Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [AMDFX] AMD Steamroller IPC Leaked! Cosmology Benchmark!
New Posts  All Forums:Forum Nav:

[AMDFX] AMD Steamroller IPC Leaked! Cosmology Benchmark! - Page 29

post #281 of 745
Quote:
Originally Posted by btupsx View Post

How is it utter nonsense? Those bench results state exactly what I said; look at the speeds of the BD/PD chips, then look at the speeds of the Thuban/Deneb chips. When a BD/PD beats a Thuban/Deneb, it's clocked in the neighborhood of 5 GHz, and the top scoring Thubans/Denebs are in 4.2 GHz territory. The point is that Vishera doesn't leave Thuban in the dust; it meanders in the same neighborhood, albeit a couple of blocks down the street. Both uarchs are good, but Vishera is absolutely not conclusively better.

Most can only get 3.8. Those pushing higher are also pushing temps and voltage. PD wins because it can push higher with less voltage and achieve more performance at the higher clock than it's previous.

People will say Intel is better, they are right, but the problem is that the improvement in their overclock to performance is just as minimal as AMD's. They really have not broke the x86 wall from SB to Ivy. Haswell is basically the same thing as Ivy.
Edited by SpeedyVT - 10/30/13 at 6:06pm
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
post #282 of 745
@azanimefan

What you said is true, but only for multi-threaded workloads. A Vishera clocked at 4.8ghz would leave the Phenom II behind in single-threaded or lightly-threaded workloads (such as PCSX2) which is why the Vishera chips have the higher results on that chart. For people who constantly do heavily-threaded workloads, they wouldn't see a huge benefit going from a Thuban (that was clocked high) to a Vishera. But for people like me who run emu's, I'd be better off with the Vishera.

It basically comes down to what you already have (do you or do you not already have a well-OC'd Thuban) and what you need performance-wise (single or multi-threaded dominant) in the end. They are equal in MT, but ST the Vishera wins.

If Piledriver didn't have that performance penalty for MT workloads due to the shared decoder between two cores in a module, it would be the all-around superior chip. This is why I would like to see AMD release Steamroller-based FX parts -- whether it's on FM2+ or AM3+.
Edited by NaroonGTX - 10/30/13 at 6:05pm
post #283 of 745
Quote:
Originally Posted by NaroonGTX View Post

@azanimefan

What you said is true, but only for multi-threaded workloads. A Vishera clocked at 4.8ghz would leave the Phenom II behind in single-threaded or lightly-threaded workloads (such as PCSX2) which is why the Vishera chips have the higher results on that chart..

I somehow doubt this is the reason.

in every bench that tests single threaded performance thuban is faster then piledriver at the same clocks. I think the problem is you're using pcsx2 as your test bench. For all we know it's more optimized for what piledriver does best. Purhaps the gpu matters with it's rendering (i know it does for me with pcsx2, when i went from a 9800gs to a 7770 i saw a huge jump in framerate in pcsx2) and those old PhII numbers were with older less capable gpus then the newer piledriver numbers.

We don't know enough about that one bench to point to it and say "that represents the comparative strength between those chips"

Thats why numerous benches are used in reviews with different chip structures. some chips work better with some benches. For what you use your piledriver for obviously it's a better chip then a thuban. that said on average thuban was and remains clock for clock a little bit faster then piledriver.
 
Deep-6
(14 items)
 
 
CPUMotherboardGraphicsRAM
Core I5-4690K Asus Maximus VI Gene MSI GAMING 4G GTX 970 gskill sniper 
Hard DriveHard DriveCoolingOS
Samsung 850 evo  Samsung 850 evo Thermalright HR-02 Macho Rev.B Windows 10 Pro 
MonitorKeyboardPowerCase
Acer K242hl Corsair Raptor K30 Seasonic X-650 Fractal Design Arc Mini R2 
MouseAudio
Logitech MX518 Sennheiser - MOMENTUM Over-the-Ear Headphones 
  hide details  
Reply
 
Deep-6
(14 items)
 
 
CPUMotherboardGraphicsRAM
Core I5-4690K Asus Maximus VI Gene MSI GAMING 4G GTX 970 gskill sniper 
Hard DriveHard DriveCoolingOS
Samsung 850 evo  Samsung 850 evo Thermalright HR-02 Macho Rev.B Windows 10 Pro 
MonitorKeyboardPowerCase
Acer K242hl Corsair Raptor K30 Seasonic X-650 Fractal Design Arc Mini R2 
MouseAudio
Logitech MX518 Sennheiser - MOMENTUM Over-the-Ear Headphones 
  hide details  
Reply
post #284 of 745
PCSX2 isn't optimized for AMD at all, in fact it has been optimized for Intel ever since the devs started working on it back around 2001 or so.

