| Page 2 of 31 |
|
|
 Posted: Wed, 8th Jun 2016 23:44 Post subject: |
|
 |
| Janz wrote: | pga socket and 1.331 pins confirmed
still dont know why they use pga, soo many dumb people out there to destroy their pins ^^ |
*raises hand* 
I had sold my 965 system and when I took the cpu out to go byebye to get my 8350 when I had it. Was putting the 965 in the old system and had a violent sneak attack sneeze and dropped the 965 on the case pin side. Pins all bent to hell. Managed to spend another hour with 2 credit cards straightening them enough to get back in the socket. |
|
| Back to top |
|
|
JBeckman
VIP Member
Posts: 34976
Location: Sweden
|
| Posted: Wed, 20th Jul 2016 08:19 Post subject: |
|
|
http://www.guru3d.com/news_story/ams_zen_engineering_sample_specs_leaked.html
( http://forums.guru3d.com/showthread.php?t=408845 )
| Quote: |
New details on AMD Zen based CPUs have surfaced, the information is based on engineering samples that AMD already has been circulating. Blogger Dresdenboy AMD has confirmed the information making the reliability of the information plausible.
The engineering samples currently are set at revision A0. The user who spread the details talks about four chips with 4, 8, 24 and 32 of cores. The first two SKUs would be for the AM4-socket, while the last two were intended for servers.
4 variants of ES Zen are available at the moment:
AM4 8 cores with 95W TDP
AM4 4 cores with 65W TDP
SP3 24 cores with 150W TDP
SP3 32 cores with 180W TDP
The two AM4 chips are quad-core and octa-core with 8 and 16 threads. The quad-core would be get 2 MB L2 cache and 8MB L3 cache, while the octa-core would get double that amount. Both engineering samples currently run a clock speed of 2.8 GHz, with a maximum boost up to 3.2 GHz. The TDP of the two would be 65 watts for the chip with four cores and 95 watts for the octa-core. In idle the clock speed can throttle down back to 550 MHz with an amzing power consumption 2.5 and 5 watts idle power.
For servers there is the SP3 platform. The leaker has details on a 24-core and 32-core chip. The boost clock speed is at the 24-core 2.75 GHz and the 32-core 2.9 GHz. The idle-clock rate is 400 MHz here with even lower. The TDP of the two is 150 and 180 watts respectively. These chips have two 64 MB L3 cache, with the 24-core 12 MB of L2 cache and get the 32-core 16MB. This suggests that every Zen-core inherits 512KB of L2 cache, while each set of four cores gets 8MB L3 cache. This corresponds to a previously surfaced die-shot of a core-Zen.
The information has been confirmed by Dresdenboy, he has a good track record, but it remains a rumor/leak and as such a grain of salt is something you should always take this info with.
|
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 18th Aug 2016 15:15 Post subject: |
|
|
http://techreport.com/review/30540/amd-gives-us-our-first-real-moment-of-zen
| Quote: | Nitty-gritty details aside, the real question on everybody's mind is whether AMD met the 40% IPC improvement goal that it's publicly committed to over the past few months. To make the point that it has, AMD put a Summit Ridge engineering sample running at 3GHz up against an eight-core, sixteen-thread Core i7-6900K artificially limited to the same 3GHz speed. AMD ran the same Blender 3D rendering workload on both chips at the same time. Watch the video above for a sense of how Summit Ridge stacks up to Broadwell-E.
While it's worth remembering that this is only one data point, the Zen chip kept pace with or slightly beat the Broadwell-E CPU in that test. If that performance level holds across a range of workloads, AMD appears to have made some of the large strides it needs to make toward closing the performance gap with Intel CPUs. Tantalizingly, AMD says the 3GHz figure isn't the final clock speed it expects production Zen chips to top out at, either. Final clock speeds for Zen, along with TDP figures and pricing, are still under wraps, but we should learn more as we draw closer to the Summit Ridge launch in the first quarter of 2017. |
ASRock B450 Pro4 || Ryzen 7 3700X || Patriot Viper 16Gb DDR4-3600 || ASUS GeForce RTX 3080 TUF GAMING || XPG GAMMIX S11 Pro 1Tb || Xiaomi Mi Surface Display 34" (3440x1440)
|
|
| Back to top |
|
|
thudo
Posts: 6309
Location: Mellonville North, Canada
|
| Posted: Thu, 18th Aug 2016 15:24 Post subject: |
|
|
Ewww.. competition! 
