture means a lot and just because the FX 8350 has more cores, higher cache and higher clock speeds, then it still is slower. What about the old FX 8150? It also had mind blowing specifications when it launched, still it was pretty bad because of the flaws of the Bulldozer architecture. SO that was a bad comparison with the 2 HDDs.

I neither still don't know why you keep on arguing about the tests not being right? They make several tests running at the same time and then take all the scores of the tests and average them.

IF you take a look at general performance, then they do all the tests individually and take all the scores and average them. They ran the same tests on every CPU of course, which brings me to the conclusion. FX 8350 simply isn't faster in those tests.


What I don't understand is how you got so far off track. My whole point up to now is that your link was useless and proved nothing because the tests were both false and not documented whatsoever, makign them incapable of proving your incorrect point. The real tests done by proper reviews such as Tom's and Anand's show that the FX-8350 is clearly between the i5s and i7s, closer to the i7s at that, when it comes to the job they are meant to do: pretty much anything that can use eight six to eight threads effectively.

IDK why you're going into irrelevant specifications and then acting as if I was focused on them when I am not. My HDD example works perfectly, hence the reason I gave it. You seem to not understand the point that I was trying to make with it, so I'm making it more blunt with this post. Your link was wrong. Tom's and Anand prove this.

Your link was wrong because of the tests being improper in several ways. First off, they were unrealistic. No one runs those software combinations together like that because it would make them all have skewed and useless results. Second, you aren't even told what was done with the software in the tests. You have no idea what exactly the tests even were because of that! Third, simply looking at the results clearly shows that the issues with the tests affected both AMD and Intel because the i5s and i7s were neck and neck instead of having a 20-40% jump between them like they should for any real multi-threading/multi-tasking performance test.

Just as clarification, all references to i5s and i7s in this post are the LGA 1155 models.

swilczak said:
There is no doubt they are most advanced. They are anywhere from 12 to 18 months ahead of AMD in their release of new technology. Intel has been selling cpu's with 22nm technology for almost a year now, while AMD is still selling 32nm chips. Lithography is one of the main reason chips become faster and more efficient over time.


How advanced the chips are is not determined specifically by the process node. If we go by architectural and conceptual advancement, AMD wins because Intel hasn't had a major architecture change since Core 2 and the modular concept that AMD is using is one of the most extreme changes in x86 CPUs ever. Unfortunately, it is also one of the least well-thought-out of those changes as AMD screwed up in implementing it, but how poorly implemented a concept is doesn't affect how advanced it is.
a b ) Power supply
February 21, 2013 8:00:51 PM

swilczak said:
I did many tests on the FX 8320 which is almost as good as the 8350, and it only beats high end i7 processors in one area, integer math calculations. I'm not sure what integer math is and if it's used to play games though.


Integer math is the type of math used by almost all consumer and most professional workloads done on a CPU, especially modern gaming. However, AMD usually still only meets or loses a little to the LGA 1155 i7s in it in real-world workloads except for a rare few examples, so I'd have to ask what i7s you are referring to in saying that AMD beat them.
a b ) Power supply
February 21, 2013 8:48:57 PM

blazorthon said:
Integer math is the type of math used by almost all consumer and most professional workloads done on a CPU, especially modern gaming. However, AMD usually still only meets or loses a little to the LGA 1155 i7s in it in real-world workloads except for a rare few examples, so I'd have to ask what i7s you are referring to in saying that AMD beat them.

The fx 8320 beats the i7 3770k in integer math by a lot, but not in other areas such as prime numbers, SSE, cpu compression, encryption, physics, sorting, and single threaded performance. I would still go with the 8320 just because it's so much cheaper.
a b ) Power supply
February 21, 2013 8:58:42 PM

Those are all floating point workloads. AMD wins a few floating point workloads, but yes, they lose in most (granted they can still be competitive).

I do agree that the 8320's pricing can make it a very attractive product so long as you intend to use it for workloads that it excels in and there's no shortage of them total.
a b ) Power supply
a b À AMD
February 21, 2013 9:11:30 PM

You pay twice as much for intel, but do you get twice the performance? Nope. But is it close? Nope.
a b ) Power supply
February 21, 2013 9:39:11 PM

bouncedk said:
You pay twice as much for intel, but do you get twice the performance? Nope. But is it close? Nope.


In many cases, you actually do get comparable or even superior performance for the money from Intel, but that's far less important than the point of people who's income depends on performance. If I'm making say 100grand a year and getting a $2000CPU instead of a $500 CPU can increase my income by even just 20%, that's extremely poor performance for the money but still somewhat faster CPU will make me much more money than the far cheaper but only somewhat slower CPU.

When it comes to tasks that can't run very parallel for whatever reason, Intel is irrefutably the better way to go unless you have reason to run a huge amount of them (if you can run enough instances of a single threaded task to make good use of a CPU with many cores such as AMD's CPUs, then they can again be the better option) and there are still some of those in the professional market.
March 9, 2013 11:33:34 AM

Any one does a proper review of the 8350 vs i5 3570K? Give link please.
AMD is a far lot cheaper than the Intel Processors. This is coming from a guy who had used a Phenom Black-Edition CPU before. It was awesome AT THE Time of it's release.
Of course AMD is still sticking on to the 32 nm fabrication process :(  .
!