Haswell i3-4150 vs FX 8320/e for budget gaming rig?

[TheELF]

AMD is so great... yet you upgrade to a 4790k rather than an FX-9xxx

How would an Fx-9xxxx be an upgrade???
I thought you can easily O/C an 8320 to reach the same speeds?

 

[JumBie]

I noticed that the most vocal anti AMD posters do not own AMD gear, they are generaly just repeating hearsay and urban legends.

Maybe because we don't have any reason to own AMD "gear". We don't have to own computer hardware to know how it will perform, all of this can be seen through benchmarking and reviews. Its not like AMD CPU's are going to magically change from what was shown on paper just by purchasing it. We know without a doubt that the AMD fx line is just an overall inferior product to intel's line of CPU's, fanboyism aside. Like one of the posters said when AMD was a better option I purchased it, now that Intel is the superior product I will continue to purchase Intel CPU's. If down the line AMD releases a product that's price/performance ratio is excellent in comparison to Intels, then you can bet I will purchase that too.

 

[Abwx]

Maybe because we don't have any reason to own AMD "gear". We don't have to own computer hardware to know how it will perform, all of this can be seen through benchmarking and reviews.

Its not like AMD CPU's are going to magically change from what was shown on paper just by purchasing it. We know without a doubt that the AMD fx line is just an overall inferior product to intel's line of CPU's, fanboyism aside. Like one of the posters said when AMD was a better option I purchased it, now that Intel is the superior product I will continue to purchase Intel CPU's. If down the line AMD releases a product that's price/performance ratio is excellent in comparison to Intels, then you can bet I will purchase that too.

If your benchmarks are Sysmark, Cinebench or other 3D particle movement then you will have no idea of how it performs, but for sure it will give you the impression that Intel gear is much better.

For the record AMD does not pay firms to rig benches while Intel has no hesitations...

For benches i look at hardware.fr who use much more relevant tasks, and the scores tell me that AMD s CPUs are just fine, moreover when accounting the price/perf :

http://www.hardware.fr/articles/924-19/indices-performance.html

That said i m interested in your benches that show AMD as inferior....
Where are they.?
And are they relevant, because if it s a task performed in 0.25s instead of 0.15s it s just a worthless argument.

 

[JumBie]

If your benchmarks are Sysmark, Cinebench or other 3D particle movement then you will have no idea of how it performs, but for sure it will give you the impression that Intel gear is much better.

For the record AMD does not pay firms to rig benches while Intel has no hesitations...

For benches i look at hardware.fr who use much more relevant tasks, and the scores tell me that AMD s CPUs are just fine, moreover when accounting the price/perf :

http://www.hardware.fr/articles/924-19/indices-performance.html

That said i m interested in your benches that show AMD as inferior....
Where are they.?
And are they relevant, because if it s a task performed in 0.25s instead of 0.15s it s just a worthless argument.

Because we can go back and forth nitpicking forever.

 

[Abwx]

Because we can go back and forth nitpicking forever.

That s a hollow statement backed by nothing; i guess that the graphs are not what would be expected, hence the nitpicking straw.

The alleged nitpicking :

3DS Max Vray
3DS Max Mental ray
Visual Studio
MinGW-W64/GCC
Winrar
7 Zip
X264 encoding
X265 encoding
Lightroom
DxO
Houdini chess
Stockfish chess.

That s the softs in the application average graph, sure that it s no Sysmark...

Games used for 3D average :

Arma III
Crysis 3
Xplane 10
F1 2013
Watchdogs
Total war Rome II
Company of Heroes
Anno 2070

You will notice that ARMA III, Xplane 10, F1 2013 and Anno 2070 are mainly single threaded, that s 50% of the games used.

 

[SlowSpyder]

My own point is it's fine to purchase AMD if you're an AMD fan. If you're an intel fan it's fine to purchase intel. But don't bring fanboyism into recommending products for other users.

That's not at all what I was saying or meant to sound like I was saying if that's how it came across.

2js said that if AMD fans demanded more then they'd make better products. I don't think that what AMD's fans want would have much bearing on what AMD produced. The point I was making is that AMD losing sales and revenue wouldn't make their products any better if their fans left them (not that I think AMD's supposed fans really contribute much to their bottom line). I wasn't saying that anyone should or shouldn't buy AMD, just simply that less sales equals less revenue equals products that aren't going to be any better than what they have now as they would have that much less money for developing products. That's it, that's all.

 

[VirtualLarry]

Agreed.... I really don't get all the hate. If you buy an AMD with realistic expectations, you'll likely be very satisfied.

