Overclock.net › Forums › Industry News › Hardware News › [xbitlabs] AMD: We Are On Track With Steamroller Micro-Architecture in 2013.
New Posts  All Forums:Forum Nav:

[xbitlabs] AMD: We Are On Track With Steamroller Micro-Architecture in 2013. - Page 40  

post #391 of 718
Quote:
Originally Posted by Particle View Post

It's worth pointing out that the 30% improvement claim was with regards to what the instruction decoders are capable of decoding. That isn't the same thing as saying each core is going to be able to complete 30% more instructions every second. Not every issue is a matter of the cores being idle when they could be chewing on more instructions.

It's one of the most notable bottlenecks the uarch has.

Besides, AMD stated '30% ops per cycle improvement'. Take it where you take it, it means the whole processor is, in AMD internal simulations, able to process 30% more data per clock cycle.

From the newer improvements, the two that definitely catch my attention are:

1) Branch misprediction penalty lowered approximately a 20%. This means they reduced the number of pipeline stages. Perhaps based upon the fact that Bulldozer's was pretty damn long (mid 30s IIRC?), and the optimal length has been estimated to be 7 stages). A problem in a pipeline that long means the processor wastes a long time running through all the stages again.

2) Independent decoders. No more idling cores when they could be crunching. This might not improve performance all that much in lightly threaded scenarios, but when more than half of the threads are used (and especially in the lower FX4 and FX6 series, as using 2-3 and even 4 cores is becoming mainstream), performance should improve a nice margin, as cores should be able to scale more linearly.
   
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  
   
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  
post #392 of 718
Quote:
Originally Posted by Stay Puft View Post

Don't start with the intel conspiracy crap in this thread.

Its not a conspiracy. I like intel and AMD. Heck I dont even currently own an AMD product, and own 6 Intel machines. Intel literally bribes their way into contracts. Go hit up google, theres been past lawsuits and it still continues. Has any intel product ever completely destroyed AMD's offerings? Even the bulldozer failure vs sandybridge, AMD was still a decent buy for the price (not the best idea, but not garbage). Look at all the other duopolies in the world, show me other duopolies that have such a mismatch of competitor sales ratio.
post #393 of 718
Quote:
Originally Posted by Artikbot View Post

Besides, AMD stated '30% ops per cycle improvement'. Take it where you take it, it means the whole processor is, in AMD internal simulations, able to process 30% more data per clock cycle.

That might be what they mean, but that isn't what the slide says.
post #394 of 718
Quote:
Originally Posted by Particle View Post

That might be what they mean, but that isn't what the slide says.

I directly extracted that from the slides.

   
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  
   
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  
post #395 of 718
Fail wrong thread!
post #396 of 718
Quote:
Originally Posted by Artikbot View Post

Quote:
Originally Posted by Particle View Post

That might be what they mean, but that isn't what the slide says.

I directly extracted that from the slides.


So lets take what they said and cut it in half. 15% increase would be pretty nice
 
Fat HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
2700k Asrock Z77 extreme 4 Sapphire HD 7950 Boost 8 GB gskill ripX 
Hard DriveHard DriveOptical DriveCooling
Samsung F3 WD 500 Blue ASUS 212+ 
OSMonitorMonitorKeyboard
Win7 x64 Prof ASUS AS248H-P LG 21.5 1080p Keys of Boardness 
PowerCaseMouseAudio
XFX 750W XXX Antec 300, fanned out! Death Adder Klipsch Promedia 2.1 
AudioAudio
Xonar Dg Superlux HD 668B 
CPUMotherboardGraphicsRAM
Phenom II 955 ASUS M4A785T-M XFX HD 5670 1 GB Crucial Ballistix 2x2 GB 
Hard DriveCoolingOSPower
WD 500 Green Hyper 212+  Win 7 x64 Prof antec earthwatts 650 
Case
NZXT source 210 
  hide details  
 
Fat HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
2700k Asrock Z77 extreme 4 Sapphire HD 7950 Boost 8 GB gskill ripX 
Hard DriveHard DriveOptical DriveCooling
Samsung F3 WD 500 Blue ASUS 212+ 
OSMonitorMonitorKeyboard
Win7 x64 Prof ASUS AS248H-P LG 21.5 1080p Keys of Boardness 
PowerCaseMouseAudio
XFX 750W XXX Antec 300, fanned out! Death Adder Klipsch Promedia 2.1 
AudioAudio
Xonar Dg Superlux HD 668B 
CPUMotherboardGraphicsRAM
Phenom II 955 ASUS M4A785T-M XFX HD 5670 1 GB Crucial Ballistix 2x2 GB 
Hard DriveCoolingOSPower
WD 500 Green Hyper 212+  Win 7 x64 Prof antec earthwatts 650 
Case
NZXT source 210 
  hide details  
post #397 of 718
Quote:
Originally Posted by Biorganic View Post

So lets take what they said and cut it in half. 15% increase would be pretty nice

Indeed. More realistic than 30 for sure. Steam will actually be a great processor if they can hit the 15 mark.
post #398 of 718
Quote:
Originally Posted by Artikbot View Post

I directly extracted that from the slides.


