Overclock.net › Forums › Industry News › Hardware News › [Inquirer] AMD sticks with Socket AM3+ for Steamroller, FM2 to get three years
New Posts  All Forums:Forum Nav:

[Inquirer] AMD sticks with Socket AM3+ for Steamroller, FM2 to get three years - Page 16

post #151 of 216
Quad channel will come into play when you can buy 16GB dimm. x8 is 128, plenty for a ram disk.

Regarding DDR4, makes less sense unless 16GB comes out the gate. But 2014 is long way from now, just nice to know the 990 chip set is good.
    
CPUMotherboardGraphicsGraphics
FX-8320 Gigabyte 990FXA UD3 Powercolor AX6850 Powercolor AX6850 
RAMHard DriveOptical DriveCooling
Corsair Vengeance LP CAS7 1600 (2 x 4GB) Samsung 840 Series Lite-On DVD-RW Corsair H80 
CoolingCoolingCoolingOS
Arctic Twin Turbo ll x2 Corsair AF140 x3 Corsair SP120 x2 Windows 7 Premium 64-bit 
MonitorKeyboardPowerCase
LG 32LH40 1080p 120Hz Interpolated 32" LCD Razer Arctosa Seasonic X650 Corsair 300R Windowed 
MouseMouse PadAudio
Razer Deathadder 2013 Razer Vespula Klipsch Promedia 2.1 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
FX-8320 Gigabyte 990FXA UD3 Powercolor AX6850 Powercolor AX6850 
RAMHard DriveOptical DriveCooling
Corsair Vengeance LP CAS7 1600 (2 x 4GB) Samsung 840 Series Lite-On DVD-RW Corsair H80 
CoolingCoolingCoolingOS
Arctic Twin Turbo ll x2 Corsair AF140 x3 Corsair SP120 x2 Windows 7 Premium 64-bit 
MonitorKeyboardPowerCase
LG 32LH40 1080p 120Hz Interpolated 32" LCD Razer Arctosa Seasonic X650 Corsair 300R Windowed 
MouseMouse PadAudio
Razer Deathadder 2013 Razer Vespula Klipsch Promedia 2.1 
  hide details  
Reply
post #152 of 216
Quote:
Originally Posted by 2010rig View Post

This is the problem.
"Competing" means different things, to different people.
http://www.anandtech.com/bench/Product/701?vs=697
http://www.anandtech.com/bench/Product/551?vs=697
The 8350 is clocked 500 - 600 MHZ higher than the CPU's it's "competing" against, what happens when they are both running the same clocks, or when they're both clocked at their MAX OC's?
In the 3570K vs 8350 comparison, the 8350 has a 600 MHZ advantage, and 4 extra cores, yet, the 3570K still wins the majority of the tests. Those 4 extra cores help give the 8350 some "wins" in multi-threaded tests, but it still loses overall.
But, but, but, the 8350 and 3570K are priced the same, so it's fair to compare them.
Of course it is, if AMD could have charged more they would've, but they can't for obvious reasons.
I really doubt AMD's ORIGINAL goals were to have a $200 "flagship" CPU.

But if you compare an FX-4300 with an 3570K at the same clock, PD doesn't compete with IB.
 
Ants In My HTPC
(8 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7-6700K (4.60Ghz @ 1.35v) Asus Z170 Pro Asus GTX 1070 8GB 32GB Kingston HyperX Fury DDR4 2133Mhz 
Hard DriveHard DriveCoolingCooling
Samsung 850 SSD (250GB) WD 3TB EK Supreme LTX CPU block (plexi) EK XT 360 radiator 
CoolingCoolingCoolingCooling
EK DCP 4.0 pump Phobya 250mm tube reservoir EK Compression Fittings Masterkleer 3/8" ID 1/2" OD tubing CLEAR 
CoolingOSMonitorMonitor
Mayhems Pastel White Windows 10 64bit iiyama 24" (1920x1080) Dell 19" (1440x900) 
KeyboardPowerCaseMouse
Corsair K70 (Cherry MX Red) XFX PRO 550W Core Edition NZXT Switch 810 (white) Razer DeathAdder Black Edition 3.5G 
Audio
Bose QC25 
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1090T (3.50Ghz @ 1.37v) AMD Sabertooth 990FX MSI GTX 460 1GB 8GB Corsair Vengeance DDR3 1333Mhz 
Hard DriveCoolingOSPower
Samsung 830 (128GB) Cooler Master Hyper 212X Windows 10 64bit Corsair VS550 
CPUGraphicsRAMHard Drive
i5 540M 2.53 GHz NVIDIA GeForce GT 330M 8 GB DDR3 1067 MHz Western Digital 1TB 
OSMonitorPowerCase
Mac OS X.6.8 17", 1920x1200 95-watt-hour lithium polymer battery Aluminum unibody MacBook Pro 
  hide details  
Reply
 