Obviously, buying an AMD-powered PC with one of their "small cores", and expecting to be able to Skype, listen to internet radio, copy files over the LAN, and web browse at the same time, is "unrealistic". I've bought them, tried that, and it... doesn't work. Poor PC is crawling along at a snail's pace, Skype sounds robotic, and LAN transfers virtually hang because they're going so slow due to lack of available CPU power.

Been there, done that, got the t-shirt, and bought Intel.

 

[Abwx]

Obviously, buying an AMD-powered PC with one of their "small cores", and expecting to be able to Skype, listen to internet radio, copy files over the LAN, and web browse at the same time, is "unrealistic". I've bought them, tried that, and it... doesn't work.

What did you bought actualy, a 2C 1.4 Kabini and then assumed that whatever AMD will be no better.?

Because it s what you did and stated in a certain thread.

Btw, i can do all the tasks you mentioned simultaneously with my Athlon 5350, i guess that you re talking without even having tested such gear.

 

[2is]

I noticed that the most vocal anti AMD posters do not own AMD gear, they are generaly just repeating hearsay and urban legends.

Sounds like you don't understand the concept of "anti" If they're anti-AMD, it would make sense that they don't own AMD gear. I actually do own a 7970 though. I have no issues with their GPU's. If I were to reject reality and substitute your own, I may agree with your urban legend and hearsay accusation. It wouldn't make me right, I'd just be in the same state of denial and delusion as you are

 

[Abwx]

Sounds like you don't understand the concept of "anti" If they're anti-AMD, it would make sense that they don't own AMD gear.

If it s so then they are not qualified to give a valuable opinion, they can state that said benches show AMD being slower if they want but they can in no way be credible based on just such estimations unless they understand something about what is benched and how, exemple is people using Cinebench to compare AMD and Intel CPU while this bench it totally irrelevant for such comparison unless you know exactly how much bias it has due to its use of, be ready for the list, ICC, Intel Thread Profiler and MKLs.

I actually do own a 7970 though. I have no issues with their GPU's. If I were to reject reality and substitute your own, I may agree with your urban legend and hearsay accusation. It wouldn't make me right, I'd just be in the same state of denial and delusion as you are

And i own Intel gear as well but that doesnt qualify me as being forcibly credible if i start to use urban legends, hearsay and other straws, thats why i m relying exclusively on numbers.

 

[VirtualLarry]

What did you bought actualy, a 2C 1.4 Kabini and then assumed that whatever AMD will be no better.?

Because it s what you did and stated in a certain thread.

Btw, i can do all the tasks you mentioned simultaneously with my Athlon 5350, i guess that you re talking without even having tested such gear.

It's a first-hand example of how poor AMD's performance is, in general. AMD bills those CPUs as being sufficient for "daily tasks", when the truth is, they are useless. (Edit: For more than one task at once.)

Don't buy any PC with anything less than an i3, a G3258 (and only if you plan to overclock it), or if you must buy AMD, then get their newest big-core APU dual-core/module chip. (2C/4T, basically AMD's equivalent of Intel's i3).

 

[Erenhardt]

fx8350 have a lot performance waiting for devs to use. Any 4 thread CPU is running at full load across all threads.

We are still waiting for software to catch and make a good use of 8 thread. 6 thread are finally showing the advantage over 4 threads.

Going from 4 to 6 threads (+50%) gets you 31% more fps.
Going from 6 to 8 threads (+33%) gets you less than 5% more fps

 

[2is]

If it s so then they are not qualified to give a valuable opinion, they can state that said benches show AMD being slower if they want but they can in no way be credible based on just such estimations unless they understand something about what is benched and how, exemple is people using Cinebench to compare AMD and Intel CPU while this bench it totally irrelevant for such comparison unless you know exactly how much bias it has due to its use of, be ready for the list, ICC, Intel Thread Profiler and MKLs.



And i own Intel gear as well but that doesnt qualify me as being forcibly credible if i start to use urban legends, hearsay and other straws, thats why i m relying exclusively on numbers.

Lucky for us, the benchmarks aren't just meaningless charts, they are titled, with an X and a Y axis and we know exactly what we are looking at. We can choose to look at cenebench, a myriad of gaming benchmarks, compression, encoding, photoshop, the list goes on.

Let me give another example. I don't need to have owned a Toyota Corrolla and Bugatti Veyron to know that the Bugatti is faster. Perhaps you do, but I have the mental fortitude to look at numbers and know which is the faster vehicle. All you need to do is understand what you're looking at. Your comments here don't prove hearsay or urban legends. In fact, what they prove is that you look at numbers and enter a state of confusion, not knowing what to do with them or how to make sense of them.

 

[Abwx]

It's a first-hand example of how poor AMD's performance is, in general. AMD bills those CPUs as being sufficient for "daily tasks", when the truth is, they are useless. (Edit: For more than one task at once.)

