FX chips better than i5 for video editing?

Page 8 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Abwx

Lifer
Apr 2, 2011
11,889
4,875
136
You moved the goal posts. I replied to perf/$. The Pentium smokes anything FX in perf/$. Period.

Have any proof that is an unadvisable overclock?

For one 4.8 is not guaranted, even with 100V, FX frequencies are guaranted.

So you are relying on completely random results, for the voltage Hardware.fr consider 1.25V as a safe limit and 1.3V only for short overclockings to check the frequency limit, that said i recommend to follow this guideline, feel free to advise people to get over them.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
Anand got 4.7 out of theirs. So let's say the average user gets 4.5. It still lays to waste any FX CPU in perf/$.

I don't suppose your favorite site reviewed it?

Edit:
Bit-tech got 4.8
Tweaktown got 4.5
Guru got 4.8
Techspot says 4.4 is a "mild" overclock on this chip.

We'll discuss "guaranteed" frequencies next time.
 
Last edited:

Abwx

Lifer
Apr 2, 2011
11,889
4,875
136
Anand got 4.7 out of theirs. So let's say the average user gets 4.5. It still lays to waste any FX CPU in perf/$.

At 1.35V also, so it will also consume close to 80W when stressed with P95, to give you an idea a FX , also with P95, will consume this amount at 3.5Ghz, with all cores loaded at 100% of course, i let you imagine the Pentium perf/watt efficency, for the perf/$ the FX is one more time better by far..

That s rare when comparing products with such prices delta, in principle the perf/$ ratio is systematicaly in favour of the cheap CPUs, you didnt knew this?

In this case though the Pentium manage to cost 50% of a FX with barely 50% of the perfs at almost 100% of the TDP, you see, you are promoting a complete disaster as a great product, although it can be quite good if you re Cinebenching on single threads..

I don't suppose your favorite site reviewed it?

Ask them but i guess that the fact that they didnt bother to test it is somewhat telling...
 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,002
126
Anand got 4.7 out of theirs. So let's say the average user gets 4.5. It still lays to waste any FX CPU in perf/$.

I don't suppose your favorite site reviewed it?

Edit:
Bit-tech got 4.8
Tweaktown got 4.5
Guru got 4.8
Techspot says 4.4 is a "mild" overclock on this chip.

We'll discuss "guaranteed" frequencies next time.


I would expect an FX6300 overclocked could put up a pretty good showing in price / performance vs. an overclocked Pentium. You can find the FX6300 for pretty near $100 these days.
 

jhu

Lifer
Oct 10, 1999
11,918
9
81
In what way does that matter?

(versus at-the-wall consumption for the entire platform, which is what the customer is paying the electrical company for when they use their computer to get the job done?)

It seems like an academic argument at best (who has the best architecture if you ignore XYZ, etc. etc.) but one that isn't relevant once you start down the road of "real world application and use models of actual customers" because those "actual customers" aren't just paying for the electricity the CPU uses.

I could have sworn you guys banned galego…
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
I would expect an FX6300 overclocked could put up a pretty good showing in price / performance vs. an overclocked Pentium. You can find the FX6300 for pretty near $100 these days.

I'll admit to not knowing anything about FX6300 overclocking. Are there any articles you would consider good for me to read?
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
I assure you that if you use the same set up and measure the actual CPU comsumption it does..

IMG0043831.png


http://www.hardware.fr/articles/913-9/cpu-consommation-overclocking.html

Can we seriously stop with the fritz? Its garbage okay. Almost no improvement vs. core 2 in IPC.
 

Abwx

Lifer
Apr 2, 2011
11,889
4,875
136
Can we seriously stop with the fritz? Its garbage okay. Almost no improvement vs. core 2 in IPC.

Actualy it s 12% or so, same results in 7zip, Haswell does a little better in Cbench though but it s still below Ibridge perf/watt wise by about 10-15% depending of the CPUs, you can check at Computerbase.de since they publish power comsumption and scores...
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
136
Can we seriously stop with the fritz? Its garbage okay. Almost no improvement vs. core 2 in IPC.

He seems to know this, kinda makes you wonder why he chose this particular benchmark.