Ants In My HTPC
(8 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7-6700K (4.60Ghz @ 1.35v) Asus Z170 Pro Asus GTX 1070 8GB 32GB Kingston HyperX Fury DDR4 2133Mhz 
Hard DriveHard DriveCoolingCooling
Samsung 850 SSD (250GB) WD 3TB EK Supreme LTX CPU block (plexi) EK XT 360 radiator 
CoolingCoolingCoolingCooling
EK DCP 4.0 pump Phobya 250mm tube reservoir EK Compression Fittings Masterkleer 3/8" ID 1/2" OD tubing CLEAR 
CoolingOSMonitorMonitor
Mayhems Pastel White Windows 10 64bit iiyama 24" (1920x1080) Dell 19" (1440x900) 
KeyboardPowerCaseMouse
Corsair K70 (Cherry MX Red) XFX PRO 550W Core Edition NZXT Switch 810 (white) Razer DeathAdder Black Edition 3.5G 
Audio
Bose QC25 
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1090T (3.50Ghz @ 1.37v) AMD Sabertooth 990FX MSI GTX 460 1GB 8GB Corsair Vengeance DDR3 1333Mhz 
Hard DriveCoolingOSPower
Samsung 830 (128GB) Cooler Master Hyper 212X Windows 10 64bit Corsair VS550 
CPUGraphicsRAMHard Drive
i5 540M 2.53 GHz NVIDIA GeForce GT 330M 8 GB DDR3 1067 MHz Western Digital 1TB 
OSMonitorPowerCase
Mac OS X.6.8 17", 1920x1200 95-watt-hour lithium polymer battery Aluminum unibody MacBook Pro 
  hide details  
Reply
post #153 of 216
Quote:
Originally Posted by Usario View Post

If you filter out the poor excuse for a benchmark suite known as sysmark as well as games that are already over 60 fps despite testing with a GTX 280, the 3570k and 8350 have the same number of wins and losses.
Wow... is that it or do you have any more special rules to handicap the 3570k? 60 fps is certainly not the be all end all with 120Hz/3D monitors and of course the higher the average fps (generally speaking) the higher the minimums will be as well.
Quote:
Plus, Anandtech's testing suite is far from the most comprehensive. They'll include single-threaded versions of multi-threaded benchmarks and all the 10 tests that sysmark runs, but they'll for example only use one 3D rendering or one encryption benchmark.
What's the issue with including single thread results along with multi-thread for synthetic benchmarks? It gives a very clear picture of the performance of the architecture, which is ultimately the point of these types of comparisons. AMD's 8 core can pull ahead under heavily-threaded situations, but Intel's ipc lead absolutely crushes them in single threaded applications. That's relevant info to me.

If anything that makes them more comprehensive than most.

Unless you specifically buy a CPU to run Cinebench all day...

OT: I am kind of envious of AMD users having an endless stream of drop-in upgrades available. Maybe steamroller will be the big payoff people have been waiting for.
Edited by Booty Warrior - 12/14/12 at 5:39am
post #154 of 216
Quote:
Originally Posted by Capt View Post

They really need a new socket. It seems as they are getting nowhere with all the CPUs that they have released recently.

Erm, what CPUs have you been looking at? Piledriver is a great chip, it matches the chips in its price range (i5 3570k) in CPU intensive tasks while having an unnoticeable difference in gaming, AMD has actually been gaining more performance from each new chip than Intel recently.

What would a new socket achieve for them, anyway? The multi-GPU bottleneck is mainly in the chipset, AMD could have a 1090FX that fixes it if they wanted to release a fix.
Quote:
Originally Posted by Liranan View Post

Releasing a socket with no improvements in the architecture serves no purpose. AMD will switch to quad channel memory but not before it's a necessity, the memory doesn't bottlenecks the CPU's and Intel have proven that bar benchmark numbers there's no difference between dual, triple and quad channel RAM.

The problems AMD have with SLI is due to the north bridge and their PCI-E configuration, rather than the CPU. A dual GPU card isn't bottlenecked by the FX8350 (slightly by the 8150).

And what Brutuz is saying about cores is totally correct. Currently you are better off with a slower quad core than a super fast dual core, which is why quite a few games have quad cores as recommended and faster dual cores as minimum requirements.

Actually, having quad channel memory can be useful in extraction as the i7 3820 proves. It's a small difference and for consumers the die space is better used by many other things.
Quote:
Originally Posted by Fuell View Post

I'm confused.

People say:
Bulldozer/Zambezi was a total fail.
Phenom II was behind Nehalem.
SB was a decent improvement over nehalem.
Bulldozer was a step back from Phenom II.

At this point the gap should be massive, considering how people talk about these chips as stated above. Almost like a 2-2.5 gen gap so to speak.

Now IB improves slightly upon SB.
PD is a decent improvement over BD, but nothing miraculous.

And now PD is competing with IB. So... how was BD garbage again? Confusing when you think about it purely based on the communities reactions. If PB is such a good competitor, I'm starting to think people severely over-exaggerated how bad BD actually was...