Don't buy any PC with anything less than an i3, a G3258 (and only if you plan to overclock it), or if you must buy AMD, then get their newest big-core APU dual-core/module chip. (2C/4T, basically AMD's equivalent of Intel's i3).

It s a first hand exemple to show how of bad faith you are, what would people say if i buy a 1.6 dual core Baytrail and then state that :

It's a first-hand example of how poor Intel's performance is, in general.

I wouldnt had expected someone to use such a worthless argument, not only that but you are stating lies :

Obviously, buying an AMD-powered PC with one of their "small cores", and expecting to be able to Skype, listen to internet radio, copy files over the LAN, and web browse at the same time, is "unrealistic". I've bought them, tried that, and it... doesn't work

You didnt buy "them", you did buy a single APU, the most weak that you could find, a 1.4 dual core Kabini, and then stated that all AMDs APUs dont work flawlessly.

 

[tential]

It's a first-hand example of how poor AMD's performance is, in general. AMD bills those CPUs as being sufficient for "daily tasks", when the truth is, they are useless. (Edit: For more than one task at once.)

Don't buy any PC with anything less than an i3, a G3258 (and only if you plan to overclock it), or if you must buy AMD, then get their newest big-core APU dual-core/module chip. (2C/4T, basically AMD's equivalent of Intel's i3).

Did you actually use Kabini and a Baytrail? Because I find it hard to believe the J1900 can do what you said. I don't think either processor excels at that.

I wanted Kabini originally until I actually saw the J1900 review. Then I changed my mind.
AMD is always "Almost there" and then they manage to royally screw up something I actually need.

Even for a budget Steam Box for InHome Streaming, I can't find an AMD competitor that can fit in the same form factors a J1900 can. I expected the AMD GPU to mean I could get great decoding for cheap and yet still the J1900 came out ahead at the same pricepoint/form factor.

There is always a compromise I'm just not willing to make when it comes to AMD's APU lineup.

If they could improve their manufacturing process though, I think we'd be telling a different story.

 

[VirtualLarry]

You didnt buy "them", you did buy a single APU, the most weak that you could find, a 1.4 dual core Kabini, and then stated that all AMDs APUs dont work flawlessly.

No, I own a C-60 Netbook, two C-70 NanoPCs (all three with SSDs), and the E1-2500 AIO (with a 5400RPM HDD). "Them."

They all suck. Badly.

I've also installed an A6-5400K Trinity APU at a friends place (with a 90GB Vertex2 SSD), and was amazed at how less-than-snappy it was. Then I realized, that AMD's marketing is what is "lies", not me, and that that APU was actually a single-core (with two threads). Which is exactly how it performs, and how Windows 8.1 identifies it in the Task Manager.

 

[Abwx]

Lucky for us, the benchmarks aren't just meaningless charts, they are titled, with an X and a Y axis and we know exactly what we are looking at. We can choose to look at cenebench, a myriad of gaming benchmarks, compression, encoding, photoshop, the list goes on.

For Cinebench you can use it for Intel/Intel or AMD/AMD comparisons but not for Intel/AMD comparisons.

Numbers say that it s optimised specificaly for Intel since version 11.5 and will give AMD 15-20% lower scores than what would be the case if there was no CPU dispatcher, this can be checked by comparing CB R10, CB 11.5 and CB R15 scores using two same recent CPUs.

Compare the successive versions scores betwen the tested Beema and the Haswell 4030U and you ll see the pattern, CB R15 recently just added more bias :

http://www.notebookcheck.net/HP-Pavilion-13-a093na-x360-Convertible-Review-Update.130928.0.html

And of course the necessary praise from Maxxon to Intel, only the guru is missing and it would be all nirvana in this video :

https://www.youtube.com/watch?v=PzLxCo5qofo

 

[VirtualLarry]

If they could improve their manufacturing process though, I think we'd be telling a different story.

Agreed. If Kabini could clock at 2.41 like Bay Trail does, then it would be a different story.

Btw, my N2830 laptop can do all of those things I mentioned. However, it cannot play back 1080P YouTube without skipping. Most of the time, the CPU is only at 15%, so it's not overloading the CPU. Then it has these spikes to 95% every few seconds. I don't know if it's a driver problem (using a RealTek USB3.0 wifi AC1200 dongle), or what.

 

[Abwx]

No, I own a C-60 Netbook, two C-70 NanoPCs (all three with SSDs), and the E1-2500 AIO (with a 5400RPM HDD). "Them."

They all suck. Badly.

Two 1Ghz dual core Bobcat and a 1.4 dual core Kabini; that s quite a representative sample, i suggest that you find a forum that is still in 2011 or so so you can compare them to clover trails, indeed i wouldnt had dared posting such exemples to sustain my point, it would have been a lost cause...