Some games in PCSX2 are more GPU-oriented (the game would be heavily utilizing the Vector Units of the PS2 in this case). The purpose of the benchmark is that it uses a very demanding section of a specific game that is known to be extremely hard on the CPU. Plus the benchmark is run in Software mode which means it doesn't use the GPU for rendering. This is why more games get boosts in Hardware mode (which would explain your performance increase from swapping out GPU's, more bandwidth and GPU power would help here but not in Software mode).

Piledriver also carries various instruction sets that PII doesn't in which case it would outperform PII obviously in those cases, and probably in some apps that use instructions that both support as well. For example, in PCSX2 you can set the gsdx graphic plugin to run using SSE2, SSE 4.1, SSSE3 or AVX. Piledriver chips tend to run better using either SSSE3 or AVX, which PII doesn't support.
post #285 of 745
Quote:
Originally Posted by azanimefan View Post

I somehow doubt this is the reason.

in every bench that tests single threaded performance thuban is faster then piledriver at the same clocks. I think the problem is you're using pcsx2 as your test bench. For all we know it's more optimized for what piledriver does best. Purhaps the gpu matters with it's rendering (i know it does for me with pcsx2, when i went from a 9800gs to a 7770 i saw a huge jump in framerate in pcsx2) and those old PhII numbers were with older less capable gpus then the newer piledriver numbers.

We don't know enough about that one bench to point to it and say "that represents the comparative strength between those chips"

Thats why numerous benches are used in reviews with different chip structures. some chips work better with some benches. For what you use your piledriver for obviously it's a better chip then a thuban. that said on average thuban was and remains clock for clock a little bit faster then piledriver.

At the same clocks is the exact problem. If you're judging a processors performance simply by IPC you'd be wrong to do so. The best way to judge it is to find both their clocks where they achieve the same performance and then see how far of improvement they acquire as you push them beyond it volt for volt, clock for clock and that'll show which is best.
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
post #286 of 745
Quote:
Originally Posted by azanimefan View Post

I somehow doubt this is the reason.

in every bench that tests single threaded performance thuban is faster then piledriver at the same clocks. I think the problem is you're using pcsx2 as your test bench. For all we know it's more optimized for what piledriver does best. Purhaps the gpu matters with it's rendering (i know it does for me with pcsx2, when i went from a 9800gs to a 7770 i saw a huge jump in framerate in pcsx2) and those old PhII numbers were with older less capable gpus then the newer piledriver numbers.

We don't know enough about that one bench to point to it and say "that represents the comparative strength between those chips"

Thats why numerous benches are used in reviews with different chip structures. some chips work better with some benches. For what you use your piledriver for obviously it's a better chip then a thuban. that said on average thuban was and remains clock for clock a little bit faster then piledriver.
I think those scenarios are quite limited and dwindling by the day. I mean I loved my 965BE but it isn't anywhere near my FX 8350. As time goes on Thuban is gonna fall further behind. Doesn't make it a bad chip at all. Just like my 3800+ was awesome for the near 10years I used it. But as software changes and drivers/support moves forward older tech will start to look more the part. At the inception of BD/PD and its initial bench results Thuban was still looking stout, but know the number of benches where Thuban is even remotely close is becoming far and few between.
post #287 of 745
PD does not use less power than Thuban. Soooo tired of hearing otherwise, the benches are out there. Anyway, fact remains that for a just above average Phenom, PD is a sidegrade, end of story. This isn't a backhanded compliment to Vishera. Let's try to steer this ship back on topic.
Up in the Air
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus M5a99fx Pro r2.0 PowerColor 7850 2GBD5-2DHE/OC + Arctic Accelero S1 Kingston HyperX 8GB 1866MHz 
Hard DriveHard DriveOptical DriveCooling
WD Blue 1 TB 7200 RPM WD Blue 1 TB 7200 RPM Asus 24x DVD-RW Zalman CNPS14X 
OSOSMonitorPower
Ubuntu 14.04 x64 Windows 8.1 x64 HP w2207 Seasonic SS-660XP2 ATX 
CaseOther
Corsair 540 Air Crucial MX100 256GB 
  hide details  
Reply
Up in the Air
(14 items)
 