MSI GT72S 6QF Dominator Pro S 29th Anniversary Intel i7 6820HK @ 4.0Ghz, 32GB DDR4-2133 RAM, 2x256GB Raid0 Toshiba NVMe 2.5 inch PCIe SSD, Nvidia Geforce GTX 980 OC'ed 200+ Core / 200+ Mem, 17.3 inch LG IPS HD Display @ 75Hz, Intel 7265AC Wifi, Windows 10 Pro BIOS version: .112 EC Firmware version: .105
Current Broadband speed record: 329.1 Mb/sec down // 21.73 Mb/sec up
http://www.dslreports.com/speedtest/3933292.png
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 18th Aug 2016 15:25 Post subject: |
|
|
I really hope AMD get their shit together, we need competition in the GPU and CPU markets.
"I think Call of Duty resonates because it's believable and relatable," Sledgehammer Games cofounder Michael Condrey says.
Believable and relatable...Yep, sounds like Call of Duty
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 18th Aug 2016 15:41 Post subject: |
|
|
I really hope they will show more benchmarks, that blender test is pure crap, i want to see single thread performance in Cinebench compared to Haswell and upwards.
|
|
| Back to top |
|
|
Werelds
Special Little Man
Posts: 15098
Location: 0100111001001100
|
| Posted: Thu, 18th Aug 2016 16:30 Post subject: |
|
|
Why is it crap? It's not as if AMD have ever done well in Blender before:
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 18th Aug 2016 22:45 Post subject: |
|
|
Last edited by paxsali on Thu, 4th Jul 2024 21:25; edited 1 time in total
|
|
| Back to top |
|
|
couleur
[Moderator] Janitor
Posts: 14330
|
| Posted: Fri, 19th Aug 2016 00:12 Post subject: |
|
|
Is it going to be the first affordable octacore or isnt it. I want to need new hardware!
"Enlightenment is man's emergence from his self-imposed nonage. Nonage is the inability to use one's own understanding without another's guidance. This nonage is self-imposed if its cause lies not in lack of understanding but in indecision and lack of courage to use one's own mind without another's guidance. Dare to know! (Sapere aude.) "Have the courage to use your own understanding," is therefore the motto of the enlightenment."
|
|
| Back to top |
|
|
|
|
| Posted: Fri, 19th Aug 2016 01:47 Post subject: I have left. |
|
|
|
|
|
| Back to top |
|
|
thudo
Posts: 6309
Location: Mellonville North, Canada
|
| Posted: Fri, 19th Aug 2016 01:59 Post subject: |
|
|
Would love to see it do some damage to Intel but, like the distant past now, I really doubt it. We need competition back in this area.
MSI GT72S 6QF Dominator Pro S 29th Anniversary Intel i7 6820HK @ 4.0Ghz, 32GB DDR4-2133 RAM, 2x256GB Raid0 Toshiba NVMe 2.5 inch PCIe SSD, Nvidia Geforce GTX 980 OC'ed 200+ Core / 200+ Mem, 17.3 inch LG IPS HD Display @ 75Hz, Intel 7265AC Wifi, Windows 10 Pro BIOS version: .112 EC Firmware version: .105
Current Broadband speed record: 329.1 Mb/sec down // 21.73 Mb/sec up
http://www.dslreports.com/speedtest/3933292.png
|
|
| Back to top |
|
|
Werelds
Special Little Man
Posts: 15098
Location: 0100111001001100
|
| Posted: Fri, 19th Aug 2016 11:11 Post subject: |
|
|
| Shoshomiga wrote: | Looks like it might be competitive with intel but draw a lot more power from what they posted
Makes sense since its amd |
.......what?
The 8C/16T part they used here will be a 95W part (the 4C/8T version will be a 65W part - at least that's what AMD have said so far) and they pitted it against a 140W Broadwell-E part. I'm not sure what you read that "they" "posted" that would indicate it consumes more power.