Essentially, a lot of people say BD is garbage because they were comparing straight IPC without OCs bar a few benchmarks. Add in stuff being compiled in a way that works better for BD and yeah.
BD was severely under-rated, PD is 5% faster in single-threaded and (iirc) 10-15% faster in multi-threaded stuff which lets the FX-83*0 match the i5 3570k, for the most part BD was a side-step and people were disappointed; the main reason it's classed as a downgrade is because power consumption is insanely high.
Quote:
Originally Posted by Homeles View Post

Quote:
Originally Posted by Fuell View Post

I'm confused.
And now PD is competing with IB. So... how was BD garbage again? Confusing when you think about it purely based on the communities reactions. If PB is such a good competitor, I'm starting to think people severely over-exaggerated how bad BD actually was...
With Bulldozer, you had two crowds: the guys that saw it for what it was, and the guys that worshiped it. It was a stark dichotomy with very few exceptions.

And in case you haven't noticed, the gains from SB -> IB were smaller than the ones from BD -> PD, so your argument has no factual backing.

Actually, you had the people who looked at review sites and claimed small differences were massive and people who owned one and could tell the difference was zero in reality. I literally paid $0 for my FX, I've got no buyers remorse, the main problem with BD was its stock clocks.
Quote:
Originally Posted by CiBi View Post

Quote:
Originally Posted by Fuell View Post

I'm confused.
People say:
Bulldozer/Zambezi was a total fail.
Phenom II was behind Nehalem.
SB was a decent improvement over nehalem.
Bulldozer was a step back from Phenom II.
At this point the gap should be massive, considering how people talk about these chips as stated above. Almost like a 2-2.5 gen gap so to speak.
Now IB improves slightly upon SB.
PD is a decent improvement over BD, but nothing miraculous.
And now PD is competing with IB. So... how was BD garbage again? Confusing when you think about it purely based on the communities reactions. If PB is such a good competitor, I'm starting to think people severely over-exaggerated how bad BD actually was...

People did, the hype built up to insane levels so people were let down by something that merely matched Phenom II. People were complaining even though the new architecture is a good thing, a few tweaks and we got PD, few more tweaks and we've got SR..That's 30% extra performance from BD to SR if rumours are to be believed, more than Intel has had in the same time. (To be fair, Haswell has TSX, etc which will dramatically increase performance)

PD is not competing with IB, its nowhere near as good. And Piledriver might be an improvement over Bulldozer but its still only slightly better then Thuban and Deneb (aka Phenom II)

Look at stuff that isn't synthetic, for the most part it manages to make use of 8+ cores and PD matches the i5 3570k (it's competitor) pretty well.

Proof.
More proof.
Quote:
Originally Posted by 2010rig View Post

Quote:
Originally Posted by CiBi View Post

PD is not competing with IB, its nowhere near as good. And Piledriver might be an improvement over Bulldozer but its still only slightly better then Thuban and Deneb (aka Phenom II)

This is the problem.

"Competing" means different things, to different people.

http://www.anandtech.com/bench/Product/701?vs=697
http://www.anandtech.com/bench/Product/551?vs=697

The 8350 is clocked 500 - 600 MHZ higher than the CPU's it's "competing" against, what happens when they are both running the same clocks, or when they're both clocked at their MAX OC's?

In the 3570K vs 8350 comparison, the 8350 has a 600 MHZ advantage, and 4 extra cores, yet, the 3570K still wins the majority of the tests. Those 4 extra cores help give the 8350 some "wins" in multi-threaded tests, but it still loses overall.

But, but, but, the 8350 and 3570K are priced the same, so it's fair to compare them.

Of course it is, if AMD could have charged more they would've, but they can't for obvious reasons.

I really doubt AMD's ORIGINAL goals were to have a $200 "flagship" CPU.

Firstly; take out the synthetic benchmarks and stuff that's made single-threaded when there's also a multi-threaded version (Who is going to use a single core for something they can use 8 for?) and games (The FPS difference is unnoticeable unless you're already out of AMDs target market), FX-8350 suddenly has most of the benchmarks over the i5 3570k, also remember that a few of those are "Lower is better". AMD wins 13, Intel wins 8. (I took out all of sysmark, all the games, the POVRay and Cinebench singlethreaded stuff)

Now also remember, the difference between those for the most part is literally 2-3 seconds, apart from a few. Also take context into account, for all of the x264 encoding ones, the 1st pass is mostly single-threaded but a hell of a lot shorter than the 2nd pass..In which AMD is a hell of a lot faster than Intel, look at how the newer versions of x264 encoding also have higher performance on AMD too, the difference between PD and IvB are pretty small, and as Usario said, thanks to Turbo the difference is more like ~300Mhz, considering AMDs Turbo implementation is (As far as I know) utter crap, it also could be even as low as 200Mhz. Average overclocks on hwbot is 4.97Ghz on the FX-8350 and 4.71Ghz on the i5 3570k too.
Quote:
Originally Posted by Booty Warrior View Post

Quote:
Originally Posted by Usario View Post

If you filter out the poor excuse for a benchmark suite known as sysmark as well as games that are already over 60 fps despite testing with a GTX 280, the 3570k and 8350 have the same number of wins and losses.
Wow... is that it or do you have any more special rules to handicap the 3570k? 60 fps is certainly not the be all end all with 120Hz/3D monitors and of course the higher the average fps (generally speaking) the higher the minimums will be as well.

