Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [Videocardz] Leaked Zen slides - Identical performance to Broadwell-E
New Posts  All Forums:Forum Nav:

[Videocardz] Leaked Zen slides - Identical performance to Broadwell-E - Page 13

post #121 of 132
What were the clock speeds of the broadwell-e 6900k ES compared to retail version?
Intel Build
(13 items)
 
AMD Build
(13 items)
 
 
CPUMotherboardGraphicsRAM
Intel Q9450 at 3.55Ghz Gigabyte GA-X48-DQ6 Sapphire HD 5770 Vapor-X G.SKILL 8GB 1066 DDR2 
Hard DriveOSPowerCase
WD 640GBs 2-Blue 6-Black Windows 7 Ultimate 64-Bit ABS Tagan BZ800 Antec 900 
CPUMotherboardGraphicsRAM
AMD PII X4 720 BE 3.0Ghz BioStar TA790GX 128M Sapphire HD 5770 Crossfire G.SKILL 8GB 1066 DDR2 
Hard DriveOSPowerCase
Twin WD Caviar Blue 640GB Windows 7 Ultimate 64-Bit Antec EA650 650W Antec 300 
  hide details  
Reply
Intel Build
(13 items)
 
AMD Build
(13 items)
 
 
CPUMotherboardGraphicsRAM
Intel Q9450 at 3.55Ghz Gigabyte GA-X48-DQ6 Sapphire HD 5770 Vapor-X G.SKILL 8GB 1066 DDR2 
Hard DriveOSPowerCase
WD 640GBs 2-Blue 6-Black Windows 7 Ultimate 64-Bit ABS Tagan BZ800 Antec 900 
CPUMotherboardGraphicsRAM
AMD PII X4 720 BE 3.0Ghz BioStar TA790GX 128M Sapphire HD 5770 Crossfire G.SKILL 8GB 1066 DDR2 
Hard DriveOSPowerCase
Twin WD Caviar Blue 640GB Windows 7 Ultimate 64-Bit Antec EA650 650W Antec 300 
  hide details  
Reply
post #122 of 132
http://www.notebookcheck.net/Intel-First-Kaby-Lake-Benchmarks-leaked.171693.0.html

1.067x faster than Skylake?

142 vs 133 in cinebench R15 singlethread
Edited by czin125 - 8/21/16 at 10:54am
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6850 evga 680i a1 XFX 5770 11.7_Drivers_Aug15-2011) 4GB DDR2 800 
Hard DriveOSMonitorPower
320 GB SATA Windows 7 Ultimate x86 22inch display 1680x1050 750W Silverstone 
Case
antec 900 
  hide details  
Reply
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
E6850 evga 680i a1 XFX 5770 11.7_Drivers_Aug15-2011) 4GB DDR2 800 
Hard DriveOSMonitorPower
320 GB SATA Windows 7 Ultimate x86 22inch display 1680x1050 750W Silverstone 
Case
antec 900 
  hide details  
Reply
post #123 of 132
Thread Starter 
Quote:
Originally Posted by umeng2002 View Post

Junk leakage, bad yields, etc.
Ah, I see.
post #124 of 132
Quote:
Originally Posted by lolerk52 View Post

The hell is quality of silicon supposed to mean?

It was a really bad way or wording things, but I think the process node will be a big factor in why Zen will not clock really high.

Quote:
Originally Posted by Seraphic View Post

What were the clock speeds of the broadwell-e 6900k ES compared to retail version?

You cant compare Intel's ES chips to AMD's ES chips.
Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
post #125 of 132
Quote:
Originally Posted by Kuivamaa View Post

Thing is that features like large L2 cache (which Zen reportedly has in bigger quantities than BW and SL) also benefit IPC in many ways and for many workloads.

And they'll benefit two threads more than one.
Quote:
Originally Posted by Kuivamaa View Post

Given its width and the fact its L3 is not totally unified (for the 8 core part that is) I tend to believe that at least for blender, it performs on a very similar level to BW both in ST and MT.

Rendering is parallel enough that inter-core bandwidth shouldn't matter in the slightest. Even if a single threaded or dependency limited multi-threaded workload would be impacted by the split LLC/modules, multi-threaded rendering shouldn't be.
Quote:
Originally Posted by Seraphic View Post

What were the clock speeds of the broadwell-e 6900k ES compared to retail version?