Take this slide:
Reduced power and at the same time higher IPC. How does that equate to it using more power than the 140W Broadwell-E? |
|
| Back to top |
|
|
|
|
| Posted: Fri, 19th Aug 2016 12:14 Post subject: |
|
|
| Breezer_ wrote: | | , i want to see single thread performance in Cinebench compared to Haswell and upwards. |
Same,i need to know if these chips do good in Arma 
| paxsali wrote: |
Now, I don't know what hardware costs in Poland, I guess it's cheaper because everything is stolen from Germany and resold... |
|
|
| Back to top |
|
|
|
|
| Posted: Fri, 19th Aug 2016 18:38 Post subject: I have left. |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Fri, 19th Aug 2016 20:12 Post subject: |
|
|
Last edited by paxsali on Thu, 4th Jul 2024 21:25; edited 1 time in total
|
|
| Back to top |
|
|
Werelds
Special Little Man
Posts: 15098
Location: 0100111001001100
|
| Posted: Fri, 19th Aug 2016 20:43 Post subject: |
|
|
| Shoshomiga wrote: | | I was looking at their 4 core part vs intel mainstream 4 cores |
Again...what? 
We haven't seen anything of that yet.
| Quote: | | The 8 core part is supposedly the same with double the cores, but it only takes 46% more power, there is gotta be some catch to that and we know how amd's "8 cores" were in the past |
You clearly haven't read into how any of this works.
AMD's "modules" were not why the last few AMD CPU generations were so crap. That was because the architecture was crap. And just the first slide on TR alone shows that this thing is absolutely nothing like Piledriver and Excavator
That modular design is still here I think. Each Zen CPU will have multiple dies that have multiple cores; for Summit Ridge (desktop) that'll be 4 cores per die, for the Opterons it'll be 8 cores per die (and each core can do SMT). Each core gets its own L1 + L2 cache, more than each core in Broadwell-E gets. L3 cache is shared across the die (so between all 4 or 8 cores) and the size of it is on par with Skylake. At least this is how I understand the slides we've seen so far.
As for the power, they save power because the dies do not include the memory controller, south bridge and what not. Those things use more power than you might think. The memory controller is the same on the 65W and 95W parts as far as I can tell, whereas Intel changes pretty much everything for every class of CPUs. AMD seem to just bake in a bunch more ALUs rather than redesigning the whole chip.
I'm not sure what you based your conclusion on, if you have a link that shows something about the 4-core parts I'd love to see it (although I think the 8C parts are where it's at for gamers soon). Considering the preliminary TDP for this part is meant to be 95W and it's matching a 140W part from Intel (even if it doesn't hit that 95W target, that's still 45W of headroom before they decide the final TDP), I'm cautiously optimistic about AMD returning to compete with Intel. Seems Jim Keller has done it again.
And if the 65W does the same thing, that basically means it's an i7-6700 (non-K, K is 91W) but with larger caches.
Obviously need more benchmarks, but the fact that they completely caught up in Blender is a very good sign. Blender is multithreaded and does benefit from multiple cores, but what it does on each core still requires good IPC. It still benefits massively from higher clockspeeds. Just look at the 4790K vs 4770K; that's a 500 MHz speed or 14.3% bump in clockspeed and it translates almost 1:1 in performance in Blender. |
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Thu, 1st Sep 2016 23:25 Post subject: |
|
|
Last edited by paxsali on Thu, 4th Jul 2024 21:24; edited 1 time in total
|
|
| Back to top |
|
|
russ80
Posts: 4679
Location: Romania
|
|
| Back to top |
|
|
|
|
| Posted: Fri, 2nd Sep 2016 09:04 Post subject: I have left. |
|
|
|
|
|
| Back to top |
|
|
Frant
King's Bounty
Posts: 24636
Location: Your Mom
|
|
| Back to top |
|
|
|
|
| Posted: Wed, 7th Sep 2016 16:29 Post subject: |
|
|
Last edited by paxsali on Thu, 4th Jul 2024 21:24; edited 1 time in total
|
|
| Back to top |
|
|
JBeckman
VIP Member
Posts: 34976
Location: Sweden
|
| Posted: Fri, 25th Nov 2016 11:15 Post subject: |
|
|
http://www.guru3d.com/news_story/amd_zen_octoacore_summit_ridge_engineering_sample_spotted_at_3_23_5_ghz.html
| Quote: |
They are coming, a new Zen octacore Summit Ridge engineering sample has been spotted. The engineering samples are internal test samples for qualification purposes. The new model that has been spotted was running 3.2 GHz on all eight cores and a boost towards 3500 MHz.
This is exciting news as these values are as fast or even faster opposed to what Intel offers, their 8-core 5960X and 10-core Core i7 6950X processor run a base clock of 3.0 GHz and a binned turbo towards 3.5 GHz. The first one costs 999 USD, the 6950X 1300 USD. If AMD can keep their IPC performance up-to snuff or at least close to Intel with prices in the rumored 300~400 USD range, then AMD might have struck gold here.