Why the hell would you compare a synthetic benchmark when there's plenty of results from real world applications there? I'm not going to get a graphics card because 3DMark said it's better, I'm going to look at real world results..I can understand the argument for using them in games (Repeatable results, although thanks to them not showing the same difference as actual games it really does need to be called into question) but when you're comparing media encoding, compression, etc it's so easy to make it exactly the same that there's absolutely no use for them.

If you're running a 3D monitor, why are you considering an AMD CPU? You're obviously mainly into gaming, in which case the i5 3570k is the better choice, if you do stuff other than gaming or game at 60Hz then AMD is practically equal to Intel until you hit above the $250 mark, there's just a tonne of people who will read a review, look at the coloured graphs (Sometimes not even noticing it's a lower is better result) and claim AMD is worse when they're competing with Intel pretty well.
Quote:
Originally Posted by Booty Warrior View Post

Quote:
Plus, Anandtech's testing suite is far from the most comprehensive. They'll include single-threaded versions of multi-threaded benchmarks and all the 10 tests that sysmark runs, but they'll for example only use one 3D rendering or one encryption benchmark.

What's the issue with including single thread results along with multi-thread for synthetic benchmarks? It gives a very clear picture of the performance of the architecture, which is ultimately the point of these types of comparisons. AMD's 8 core can pull ahead under heavily-threaded situations, but Intel's ipc lead absolutely crushes them in single threaded applications. That's relevant info to me.

If anything that makes them more comprehensive than most.

Unless you specifically buy a CPU to run Cinebench all day...

OT: I am kind of envious of AMD users having an endless stream of drop-in upgrades available. Maybe steamroller will be the big payoff people have been waiting for.

IPC has zero to do with core count..AMD gets ahead because they have more cores than Intel at the same price point, and if you're doing something that truly does max out your CPU chances are it's going to be using all 8 cores anyway. There's no point in comparing single-threaded performance when you're trying to do a real-world comparison simply because why the hell would I run (For example) POVRay in single-threaded mode when I've got 7 more cores sitting there doing nothing? You can talk single-threaded performance all day but for the most part, it matters very little. AMD is ahead in compression, encryption, transcoding and rendering. Steamroller looks to increase single-threaded performance quite a bit thanks to giving each core its own decoder.
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
post #155 of 216
Quote:
Originally Posted by Liranan View Post

It has 4 cores and 8 threads, your point is?

Alatar was talking about the hexacores.
   
AGP bencher
(14 items)
 
CPUMotherboardGraphicsRAM
Ryzen R7 1700 Gigabyte GA-AX370-Gaming 5 Sapphire HD 6950 2GiB 2x8GB KFA2 HOF DDR4-3600 
Hard DriveHard DriveHard DriveHard Drive
Crucial MX100 256GB Seagate 600 Series 240GB Seagate 7200.14 2TB Samsung F3 1TB 
CoolingCoolingCoolingCooling
EKWB Supreme HF XSPC Rasa GPU EK XT360 EK 4.0 
OSMonitorMonitorKeyboard
W10 Pro LG IPS235 LG E2250V KUL ES-87 
PowerCaseMouseAudio
SF Leadex II 650W Lian Li PC-A05NB Logitech G9 Xonar DX 
AudioAudio
SMSL SA-S3+Technics CB-250 Sennheiser HD555 
CPUMotherboardRAMHard Drive
AMD A10-5700 Gigabyte F2A75M-HD2 G.SKILL Ares 2133 CL9 Hitachi 5K750 
Hard DriveCoolingOSMonitor
Momentus .7 200GB Noctua NH-L9a Server 2012 R2 Standard AUO B156HW01 
PowerCaseOther
PicoPSU-80-WI-25V AIO Aluminium Handmade TP-Link Archer Something Something Wi-Fi AC 
CPUCPUCPUMotherboard
Core2Duo E6400 Core2Quad Q6600 Pentium Dual Core E5200 AsRock 4COREDUAL-SATA2 R2.0 
GraphicsRAMHard DriveOptical Drive
A dumpload of ancient AGP cards Kingston Value DDR2-667 CL4 2T @CL3 1T Seagate 160GB 7200.10 LG IDE DVD-ROM 
CoolingCoolingOSMonitor
Ghettomade CPU waterblock 49cc 2stroke engine copper radiator WinXP SP2 32bit ProView 17" 
PowerCase
Tacens Radix V 550W Ghetto aluminium bench 
  hide details  
Reply
   
AGP bencher
(14 items)
 