The 6900K ES? There were likely hundreds of Broadwell-E engineering samples of widely varying clock potential.

Generally ES aren't directly representative of specific SKUs, just core and/or feature configurations.

The earliest Broadwell-E samples I remember was a 10-core 3GHz part (equivalent to a retail 6950X), but that was barely a month before the official release and were clearly mature samples (same R0 release stepping and manufacture dates no older than the last week of 2015).
Quote:
Originally Posted by EniGma1987 View Post

You cant compare Intel's ES chips to AMD's ES chips.

Especially when Broadwell-E was a derivative of an already very mature process and architecture.
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
post #126 of 132
Quote:
Originally Posted by Blameless View Post

And they'll benefit two threads more than one.
Rendering is parallel enough that inter-core bandwidth shouldn't matter in the slightest. Even if a single threaded or dependency limited multi-threaded workload would be impacted by the split LLC/modules, multi-threaded rendering shouldn't be.

About cache you simply can't state that generally 2 SMT threads are gonna benefit more than one because this depends on OS, application/workload and processor used. As for a single threaded workload , again it depends- for a theoretical ST load that saturates L3 cache could see latency due to Zen's L3 configuration. Definitely doable although not sure in what scenarios is this relevant. At least when benching ST stuff I don't expect to see context switching and such.
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Sapphire RX Vega 64 reference 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 Sapphire RX Vega 64 reference 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 #127 of 132
Quote:
Originally Posted by Kuivamaa View Post

About cache you simply can't state that generally 2 SMT threads are gonna benefit more than one because this depends on OS, application/workload and processor used.

Two threads scheduled to the same core automatically implies increased cache contention over one. If the caches are large enough there will be rapidly diminishing returns from making them larger, but there is evidence that the 256KiB L2 in the various Core ix parts is often not at this point of rapidly diminishing returns. Intel's L2s are very fast, but the size was a compromise when the first Nehalem was being designed and is certainly even more of a compromise now that it's been a decade since that 256K was deemed sufficient.

More cache contention doesn't automatically means that the architecture with the larger L2 will do better, or see more of a benefit from SMT, but it's certainly one feature of Zen that makes me think it could very well scale better with SMT than Intel's x86-64 parts.
Quote:
Originally Posted by Kuivamaa View Post

As for a single threaded workload , again it depends- for a theoretical ST load that saturates L3 cache could see latency due to Zen's L3 configuration. Definitely doable although not sure in what scenarios is this relevant. At least when benching ST stuff I don't expect to see context switching and such.

It's much less likely for a single thread to see tangible improvement from more than 8MiB of L3 than it is from one or two threads to see an improvement from more than 256KiB of L2, but there are still plenty of cases where this has been demonstrated. Any lightly threaded app that runs faster on an HEDT part than a mainstream part, of a given generation, is either benefiting from more memory channels, or a larger cache, because the cores themselves are identical. Games are the most notable examples of relatively lightly threaded apps where large caches can make a difference; I recall Skyrim being a notable example of something that ran better on HEDT parts, all other things being equal.

Those few tasks that can benefit from having more than 8MiB in a fast local cache will likely see more of a penalty on a dual module Zen than on an HEDT Intel part with a single, fast, unified 15-25MiB L3.
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
Primary
(15 items)
 
Secondary
(13 items)
 
In progress
(10 items)
 