I've also installed an A6-5400K Trinity APU at a friends place (with a 90GB Vertex2 SSD), and was amazed at how less-than-snappy it was. Then I realized, that AMD's marketing is what is "lies", not me, and that that APU was actually a single-core (with two threads). Which is exactly how it performs, and how Windows 8.1 identifies it in the Task Manager.

Less snappy than what exactly.??.

Even for a budget Steam Box for InHome Streaming, I can't find an AMD competitor that can fit in the same form factors a J1900 can. I

What is this rare form factor.?

 

[Abwx]

Agreed. If Kabini could clock at 2.41 like Bay Trail does, then it would be a different story.

Kabini has 20-25% better Integer IPC than BT, at least, and 30-35% better IPC in FP, yet another argument that is void of sense, it takes a 2.5 BT to match an Athlon 5350 in Integer and 2.7ghz in FP..

I guess that discussing further is just useless given that s it s worse than hearsay, it s plain denial..

 

[Sweepr]

Kabini has 20-25% better Integer IPC than BT, at least, and 30-35% better IPC in FP, yet another argument that is void of sense, it takes a 2.6 BT to match an Athlon 5350 in Integer and 3ghz in FP..

I guess that discussing further is just useless given that s it s worse than hearsay, it s plain denial..

Yes, you're simply denying that for basic stuff like web browsing and light office use Bay Trail-D offers almost same CPU performance as Kabini in a fanless package. A faster GPU don't make your beloved Kabini automatically the best choice for everyone, deal with it instead of calling everyone who disagrees a fanboy.

 

[Abwx]

Yes, you're simply denying that for basic stuff like web browsing and light office use Bay Trail-D offers almost same CPU performance as Kabini in a fanless package..

Point me where i stated what i bolded in your post, you ll be hard pressed to find something, what i m denying is that the guy is saying that a 2.05 Kabini is not as potent as a J1900 and that it wouldnt perform better than a 2C 1.4 variant or 1Ghz 2C Bobcats.

http://nl.hardware.info/reviews/5330/amd-am1-vs-intel-bay-trail-d-review-goedkope-desktopplatforms

 

[VirtualLarry]

Point me where i stated what i bolded in your post, you ll be hard pressed to find something, what i m denying is that the guy is saying that a 2.05 Kabini is not as potent as a J1900 and that it wouldnt perform better than a 2C 1.4 variant or 1Ghz 2C Bobcats.

Maybe the 5350 performs better somehow, but in ST, going from 1.4 (Kabini) or 1.6 (Brazos) to 2.05 (Kabini AM1) doesn't seem like that much of an improvement to me.

My argument is basically that across the board, AMD chips have inferior performance to Intel, when compared directly against their "equivalent" processor from the other camp. Not that there aren't outlying corner cases where AMD's higest-end quad-core small-core CPU beats a mid-range Intel small-core.

 

[Abwx]

Maybe the 5350 performs better somehow, but in ST, going from 1.4 (Kabini) or 1.6 (Brazos) to 2.05 (Kabini AM1) doesn't seem like that much of an improvement to me.

But the 5350 has better ST overall than a J1900 and your Kabini not only is clocked at 1.4 but it s also a 2C only, yet seems to me that in previous thread i pointed to you that an Athlon 5350 has 3 times the processing power of a 1.6 2C Bobcat, for the record Bobcat has 10% better integer IPC than Baytrail in Integer and 15-20% in FP.

Anyway if possible find someone who has a 5350 and do some usage otherwise you ll be left doing assumptions that are indeed at odd with all 5350 onwers experience.

Edit : I linked Hardware.info review above, you have the 5350 and a J1900 compared.

 

[2is]

For Cinebench you can use it for Intel/Intel or AMD/AMD comparisons but not for Intel/AMD comparisons.

Numbers say that it s optimised specificaly for Intel since version 11.5 and will give AMD 15-20% lower scores than what would be the case if there was no CPU dispatcher, this can be checked by comparing CB R10, CB 11.5 and CB R15 scores using two same recent CPUs.

Compare the successive versions scores betwen the tested Beema and the Haswell 4030U and you ll see the pattern, CB R15 recently just added more bias :

http://www.notebookcheck.net/HP-Pavilion-13-a093na-x360-Convertible-Review-Update.130928.0.html

And of course the necessary praise from Maxxon to Intel, only the guru is missing and it would be all nirvana in this video :

https://www.youtube.com/watch?v=PzLxCo5qofo

I wouldn't doubt if some apps are compiled to be optimized for Intel processors, but that doesn't really matter. You're providing [part of] the reason why it performs better on Intel, but the end result is that it performs better. That's like saying the only reason this game works better on an AMD GPU is because of Mantle so it doesn't count.