CPUMotherboardGraphicsRAM
Ryzen R7 1700 Gigabyte GA-AX370-Gaming 5 Sapphire HD 6950 2GiB 2x8GB KFA2 HOF DDR4-3600 
Hard DriveHard DriveHard DriveHard Drive
Crucial MX100 256GB Seagate 600 Series 240GB Seagate 7200.14 2TB Samsung F3 1TB 
CoolingCoolingCoolingCooling
EKWB Supreme HF XSPC Rasa GPU EK XT360 EK 4.0 
OSMonitorMonitorKeyboard
W10 Pro LG IPS235 LG E2250V KUL ES-87 
PowerCaseMouseAudio
SF Leadex II 650W Lian Li PC-A05NB Logitech G9 Xonar DX 
AudioAudio
SMSL SA-S3+Technics CB-250 Sennheiser HD555 
CPUMotherboardRAMHard Drive
AMD A10-5700 Gigabyte F2A75M-HD2 G.SKILL Ares 2133 CL9 Hitachi 5K750 
Hard DriveCoolingOSMonitor
Momentus .7 200GB Noctua NH-L9a Server 2012 R2 Standard AUO B156HW01 
PowerCaseOther
PicoPSU-80-WI-25V AIO Aluminium Handmade TP-Link Archer Something Something Wi-Fi AC 
CPUCPUCPUMotherboard
Core2Duo E6400 Core2Quad Q6600 Pentium Dual Core E5200 AsRock 4COREDUAL-SATA2 R2.0 
GraphicsRAMHard DriveOptical Drive
A dumpload of ancient AGP cards Kingston Value DDR2-667 CL4 2T @CL3 1T Seagate 160GB 7200.10 LG IDE DVD-ROM 
CoolingCoolingOSMonitor
Ghettomade CPU waterblock 49cc 2stroke engine copper radiator WinXP SP2 32bit ProView 17" 
PowerCase
Tacens Radix V 550W Ghetto aluminium bench 
  hide details  
Reply
post #156 of 216
Quote:
Originally Posted by Booty Warrior View Post

Quote:
Originally Posted by Usario View Post

If you filter out the poor excuse for a benchmark suite known as sysmark as well as games that are already over 60 fps despite testing with a GTX 280, the 3570k and 8350 have the same number of wins and losses.
Wow... is that it or do you have any more special rules to handicap the 3570k? 60 fps is certainly not the be all end all with 120Hz/3D monitors and of course the higher the average fps (generally speaking) the higher the minimums will be as well.
Quote:
Plus, Anandtech's testing suite is far from the most comprehensive. They'll include single-threaded versions of multi-threaded benchmarks and all the 10 tests that sysmark runs, but they'll for example only use one 3D rendering or one encryption benchmark.
What's the issue with including single thread results along with multi-thread for synthetic benchmarks? It gives a very clear picture of the performance of the architecture, which is ultimately the point of these types of comparisons. AMD's 8 core can pull ahead under heavily-threaded situations, but Intel's ipc lead absolutely crushes them in single threaded applications. That's relevant info to me.

If anything that makes them more comprehensive than most.

Unless you specifically buy a CPU to run Cinebench all day...

OT: I am kind of envious of AMD users having an endless stream of drop-in upgrades available. Maybe steamroller will be the big payoff people have been waiting for.

How is filtering useless, irrelevant benchmarks like sysmark and games tested at 1024x768 with a GTX 280 pulling 200fps "handicapping" the 3570k? Oh, my bad, the i5 is only better if you obsess over graphs showing synthetic performance instead of actually using your computer.

Single-threaded versions of multi-threaded benchmarks are worth zero. If a workload can be adequately parallelized, there's no reason at all to factor in silly what-if scenarios when deciding what chip to buy. Legitimate lightly-threaded workloads, on the other hand, are definitely fair game.

If you're pulling 60 fps with a 280 I think it'll be pretty easy to pull 120 with a 480, let alone a 680.
Kalki
(25 items)
 
Alienware
(18 items)
 
ASUS A53Z
(13 items)
 
