Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

Page 251 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

The Stilt

Golden Member
Dec 5, 2015
1,709
3,057
106
That makes the question of how AMD is going to address the inter-CCX quirks in Naples and ThreadRipper even more intriguing.

Why would the inter-CCX quirks be any different to normal Ryzen CPUs, since it is the same exact die (Zeppelin) in all of them?
The potential inter-die (GMI for TR and GMI / xGMI for Epyc) quirks meanwhile are another thing and there is no information about them at the moment.
 
Last edited:

Shivansps

Diamond Member
Sep 11, 2013
3,851
1,518
136
Today i builded my first Ryzen config for a client and i think i ran intro the VME bug that completely screws my network boot system that uses legacy PXE boot. I think the problem seems to be the pxeboot.n12 file provided by the Windows 10 SDK to chainload PXE boot, is either that or the bootmgr.exe itselft. I finally managed to work around it by using the UEFI PXE boot instead, but that is really a non-easy fix for me, as i always use the legacy PXE for every system, old and new ones.


I going to report the issue to MS but i dont think they will do anything about it.
 

dnavas

Senior member
Feb 25, 2017
355
190
116
In typical benchmarks or real world applications / games.
Zeppelin cannot clock high enough for the bandwidth / latency available at 3200MHz MEMCLK (w/ decent latencies) to become a problem.

I went from bog-standard 2166 to 3200 and noticed no difference in the software I use (video editing). It feels like the inter-ccx latency/"infinity"-bandwidth "issue" is a mite overblown.

If AMD was to release a refresh of Zeppelin on a better process, offering significantly higher Fmax (e.g. 4.5GHz), then that would be another story.

I would dearly love to see 4.5Ghz OCs, but from a competitive point of view, boosting into that region seems a necessary step. Would you notice issues if there were only a few cores active, or is the issue only likely to present itself under full loads?
 

DrMrLordX

Lifer
Apr 27, 2000
21,617
10,824
136
I went from bog-standard 2166 to 3200 and noticed no difference in the software I use (video editing). It feels like the inter-ccx latency/"infinity"-bandwidth "issue" is a mite overblown.

It seems to have the greatest impact on software with a lot of thread movement between CCXs and/or interthread communication between CCXs.
 

imported_jjj

Senior member
Feb 14, 2009
660
430
136
Where / when has she stated that?

At a financial presentation last week i believe, will have to look for it and edit this post.
Ofc it's debatable what one means by final silicon or maybe she misspoke so we'll see.

Edit: i guess others have already provided the answer and linked to a transcript.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,770
3,590
136
Why would the inter-CCX quirks be any different to normal Ryzen CPUs, since it is the same exact die (Zeppelin) in all of them?
The potential inter-die (GMI for TR and GMI / xGMI for Epyc) quirks meanwhile are another thing and there is no information about them at the moment.
I intended to talk about inter-die but my wording was wrong.

Is GMI/xGMI a different hierarchy of IF or something else altogether?
 

tamz_msc

Diamond Member
Jan 5, 2017
3,770
3,590
136
https://www.lowyat.net/2017/132732/...20zc-worlds-first-notebook-powered-ryzen-cpu/

My dreams comes true 8 core notebook.

Asus Ryzen Notebook (Ryzen 7 1700 )
That low res image makes the text on the card illegible. Can't say if this is a Raven Ridge or Ryzen 7 based.

EDIT: That thing is a desktop 1700 with an RX 580. TDP is listed as 65W as usual, but I bet it must be having some form of cTDP. That they were able to fit everything at around 3 kg is impressive.

Now I want this CPU paired with a GeForce MX150 and it will be the ultimate portable workstation. Looking at you Dell.
 
Last edited:

imported_jjj

Senior member
Feb 14, 2009
660
430
136

That's funny considering that Core I9 starts with 10 cores at 999$, nothing to do with AMD's bellow 500$ SKUs.
And ofc AMD has no reason not to cover the 500$ to 999$ band with 12 and 16 cores.
Would be good for them if they have a 1820X at slightly higher clocks at 499$ and/or a SKU with fewer than 12 cores Threadripper bellow 599$.

The 6 cores Skylake X will be interesting to many, if it can OC a bit. Not a great deal when you factor in the entire platform but better than all the other SkylakeX and Kaby Lake X SKUs for the more mainstream users. Mainstream in this context meaning folks that don't need more than 8 cores.
 

Neptunemystic

Junior Member
Mar 1, 2017
3
8
81
I'm not sure if this is the right thread but....there is a new patch for Rise of the Tomb Raider. Ryzen's performance is significantly improved, some users are reporting a 50% increase @ 1080p.