The information on this latest leak is based upon a user called DresdenBoy who has shared more reliable information in the past. The engineering sample processor spotted is tagged at product code number 1D3201A2M88F3_35/32_N which indeed indicates the clock frequencies and number of cores.
1 - Unknown sample type
D - Desktop
32 - 3200 MHz Clock (don't know whether base or peak - given then inclusion of 35/32 later on, we could speculate 3.2 base, 3.5 boost)
01 - Single socket ?
A2 - Unknown, assumed reference to socket type
M - Not sure. Tradition would have it mean 638 pins. We know that can't be right.
8 - 8 Cores
8 - Unknown, traditionally ref to cache size
F3 - Micro-architecture & Revision
Earlier on hardware partner MAXSUN also kind of confirmed the clock frequency range for AMD's next processors as they are quoting 3.15-3.30 GHz base clocks with 3.5 GHz boost clocks. On top of that they claim (highly speculative of course) the processors can be clocked to 4.2 GHz fairly easily with conventional cooling and up to 5 GHz with LN2. In the posted screenshots they mention that the top-end SKU can compete with the Intel 6850K.
[]
The initial “Zen” CPU core will come to market first in an 8-core, 16-thread system-on-chip for desktops (=Summit Ridge). The "Summit Ridge" Zen family will feature a unified AM4 socket with its GPU-equipped "Bristol Ridge" APU counterparts, and feature DDR4 support and a an expected 95W TDP. We expect each Zen core will have four integer units, two address generation units and four floating point units, and the decoder can decode four instructions per clock cycle. L1 data cache size is 32 KiB and L2 cache size 512 KiB per core. 2 CCUs = 2x8MB (L3) + 8x512KB (L2) = 20MB.
Much like the competition AMD would be selling Summit Ridge in three segments, low-end, mid-range and high-end much like Intel's Core i3, i5 and i7 series. When you look at the slide it states SR7, SR5 and SR3, matching up with Summit Ridge (RS) and thus a performance segment denominator. We thing that the SR5 series will not have any form of hyper threading and/or perhaps be the 4-core models.
[]
When you look at the left side of the slide you'll notice stuff like RMB1500. This relates to the price, this is renminbi (which may also be used for the yuan) is CNY. The currency is often abbreviated RMB, or indicated by the yuan sign ¥, now 1500 Chinese Yuan is roughly €200,- for likely the SR5 range.
There very likely will be an Sr7 model that remains unlocked released. It is expected that ZEN Summit Ridge based 8-core processors will launch in week 2 of January.
|
|
|
| Back to top |
|
|
|
|
| Posted: Tue, 13th Dec 2016 21:07 Post subject: |
|
|
|
|
|
| Back to top |
|
|
couleur
[Moderator] Janitor
Posts: 14330
|
|
| Back to top |
|
|
|
|
| Posted: Tue, 13th Dec 2016 21:36 Post subject: |
|
|
some pictures, with an embargo watermark posted somewhere before the embargo ends (which it does in 25 minutes). wouldnt take that too serious (even if those shit comes directly from amd, or lets say especially when it comes from amd -> remember the radeon fury benchmarks?)
edit: they are real, but amd hasnt shown off anything important. just showed a ryzen cpu @3,4ghz encoding some stuff 2 secs faster than an i7 6900K (base clock 3,2ghz,dunno if they had activated the boost though...). and they showed the game performance with battlefield1 with a titan sli system together with the zen cpu. but guess what? they havent shown the fps counter... that my dear friends are not good facts, no balls to show the actual fps? that can only mean it sucks against casual "cheap" i7 skywell system
|
|
| Back to top |
|
|
Przepraszam
VIP Member
Posts: 14491
Location: Poland. New York.
|
| Posted: Tue, 13th Dec 2016 22:34 Post subject: |
|
|
|
|
|
| Back to top |
|
|
|
|
| Posted: Tue, 13th Dec 2016 22:37 Post subject: |
|
|
Last edited by Interinactive on Mon, 4th Oct 2021 10:13; edited 1 time in total
|
|
| Back to top |
|
|
couleur
[Moderator] Janitor
Posts: 14330
|
|
| Back to top |
|
|
|
|
|
| Back to top |
|
|
| Page 2 of 31 |
All times are GMT + 1 Hour |