CPUMotherboardGraphicsRAM
5820K @ 4.2/3.5GHz core/uncore, 1.175/1.15v Gigabyte X99 SOC Champion (F22n) Gigabyte AORUS GTX 1080 Ti (F3P) @ 2025/1485, 1... 4x4GiB Crucial @ 2667, 12-12-12-28-T1, 1.34v 
Hard DriveHard DriveHard DriveCooling
Plextor M6e 128GB (fw 1.06) M.2 (PCI-E 2.0 2x) 2x Crucial M4 256GB 4x WD Scorpio Black 500GB Noctua NH-D15 
OSMonitorKeyboardPower
Windows 7 Professional x64 SP1 BenQ BL3200PT Filco Majestouch Tenkeyless (MX Brown) Corsair RM1000x 
CaseMouseAudio
Fractal Design Define R4 Logitech G402 Realtek ALC1150 + M-Audio AV40 
CPUMotherboardGraphicsRAM
X5670 @ 4.4/3.2GHz core/uncore, 1.36 vcore, 1.2... Gigabyte X58A-UD5 r2.0 w/FF3mod10 BIOS Sapphire Fury Nitro OC+ @ 1053/500, 1.225vGPU/1... 2x Samsung MV-3V4G3D/US @ 2000, 10-11-11-30-T1,... 
RAMHard DriveHard DriveHard Drive
1x Crucial BLT4G3D1608ET3LX0 @ 2000, 10-11-11-3... OCZ (Toshiba) Trion 150 120GB Hyundai Sapphire 120GB 3x Hitachi Deskstar 7k1000.C 1TB 
CoolingOSPowerCase
Noctua NH-D14 Windows 7 Pro x64 SP1 Antec TP-750 Fractal Design R5 
Audio
ASUS Xonar DS 
CPUMotherboardGraphicsRAM
i7-6800K @ 4.3/3.5GHz core/uncore, 1.36/1.2v ASRock X99 OC Formula (P3.10) GTX 780 (temporary) 4x4GiB Crucial DDR4-2400 @ 11-13-12-28-T2, 1.33v 
Hard DriveHard DriveCoolingOS
Intel 600p 256GB NVMe 2x HGST Travelstar 7k1000 1TB Corsair H55 (temporary) Windows Server 2016 Datacenter 
PowerCase
Seasonic SS-860XP2 Corsair Carbide Air 540 
  hide details  
Reply
post #128 of 132
ffs... I just recently decided to say **** the wait for Zen and go with Broadwell-E. Now I'm going to have to play the waiting game for another month or 2 before I make the decision. Well, the performance will be making the decision... If I can get within 5-10% of the single threaded performance of a skylake with a zen for 70% of the price, I'll go with zen. If Zen is same performance AND same price, I'll probably go with zen as well.

I'm an AMD fanboy, but I'm sitting here with an i5 4690k and GTX 1070 in my computer and it's painful. I just want AMD to compete again =(
post #129 of 132
Quote:
Originally Posted by Blameless View Post

Two threads scheduled to the same core automatically implies increased cache contention over one. If the caches are large enough there will be rapidly diminishing returns from making them larger, but there is evidence that the 256KiB L2 in the various Core ix parts is often not at this point of rapidly diminishing returns. Intel's L2s are very fast, but the size was a compromise when the first Nehalem was being designed and is certainly even more of a compromise now that it's been a decade since that 256K was deemed sufficient.

In the case of two threads scheduled to the same core we are already dealing with heavy workloads, be it many light ones running concurrently or a fully parallel one. Bigger L2 will greatly benefit single threaded workloads on a single core too. We shouldn't understate this. Haswell managed a noticeable ST perf boost over ivy and at the same time better SMT scaling , depending on the workload of course. I am just waiting to see how the particular mix of AGU,ALU,μοp etc works in the case of Zen. In other words to see If all that logic arsenal can be fully leveraged in ST scenarios or if there are certain parts redundant there but designed to boost SMT. My guess? Very similar ratio to what intel sees from Haswell onwards.
Mastodon Ryzen
(12 items)
 
HP Z220
(8 items)
 
 
CPUMotherboardGraphicsRAM
R7 1800X Asus Crosshair VI Hero Sapphire RX Vega 64 reference 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 Sapphire RX Vega 64 reference 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 #130 of 132
Quote:
Originally Posted by EniGma1987 View Post

Not sure why everyone is so completely skeptical. This is identical performance in a blender CPU render test, that means a lot of FP operations. It has been predicted through math extrapolation that Zen should have slightly better FP throughput than Haswell, and Broadwell is the same as Haswell. So given that it has a small technical advantage in this test comparison, Id say that if it only matches the Broadwell-E then that means we are going to see worse hyper-thread scaling than what Intel has.

The fact that those slides/benchmarks are coming straight from AMD is good reason to be skeptical
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Rumors and Unconfirmed Articles
Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [Videocardz] Leaked Zen slides - Identical performance to Broadwell-E