CPUMotherboardGraphicsRAM
AMD FX-8320 @ 4.73GHz Gigabyte 990FXA-UD3 Radeon HD 7950 @ 1200/1800 8GB Samsung 30nm DDR3 @ 2000MHz 11-11-11-28-1T 
Hard DriveHard DriveHard DriveOptical Drive
128GB Crucial M4 1TB Seagate Barracuda 7200.12 2TB Seagate Barracuda ST 7200 RPM who cares 
CoolingCoolingCoolingCooling
XSPC Raystorm XSPC Razor HD 7970 XSPC EX240 Swiftech MCP50X 
CoolingCoolingCoolingCooling
Black Ice GT Stealth 240 2x XSPC Xinruilian 1650rpm 2x Yate Loon Slim 1800rpm Alphacool Light Tower 
OSOSMonitorKeyboard
Windows 7 Gentoo GNU/Linux Dell UltraSharp U2412M CM Storm Trigger MX Brown 
PowerCaseMouse
Corsair AX750 Corsair Carbide 300R Logitech G400 
CPUMotherboardGraphicsRAM
AMD OPTERON BULLDOZER 16-CORE 2GHZ (32GHZ TOTAL) ALIENWARE 3+1 PHASE ITX GAMING BOARD 4GB NVIDIA GEFORCE GT 430 32GB GAMING DDR3 800MHZ CL13 
Hard DriveOptical DriveCoolingOS
2TB 5400RPM GAMING HDD GAMING BLU RAY BURNER ONE GAMING CASE FAN UBUNTU W/ UNITY (I AM 1337 H4X0R) 
MonitorKeyboardPowerCase
HD-READY 24" 1366X768 PENTILE TN PANEL GAMING M... ALIENWARE TACTX UBER 1337 BACKLIT RUBBER MEMBRA... OFF-BRAND 200W GAMING PSU; 65% EFFICIENCY  ALIENWARE GAMING CASE WITH THE COOL GLOWING ALIEN 
MouseMouse PadAudioAudio
ALIENWARE TACTX GAMING MOUSE ALIENWARE MOUSEPAD IT'S DESIGNED FOR GAMING AND... BEATS AUDIO 1337 ALIENWARE TACTX 8-CHANNEL GAMING HEADSET/M... 
OtherOther
16 CORES!!!! 4GB VRAM!!!! FASTER THAN 3960X + 7... ALIENWARE GAMING CARPET 
CPUMotherboardGraphicsRAM
AMD A6-3420M OC 2.2GHz / 3GHz Turbo 1.10625v ASUS A53Z AMD Radeon HD 6520G 6GB DDR3 1333MHz 
Hard DriveOptical DriveCoolingOS
320GB WD Scorpio Black yes. really good Windows 7 Home Premium 64-bit 
OSMonitorKeyboardPower
Fedora GNU/Linux 1366x768 LED-backlit TN panel Decent + numpad ~5hr 6-cell li-ion 
Audio
Altec Lansing 
  hide details  
Reply
Kalki
(25 items)
 
Alienware
(18 items)
 
ASUS A53Z
(13 items)
 
CPUMotherboardGraphicsRAM
AMD FX-8320 @ 4.73GHz Gigabyte 990FXA-UD3 Radeon HD 7950 @ 1200/1800 8GB Samsung 30nm DDR3 @ 2000MHz 11-11-11-28-1T 
Hard DriveHard DriveHard DriveOptical Drive
128GB Crucial M4 1TB Seagate Barracuda 7200.12 2TB Seagate Barracuda ST 7200 RPM who cares 
CoolingCoolingCoolingCooling
XSPC Raystorm XSPC Razor HD 7970 XSPC EX240 Swiftech MCP50X 
CoolingCoolingCoolingCooling
Black Ice GT Stealth 240 2x XSPC Xinruilian 1650rpm 2x Yate Loon Slim 1800rpm Alphacool Light Tower 
OSOSMonitorKeyboard
Windows 7 Gentoo GNU/Linux Dell UltraSharp U2412M CM Storm Trigger MX Brown 
PowerCaseMouse
Corsair AX750 Corsair Carbide 300R Logitech G400 
CPUMotherboardGraphicsRAM
AMD OPTERON BULLDOZER 16-CORE 2GHZ (32GHZ TOTAL) ALIENWARE 3+1 PHASE ITX GAMING BOARD 4GB NVIDIA GEFORCE GT 430 32GB GAMING DDR3 800MHZ CL13 
Hard DriveOptical DriveCoolingOS
2TB 5400RPM GAMING HDD GAMING BLU RAY BURNER ONE GAMING CASE FAN UBUNTU W/ UNITY (I AM 1337 H4X0R) 
MonitorKeyboardPowerCase
HD-READY 24" 1366X768 PENTILE TN PANEL GAMING M... ALIENWARE TACTX UBER 1337 BACKLIT RUBBER MEMBRA... OFF-BRAND 200W GAMING PSU; 65% EFFICIENCY  ALIENWARE GAMING CASE WITH THE COOL GLOWING ALIEN 
MouseMouse PadAudioAudio
ALIENWARE TACTX GAMING MOUSE ALIENWARE MOUSEPAD IT'S DESIGNED FOR GAMING AND... BEATS AUDIO 1337 ALIENWARE TACTX 8-CHANNEL GAMING HEADSET/M... 
OtherOther
16 CORES!!!! 4GB VRAM!!!! FASTER THAN 3960X + 7... ALIENWARE GAMING CARPET 
CPUMotherboardGraphicsRAM
AMD A6-3420M OC 2.2GHz / 3GHz Turbo 1.10625v ASUS A53Z AMD Radeon HD 6520G 6GB DDR3 1333MHz 
Hard DriveOptical DriveCoolingOS
320GB WD Scorpio Black yes. really good Windows 7 Home Premium 64-bit 
OSMonitorKeyboardPower
Fedora GNU/Linux 1366x768 LED-backlit TN panel Decent + numpad ~5hr 6-cell li-ion 
Audio
Altec Lansing 
  hide details  
Reply
post #157 of 216
Quote:
Originally Posted by Usario View Post

How is filtering useless, irrelevant benchmarks like sysmark and games tested at 1024x768 with a GTX 280 pulling 200fps "handicapping" the 3570k? Oh, my bad, the i5 is only better if you obsess over graphs showing synthetic performance instead of actually using your computer.
Single-threaded versions of multi-threaded benchmarks are worth zero. If a workload can be adequately parallelized, there's no reason at all to factor in silly what-if scenarios when deciding what chip to buy. Legitimate lightly-threaded workloads, on the other hand, are definitely fair game.
If you're pulling 60 fps with a 280 I think it'll be pretty easy to pull 120 with a 480, let alone a 680.