  
CPUMotherboardGraphicsRAM
FX-8350 Asus M5a99fx Pro r2.0 PowerColor 7850 2GBD5-2DHE/OC + Arctic Accelero S1 Kingston HyperX 8GB 1866MHz 
Hard DriveHard DriveOptical DriveCooling
WD Blue 1 TB 7200 RPM WD Blue 1 TB 7200 RPM Asus 24x DVD-RW Zalman CNPS14X 
OSOSMonitorPower
Ubuntu 14.04 x64 Windows 8.1 x64 HP w2207 Seasonic SS-660XP2 ATX 
CaseOther
Corsair 540 Air Crucial MX100 256GB 
  hide details  
Reply
post #288 of 745
Sidegrade? It is a full,hefty upgrade everywhere. Be it encoding, rendering,gaming or what have you. Thuban can't compete with PD.
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Asus R9 Nano Gskill TridentZ 
Hard DriveOptical DriveCoolingOS
Pny SSD 240GB Crucial MX100 CM Nepton 280L Win 10 
MonitorPowerCaseMouse
Acer Predator XG270HU Freesync XFX 750W Pro HAF XM Logitech G502 
CPUMotherboardGraphicsCooling
i7 3770 HP Quadro K2000 HP 
OSPowerCaseMouse
Win 7  HP 400W HP CMT RAT 7 
  hide details  
Reply
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Asus R9 Nano Gskill TridentZ 
Hard DriveOptical DriveCoolingOS
Pny SSD 240GB Crucial MX100 CM Nepton 280L Win 10 
MonitorPowerCaseMouse
Acer Predator XG270HU Freesync XFX 750W Pro HAF XM Logitech G502 
CPUMotherboardGraphicsCooling
i7 3770 HP Quadro K2000 HP 
OSPowerCaseMouse
Win 7  HP 400W HP CMT RAT 7 
  hide details  
Reply
post #289 of 745
Power consumption on the desktop is a goddamned joke anyway, there's no point in even bringing it up. This is a website where people have GTX Titans and triple-monitors and OC their CPU's to the moon. I don't even know why people pull out the power consumption card as if it means anything. Phenom II was and is good, but there's no point in acting like it's better than it really is. For people who run heavily-threaded workloads, PD is a definite upgrade if they went to an octocore.
post #290 of 745
Quote:
Originally Posted by azanimefan View Post

I somehow doubt this is the reason.

in every bench that tests single threaded performance thuban is faster then piledriver at the same clocks. I think the problem is you're using pcsx2 as your test bench. For all we know it's more optimized for what piledriver does best. Purhaps the gpu matters with it's rendering (i know it does for me with pcsx2, when i went from a 9800gs to a 7770 i saw a huge jump in framerate in pcsx2) and those old PhII numbers were with older less capable gpus then the newer piledriver numbers.

We don't know enough about that one bench to point to it and say "that represents the comparative strength between those chips"

Thats why numerous benches are used in reviews with different chip structures. some chips work better with some benches. For what you use your piledriver for obviously it's a better chip then a thuban. that said on average thuban was and remains clock for clock a little bit faster then piledriver.

Again, why does clock for clock matter when the chips are designed for different clocks? It's stupid to compare clock for clock. The only valid comparisons are stock clock vs stock clock and max overclock vs max overclock, that's it. Now, it just so happened that SB's max overclock was similar to Vishera's max overclock, that's why IPC played a big role in the comparisons because the clocks were similar. Haswell and Piledriver on the other hand are achieving different max overclocks, and should be compared as thus. Same with Piledriver vs Phenom II.

Again, IPC and clock for clock mean absolutely nothing by themselves.
Millenium Falcon
(24 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 4930k MSI Big Bang Xpower II EVGA GTX 690 Patriot Viper II Sector 7 
Hard DriveHard DriveHard DriveHard Drive
OCZ Deneva 2 Corsair Force 3 Maxtor Western Digital Green 
Optical DriveCoolingCoolingCooling
Samsung BD/DVD-RW Swiftech MCP655 x2 Black Ice GTX 480 Black Ice GTX 280 
CoolingCoolingCoolingCooling
Alphacool Repack Dual D5 Watercool Heatkiller 3.0 Alphacool GTX 690 fullcover Bitspower Big Bang Xpower II fullcover 
OSMonitorKeyboardPower
Windows 8.1 64-bit Professional 3x Dell S2340 Max Keyboard Durandal CoolerMaster V1000 
CaseMouseMouse PadAudio
Azza Genesis 9000B Logitech G700 Roccat Alumic Onkyo HT-S9100THX 
  hide details  
Reply
Millenium Falcon
(24 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 4930k MSI Big Bang Xpower II EVGA GTX 690 Patriot Viper II Sector 7 
Hard DriveHard DriveHard DriveHard Drive
OCZ Deneva 2 Corsair Force 3 Maxtor Western Digital Green 
Optical DriveCoolingCoolingCooling
Samsung BD/DVD-RW Swiftech MCP655 x2 Black Ice GTX 480 Black Ice GTX 280 
CoolingCoolingCoolingCooling
Alphacool Repack Dual D5 Watercool Heatkiller 3.0 Alphacool GTX 690 fullcover Bitspower Big Bang Xpower II fullcover 
OSMonitorKeyboardPower
Windows 8.1 64-bit Professional 3x Dell S2340 Max Keyboard Durandal CoolerMaster V1000 
CaseMouseMouse PadAudio
Azza Genesis 9000B Logitech G700 Roccat Alumic Onkyo HT-S9100THX 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Rumors and Unconfirmed Articles
Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [AMDFX] AMD Steamroller IPC Leaked! Cosmology Benchmark!