And even Intel ones since a Haswell is only 10-12% better IPC wise in Fritz or 7zip than a core 2 but i guess that you thought that it was otherwise, hence your belief that it would do well only compared to AMD...

http://forums.anandtech.com/showpost.php?p=36786311&postcount=104
 

Abwx

Lifer
Apr 2, 2011
11,889
4,875
136
He seems to know this, kinda makes you wonder why he chose this particular benchmark.

http://forums.anandtech.com/showpost.php?p=36786311&postcount=104

About any bench will yield these results since the perf delta between IB and HW is 8-10% and that the power comsumption delta far exceed thoses 8-10% at 20-30%...

http://www.computerbase.de/2014-09/...2/#diagramm-abschliessendes-performancerating

http://www.computerbase.de/2014-09/...mm-leistungsaufnahme-volllast-cinebench-x-cpu
 

DominionSeraph

Diamond Member
Jul 22, 2009
8,386
32
91
Except it doesn't. Oops.

wh.png

Then let's not forget:

51143.png


So you pay more when you load it, and you pay more when you don't.

We're lucky that AMD doesn't own any power plants or Abwx would be going on about the "privilege" of using more of their product. I have a feeling that AMD is the reason his power stays on.
 
Last edited:

Abwx

Lifer
Apr 2, 2011
11,889
4,875
136
Then let's not forget:

51143.png


So you pay more when you load it, and you pay more when you don't.

12-13w more..?.

And looking at the CPU at the bottom it make me thinks that no one seemed to pay attention to all this not so long ago, there must be something that escaped me...
 

DominionSeraph

Diamond Member
Jul 22, 2009
8,386
32
91
And looking at the CPU at the bottom it make me thinks that no one seemed to pay attention to all this not so long ago, there must be something that escaped me...

To get this performance 20 years ago took megawatts and a room full of Crays. Technology advances. Except for AMD.
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
136
About any bench will yield these results since the perf delta between IB and HW is 8-10% and that the power comsumption delta far exceed thoses 8-10% at 20-30%...

http://www.computerbase.de/2014-09/...2/#diagramm-abschliessendes-performancerating

http://www.computerbase.de/2014-09/...mm-leistungsaufnahme-volllast-cinebench-x-cpu

13% higher system power consumption in CB11.5 and ~8.2% performance uplift. Again, very similar to other reviews (Ivy Bridge ~ Haswell). Perhaps if you repeat this 100x times it might come true, or not.

AnandTech said:
Here I’m showing an 11.8% increase in power consumption, and in this particular test the Core i7-4770K is 13% faster than the i7-3770K. Power consumption goes up, but so does performance per watt.

55330.png


So you pay more when you load it, and you pay more when you don't.

We're lucky that AMD doesn't own any power plants or Abwx would be going on about the "privilege" of using more of their product. I have a feeling that AMD is the reason his power stays on.

Good point. ;)

55329.png
 

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
12-13w more..?.

And looking at the CPU at the bottom it make me thinks that no one seemed to pay attention to all this not so long ago, there must be something that escaped me...
http://www.bit-tech.net/hardware/cpus/2010/12/07/amd-phenom-ii-x6-1100t-review/7

No, just that times change, and so do our PC parts. Nehalem's efficiency was lost with the bigger and faster chips, even with an efficient mobo, while the Phenoms had the reverse issue, in that while power hungry in general, the X6 weren't much more power hungry than the X4, and the X4 barely used more power than X2 or X3. With a good mobo, the Phenom IIs could idle lower than the high-end Intel CPUs (which can be seen in many reviews, but don't expect the average $70 mobo to do pull that off).

In ST workloads, and MT where parallelism was high, Intel's were better, but where IO was the issue, or MT with lots of shared data, that era of AMDs were pretty strong, and usually better for the money if that fit your needs (Phenom IIs where noticeably smoother in multitasking than Nehalems, with or without HT, as well, under Windows Vista and 7--hard, if not impossible, to pick out from benchmarks, but pretty much a universally accepted observation from experience). Intel's IGP also still sucked donkey balls, to the point regular non-techies noticed the difference, IME, and would be more pleased with a lower-performing AMD CPU than a higher-performing Intel, given IGP (on top of better compatibility with productivity apps), which was also good for overall non-gaming PC value.