Thank you, sums up my thoughts on it thumb.gif
The Struggle (4k)
(20 items)
 
File Server
(12 items)
 
Lenovo G50-45
(6 items)
 
CPUGraphicsRAMOS
Athlon II X2 250u Nvidia 6150SE  2gb DDR3 1066mhz Windows 10 Home 64 bit 
CPUGraphicsRAMHard Drive
AMD A8-6410 AMD R5 Crucial Ballistix 8GB DDR3L 1866 CAS10 Crucial BX100 250gb 
Optical DriveOS
DVD Windows 10 Home 
  hide details  
Reply
The Struggle (4k)
(20 items)
 
File Server
(12 items)
 
Lenovo G50-45
(6 items)
 
CPUGraphicsRAMOS
Athlon II X2 250u Nvidia 6150SE  2gb DDR3 1066mhz Windows 10 Home 64 bit 
CPUGraphicsRAMHard Drive
AMD A8-6410 AMD R5 Crucial Ballistix 8GB DDR3L 1866 CAS10 Crucial BX100 250gb 
Optical DriveOS
DVD Windows 10 Home 
  hide details  
Reply
post #158 of 216
Quote:
Originally Posted by Usario View Post

Quote:
Originally Posted by Booty Warrior View Post

Quote:
Originally Posted by Usario View Post

If you filter out the poor excuse for a benchmark suite known as sysmark as well as games that are already over 60 fps despite testing with a GTX 280, the 3570k and 8350 have the same number of wins and losses.
Wow... is that it or do you have any more special rules to handicap the 3570k? 60 fps is certainly not the be all end all with 120Hz/3D monitors and of course the higher the average fps (generally speaking) the higher the minimums will be as well.
Quote:
Plus, Anandtech's testing suite is far from the most comprehensive. They'll include single-threaded versions of multi-threaded benchmarks and all the 10 tests that sysmark runs, but they'll for example only use one 3D rendering or one encryption benchmark.
What's the issue with including single thread results along with multi-thread for synthetic benchmarks? It gives a very clear picture of the performance of the architecture, which is ultimately the point of these types of comparisons. AMD's 8 core can pull ahead under heavily-threaded situations, but Intel's ipc lead absolutely crushes them in single threaded applications. That's relevant info to me.

If anything that makes them more comprehensive than most.

Unless you specifically buy a CPU to run Cinebench all day...

OT: I am kind of envious of AMD users having an endless stream of drop-in upgrades available. Maybe steamroller will be the big payoff people have been waiting for.

How is filtering useless, irrelevant benchmarks like sysmark and games tested at 1024x768 with a GTX 280 pulling 200fps "handicapping" the 3570k? Oh, my bad, the i5 is only better if you obsess over graphs showing synthetic performance instead of actually using your computer.

Single-threaded versions of multi-threaded benchmarks are worth zero. If a workload can be adequately parallelized, there's no reason at all to factor in silly what-if scenarios when deciding what chip to buy. Legitimate lightly-threaded workloads, on the other hand, are definitely fair game.

If you're pulling 60 fps with a 280 I think it'll be pretty easy to pull 120 with a 480, let alone a 680.

And before anyone quotes this and says "Yeah, but that shows the performance difference between the CPUs! Intel will be more future proof because it performs better", that logic is very flawed.

1) Even if the i5 3570k continues to perform better (ie. Games don't start heavily using the CPU), it will be different games and different performance differences when it does apply...So there's no use comparing current games to get an idea of future performance.
2) Considering that it's getting harder and harder to make obviously better graphics in games these days (Compare the differences between GTA III and GTA SA to GTA IV and GTA V in trailers, there's a much bigger difference between III and SA, hell, even III and VC than IV and V, this is true of most series') CPU performance will get more important, but games will also make use of more cores thanks to the use of physics calculations and vastly improved AI, I can see the FX-8350 aging a lot better than the i5 3570k if that move starts happening with the PS4 and next Xbox, much like the Q6600 being a better CPU to own than an E8400 today, whereas the E8400 was faster previously.
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i5 3570k @ 4.5Ghz ASRock Z77 Pro3 Powercolor Radeon HD7950 3GB @ 1150/1350 4x4GB G.Skill Ares 2000Mhz CL9 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 250GB Western Digital Black 1TB WD1002FAEX Seagate Barracuda 3TB ST3000DM001 Samsung Spinpoint EcoGreen 2TB 
Optical DriveCoolingCoolingCooling
Pioneer DVR-220LBKS Noctua NH-D14 Scythe Gentle Typhoon 1850rpm Corsair AF140 Quiet Edition 
CoolingOSMonitorMonitor
Arcitc Cooling Acclero Twin Turbo II Arch Linux x86-64, amdgpu BenQ G2220HD BenQ G2020HD 
KeyboardPowerCaseMouse
Ducky Shine III Year of the Snake, Cherry Blue Silverstone Strider Plus 600w CoolerMaster CM690 II Black and White SteelSeries Sensei Professional 
Mouse PadAudioOther
Artisan Hien Mid Japan Black Large ASUS Xonar DX NZXT Sentry Mesh 30w Fan Controller 
  hide details  
Reply
post #159 of 216
Quote:
Originally Posted by Brutuz View Post