Me too. We're seeing the same slide but getting two different messages from it. I see that slide which is focused on fetch & decode logic and see the message to the left as applying to what else is on the slide, not to the core as a whole. I'm also not saying that your interpretation isn't what the slide is meant to convey, but I am saying that it isn't what the slide does say in its current form.

I choose to hedge and hope for a 15% improvement in ops per cycle per core.
post #399 of 718
Quote:
Originally Posted by Biorganic View Post

I am not an electrical engineer but here is my partial understanding of the situation. Bulldozer/PD has a bottleneck in the front end decoder units. The front end decode unit feeds 2 integer cores, but cannot feed both simultaneously, it must alternate on a per clock basis. This is one thing reported to be fixed in SR. IIRC there were memory latency or cache latency issues as well.

The decoder issues should affect highly threaded workloads more so than lightly threaded. Whereas a Cache latency issue would affect all operations. However PD does not have as much issue keeping up with intel solutions in heavily threaded apps, so I don't know where the primary problem is.

It is pretty obvious from various benchmarks that BD/PD underperform relative to the competition, they need to increase IPC to compete with Ivy or Haswell (not likely). PD and Ivy will both clock to around the same frequency, so the chip with better per core IPC in many cases will perform better. On certain highly threaded tasks, the BD arch shines, but not everything is so designed.

Also next time could you be less of an A$$ in your tone. And if you understand it all so well, why don't you explain how to fix BD/PD and whats wrong with them. Thanks

You seem to have dodged my question completely. Your reply mimics that of a regurgitated review. You keep saying IPC this and IPC that and I'm asking you specifically which instruction sets is it lacking in execution per cycle? Also, have you even owned a BD/PD cpu or are you just on the anti AMD bandwagon? These review sites need to stop using the term IPC. Or at the very least stop implying that IPC is the sole aspect that determines the performance of a cpu. it's getting ridiculous. First it was all about mhz, then ghz, then IMC, now IPC. What's next?
Edited by computerparts - 4/4/13 at 2:19pm
post #400 of 718
When did I ever mention an instruction set? And I don't know the answer to your question, As far as specific instruction set weaknesses for BD/PD. I'm sure I could google it if I felt the need to...

Single core performance on BD/PD is behind that of SB or Ivy, are you debating this? I don't really see your point, besides offhandedly attacking me for no readily apparent reason.

I have no preference for brand and I would really like AMD to come back in the desktop market. I own all AMD GPUs and still have my trusty phenom II. Where you get that I am anti AMD is beyond me. And No I have not owned a BD or PD chip, I was on the fence about purchasing them, but got the z77 and 2700k at MCenter instead for a ridiculous deal.


Also, instead of berating me and my post, why not enlighten us with your knowledge/opinions on the matter. You know, teach me the way master Yoda, instead of stabbing me with your light saber. (that sounded dirty) smile.gif
Edited by Biorganic - 4/4/13 at 2:28pm
 
Fat HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
2700k Asrock Z77 extreme 4 Sapphire HD 7950 Boost 8 GB gskill ripX 
Hard DriveHard DriveOptical DriveCooling
Samsung F3 WD 500 Blue ASUS 212+ 
OSMonitorMonitorKeyboard
Win7 x64 Prof ASUS AS248H-P LG 21.5 1080p Keys of Boardness 
PowerCaseMouseAudio
XFX 750W XXX Antec 300, fanned out! Death Adder Klipsch Promedia 2.1 
AudioAudio
Xonar Dg Superlux HD 668B 
CPUMotherboardGraphicsRAM
Phenom II 955 ASUS M4A785T-M XFX HD 5670 1 GB Crucial Ballistix 2x2 GB 
Hard DriveCoolingOSPower
WD 500 Green Hyper 212+  Win 7 x64 Prof antec earthwatts 650 
Case
NZXT source 210 
  hide details  
 
Fat HTPC
(9 items)
 
 
CPUMotherboardGraphicsRAM
2700k Asrock Z77 extreme 4 Sapphire HD 7950 Boost 8 GB gskill ripX 
Hard DriveHard DriveOptical DriveCooling
Samsung F3 WD 500 Blue ASUS 212+ 
OSMonitorMonitorKeyboard
Win7 x64 Prof ASUS AS248H-P LG 21.5 1080p Keys of Boardness 
PowerCaseMouseAudio
XFX 750W XXX Antec 300, fanned out! Death Adder Klipsch Promedia 2.1 
AudioAudio
Xonar Dg Superlux HD 668B 
CPUMotherboardGraphicsRAM
Phenom II 955 ASUS M4A785T-M XFX HD 5670 1 GB Crucial Ballistix 2x2 GB 
Hard DriveCoolingOSPower
WD 500 Green Hyper 212+  Win 7 x64 Prof antec earthwatts 650 
Case
NZXT source 210 
  hide details  
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
This thread is locked  
Overclock.net › Forums › Industry News › Hardware News › [xbitlabs] AMD: We Are On Track With Steamroller Micro-Architecture in 2013.