Much has changed since 2010 :). The Phenom II X4 and X6, so long as you didn't get one of the bastard children w/ no L3, were excellent buys, several years ago, and helped keep Intel's prices in check, outside of the fastest i7 CPUs. With the OP's use case, a Phenom II would have been an excellent value in 2009-11. But, that was 3 CPU generations ago.
 
Last edited:

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
Last edited:

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,002
126
I'll admit to not knowing anything about FX6300 overclocking. Are there any articles you would consider good for me to read?

I'm not seeing anything other than forum posts. But, it seems 4.5 - 4.8GHz is fairly routine. I imagine a 4.6GHz FX6300 vs. a 4.5GHz Pentium would be close when you average benchmarks. That's just my semi educated guess though.
 

crashtech

Lifer
Jan 4, 2013
10,695
2,294
146

Yuriman

Diamond Member
Jun 25, 2004
5,530
141
106
I'm not seeing anything other than forum posts. But, it seems 4.5 - 4.8GHz is fairly routine. I imagine a 4.6GHz FX6300 vs. a 4.5GHz Pentium would be close when you average benchmarks. That's just my semi educated guess though.

It would depend very heavily on what benchmarks you included. I consider a stock i3 and FX-6 to be fairly good competitors, and it's a shame you can't overclock an i3. Pentium's lack of HT would make me reluctant to buy one.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
It would depend very heavily on what benchmarks you included. I consider a stock i3 and FX-6 to be fairly good competitors, and it's a shame you can't overclock an i3. Pentium's lack of HT would make me reluctant to buy one.

I'm tempted for a cheap Quicksync box, but I'd really have look into it as I use some pretty uncommon software for video editing that is highly optimized already.
 

Phynaz

Lifer
Mar 13, 2006
10,140
819
126
Ah, those were the days. I'm still quietly hoping AMD will spring some kind of miracle on us someday.

Here's another you may remember.

https://web.archive.org/web/20051109014729/http://www.amd.com/us-en/0,,3715_12353,00.html

I think they use the word "power" in every sentence :)

That miracle would be nice, but it would require a corporate change of direction. AMD has been saying for years they aren't going to play in the high end. They just don't have the cash, and the last time they tried (Abu Dabi) the product flopped as bad as Magny Cours did.
 
Last edited:

Samus

Golden Member
Jan 12, 2001
1,405
7
81
AMD CPU's can be competitive with i5's in performance, platform and price...until you consider the overclocking headroom the i5's have. My i5-4690 @ 4.6GHz on air cooling will outperform an FX-9590 at 6GHz on liquid nitrogen, while using half the power (and we're talking over 100-watts less!)

Then there's some sub-$200 Xeon's that are just rebadged i7's for those not interested in overclocking, such as the E3-1230v3 that'll drop into any 1150 motherboard. It'll be on par with a FX-9590 at 1/3rd the TDP, and you could put it in a Mini-ITX shoebox.

I hope AMD's new CEO turns this around, because they've proven before that Intel can be defeated, and there is a certain charm to AMD systems than Intel has always lacked; simpler sockets, better retension clip, lower-cost motherboards, etc.
 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,002
126
AMD CPU's can be competitive with i5's in performance, platform and price...until you consider the overclocking headroom the i5's have. My i5-4690 @ 4.6GHz on air cooling will outperform an FX-9590 at 6GHz on liquid nitrogen, while using half the power (and we're talking over 100-watts less!)

Then there's some sub-$200 Xeon's that are just rebadged i7's for those not interested in overclocking, such as the E3-1230v3 that'll drop into any 1150 motherboard. It'll be on par with a FX-9590 at 1/3rd the TDP, and you could put it in a Mini-ITX shoebox.

I hope AMD's new CEO turns this around, because they've proven before that Intel can be defeated, and there is a certain charm to AMD systems than Intel has always lacked; simpler sockets, better retension clip, lower-cost motherboards, etc.


I don't really have many games that have benches built in, but I'd be happy to run some CPU benches vs. your i5. I think the FX is more competitive than it is given credit for. Obviously not as efficient, but I don't think the performance (overall) is too bad.
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
AMD CPU's can be competitive with i5's in performance, platform and price...until you consider the overclocking headroom the i5's have. My i5-4690 @ 4.6GHz on air cooling will outperform an FX-9590 at 6GHz on liquid nitrogen, while using half the power (and we're talking over 100-watts less!)

Not on fully MT tasks.