I can see the FX-8350 aging a lot better than the i5 3570k if that move starts happening with the PS4 and next Xbox, much like the Q6600 being a better CPU to own than an E8400 today, whereas the E8400 was faster previously.

That was my justification for getting the 8120 (then 8350) coming from a socket 939 dual core. That dual core lasted me till 2011, yet people back then said dual cores were pointless. Not me, I keep my systems for like 5 years lol
The Struggle (4k)
(20 items)
 
File Server
(12 items)
 
Lenovo G50-45
(6 items)
 
CPUGraphicsRAMOS
Athlon II X2 250u Nvidia 6150SE  2gb DDR3 1066mhz Windows 10 Home 64 bit 
CPUGraphicsRAMHard Drive
AMD A8-6410 AMD R5 Crucial Ballistix 8GB DDR3L 1866 CAS10 Crucial BX100 250gb 
Optical DriveOS
DVD Windows 10 Home 
  hide details  
Reply
The Struggle (4k)
(20 items)
 
File Server
(12 items)
 
Lenovo G50-45
(6 items)
 
CPUGraphicsRAMOS
Athlon II X2 250u Nvidia 6150SE  2gb DDR3 1066mhz Windows 10 Home 64 bit 
CPUGraphicsRAMHard Drive
AMD A8-6410 AMD R5 Crucial Ballistix 8GB DDR3L 1866 CAS10 Crucial BX100 250gb 
Optical DriveOS
DVD Windows 10 Home 
  hide details  
Reply
post #160 of 216
Quote:
Originally Posted by Brutuz View Post


Why the hell would you compare a synthetic benchmark when there's plenty of results from real world applications there? I'm not going to get a graphics card because 3DMark said it's better, I'm going to look at real world results..
I'd absolutely agree that real world benchmarks take precedence over synthetics, but that doesn't make synthetics useless either. If nothing else they can show potential strengths and weaknesses depending on the application.

Take Heaven for instance. When it was released it showed the stark contrast in tessellation power between previous gen AMD and Nvidia cards. At the time, no games used tess at that level, so the benchmark was considered largely irrelevant. Then Crysis 2 launched.
Quote:
If you're running a 3D monitor, why are you considering an AMD CPU? You're obviously mainly into gaming, in which case the i5 3570k is the better choice, if you do stuff other than gaming or game at 60Hz then AMD is practically equal to Intel until you hit above the $250 mark, there's just a tonne of people who will read a review, look at the coloured graphs (Sometimes not even noticing it's a lower is better result) and claim AMD is worse when they're competing with Intel pretty well.
laughingsmiley.gif Give people a little credit here. Even if some people completely misinterpret the results, I'd rather have more information available than less when making any hardware-based decision.
Quote:
IPC has zero to do with core count..AMD gets ahead because they have more cores than Intel at the same price point, and if you're doing something that truly does max out your CPU chances are it's going to be using all 8 cores anyway. There's no point in comparing single-threaded performance when you're trying to do a real-world comparison simply because why the hell would I run (For example) POVRay in single-threaded mode when I've got 7 more cores sitting there doing nothing?
They include results like that because the overwhelming majority of software is still unoptimized for using 4+ threads. And in those situations Intel's IPC advantage absolutely matters. With that in mind, a core to core comparison makes quite a bit of sense.

It's not as if they exclude the multi-threaded results anyway, they just include single-thread along side them to give the full picture. I don't see the issue.
Quote:
Originally Posted by Usario View Post

How is filtering useless, irrelevant benchmarks like sysmark and games tested at 1024x768 with a GTX 280 pulling 200fps "handicapping" the 3570k? Oh, my bad, the i5 is only better if you obsess over graphs showing synthetic performance instead of actually using your computer.
You criticize them for their test suite while advocating your own biased choices of what should count and what shouldn't. You literally just said "throw out a bunch of tests where the i5 wins and they're equal!"

The 200+fps results, much like the synthetic benchmarks may not seem practical but they do show potential performance. If you're arguing the 8350 is competitive I would think you'd want to keep them considering they were closer than the more "real world" results from DoW/SC2/WoW.

The i5 was 20% faster in DoW2, 38% faster in WoW and 41% faster in Starcraft 2 (where the 8350 was under 60fps). For someone interested in gaming that is relevant information.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Inquirer] AMD sticks with Socket AM3+ for Steamroller, FM2 to get three years