The following changes are included in this patch

- Fix certain DX12 crashes reported by users on the forums.
- Improve DX12 performance across a variety of hardware, in CPU bound situations. Especially performance on AMD Ryzen CPUs can be significantly improved.

While we expect this patch to be an improvement for everyone, if you do have trouble with this patch and prefer to stay on the old version we made a Beta available on Steam, build 767.2, which can be used to switch back to the previous version.

We will keep monitoring for feedback and will release further patches as it seems required. We always welcome your feedback!

http://forums.eu.square-enix.com/showthread.php?t=254796&s=ea21fd6c2d46877c189c6e647153b29d
 

tamz_msc

Diamond Member
Jan 5, 2017
3,770
3,590
136
I'm not sure if this is the right thread but....there is a new patch for Rise of the Tomb Raider. Ryzen's performance is significantly improved, some users are reporting a 50% increase @ 1080p.

The following changes are included in this patch

- Fix certain DX12 crashes reported by users on the forums.
- Improve DX12 performance across a variety of hardware, in CPU bound situations. Especially performance on AMD Ryzen CPUs can be significantly improved.

While we expect this patch to be an improvement for everyone, if you do have trouble with this patch and prefer to stay on the old version we made a Beta available on Steam, build 767.2, which can be used to switch back to the previous version.

We will keep monitoring for feedback and will release further patches as it seems required. We always welcome your feedback!

http://forums.eu.square-enix.com/showthread.php?t=254796&s=ea21fd6c2d46877c189c6e647153b29d
People on reddit are reporting up to 40% gains with 1080Ti and 1700@3.8 at 1080p.
 
  • Like
Reactions: richierich1212

Glo.

Diamond Member
Apr 25, 2015
5,704
4,548
136
I think you misunderstood his wording as leaked information. It's just an opinion.

Either way, BitsAndChips sits in my fake news drawer as w-junk-tech.
It is great, that you read out what I have understood, out of just a link.

I just posted a link to TT of B&C, without any comment on it.
 

DrMrLordX

Lifer
Apr 27, 2000
21,617
10,824
136
https://www.lowyat.net/2017/132732/...20zc-worlds-first-notebook-powered-ryzen-cpu/

My dreams comes true 8 core notebook.

Asus Ryzen Notebook (Ryzen 7 1700 )

Now wait a second. If they can make something outlandish like a R7 1700 notebook, why did they have so much trouble making a halfway-decent A12-9800p notebook when those launched? I mean, really? This notebook breaks all the rules of what "OEMs really want in a notebook thesedays". Try to tell someone to put a lowly Bristol Ridge or Carrizo in a notebook and it's like "it uses too much power wahhhh" good grief.

Yeah yeah different market segments but still.
 

DrMrLordX

Lifer
Apr 27, 2000
21,617
10,824
136
Because Bristol Ridge was a squeak compared to Ryzen's deafening thunder in the CPU market.

Bristol Ridge actually belonged in a notebook, though. As much as I love R7, it doesn't really belong there except in an always-plugged-into-the-wall desktop replacement lappy.
 

Mopetar

Diamond Member
Jan 31, 2011
7,831
5,980
136
Now wait a second. If they can make something outlandish like a R7 1700 notebook, why did they have so much trouble making a halfway-decent A12-9800p notebook when those launched? I mean, really? This notebook breaks all the rules of what "OEMs really want in a notebook thesedays". Try to tell someone to put a lowly Bristol Ridge or Carrizo in a notebook and it's like "it uses too much power wahhhh" good grief.

Yeah yeah different market segments but still.

Completely different animals. The A12 was pointless to put into a notebook. Get a far better Intel chip and match with a discreet GPU for far better performance.

Were getting the R7 1700 in a notebook because you can't get an 8C/16T CPU in a 65W TDP budget any other way (except maybe one of those high-core Atom chips Intel made) and there are people who will pay big bucks for desktop replacements in a portable form. The Ryzen chip is the best for that right now so it gets chosen. The price makes it even sweeter.

No one would have bought that A12 notebook, because there were better alternatives. Right now you can't do better than the R7 1700. Even looking at the leaks for the upcoming Intel chips they won't have anything that can touch the 65W of the 1700, never mind that prices will be higher for no or little performance advantage.
 

The Stilt

Golden Member
Dec 5, 2015
1,709
3,057
106
I intended to talk about inter-die but my wording was wrong.

Is GMI/xGMI a different hierarchy of IF or something else altogether?

GMI and xGMI are basically the same link, but with different type of configuration (width), speed and error tolerance.

GMI is the internal form (inter-die) and has bandwidth of 42.6GB/s per direction at 2666MHz MEMCLK.
xGMI is the external form (inter-node, i.e. EPYC only), has the same bandwidth at the same MEMCLK but different speed, width and error tolerance.