PC Perspective Podcast #443 – 03/30/17
Join us for Thermoelectric Coolers, Tiny PSUs, Lots o' Storage, some trips down nostaglia lane, and more!
You can subscribe to us through iTunes and you can still access it directly through the RSS page HERE.
The URL for the podcast is: https://pcper.com/podcast – Share with your friends!
- iTunes – Subscribe to the podcast directly through the iTunes Store (audio only)
- Google Play – Subscribe to our audio podcast directly through Google Play!
- RSS – Subscribe through your regular RSS reader (audio only)
- MP3 – Direct download link to the MP3 file
Hosts: Ryan Shrout, Jeremy Hellstrom, Josh Walrath, Allyn Malventano
Program length: 1:34:48
Podcast topics of discussion:
-
Week in Review:
-
News items of interest:
-
1:16:50 AIDA64 Version 5.90 Released
-
1:18:09 Move Over Nintendo PowerGlove
-
Hardware/Software Picks of the Week
-
Ryan:
-
Jeremy: Planescape: Torment EE obviously
-
-
Closing/outro
Optane needs a driver. Fail.
Optane needs a driver. Fail.
No caching tech that passes
No caching tech that passes cached IO's through a different interface can do so without a driver or some other form of software / OS support.
I thought that there is some
I thought that there is some new feature to build asymmetric RAID 1 available only in 2-series chip-set.
Intel says 2xx chip-set is requirement. The truth is that software is CRIPPLED and will not run on anything else.
Ryzen optimizations incoming
Ryzen optimizations incoming and hopefully more than AOTS will see some improvment. But what about that AM4 motherboard supply situation, are the Taichi motherboards still in short supply. When will there be a major AM4 motherboard roundup.
and
“Update: 03/30/2017 at 9am CT: AMD has released the CodeXL Power Profiler as an Open Source tool for GPU debugging and GPU/CPU profiling, so AMD does have a tool for optimizations.”(1)
(1)
“AMD Ryzen Performance Update Released For Ashes of the Singularity”
http://www.legitreviews.com/amd-ryzen-performance-update-released-ashes-singularity_193137
From: “AMD Ryzen™ Community
From: “AMD Ryzen™ Community Update #2”
“Let’s talk BIOS updates
Finally, we wanted to share with you our most recent work on the AMD Generic Encapsulated Software Architecture for AMD Ryzen™ processors. We call it the AGESA™ for short.
As a brief primer, the AGESA is responsible for initializing AMD x86-64 processors during boot time, acting as something of a “nucleus” for the BIOS updates you receive for your motherboard. Motherboard vendors take the baseline capabilities of our AGESA releases and build on that infrastructure to create the files you download and flash.
We will soon be distributing AGESA point release 1.0.0.4 to our motherboard partners. We expect BIOSes based on this AGESA to start hitting the public in early April, though specific dates will depend on the schedules and QA practices of your motherboard vendor.
BIOSes based on this new code will have four important improvements for you
1.We have reduced DRAM latency by approximately 6ns. This can result in higher performance for latency-sensitive applications.
2.We resolved a condition where an unusual FMA3 code sequence could cause a system hang.
3.We resolved the “overclock sleep bug” where an incorrect CPU frequency could be reported after resuming from S3 sleep.
4.AMD Ryzen™ Master no longer requires the High-Precision Event Timer (HPET).
We will continue to update you on future AGESA releases when they’re complete, and we’re already working hard to bring you a May release that focuses on overclocked DDR4 memory.” (1)
(1)
“AMD Ryzen™ Community Update #2”
https://community.amd.com/community/gaming/blog/2017/03/30/amd-ryzen-community-update-2
It looks like Intel is going
It looks like Intel is going modular SOC also(1), with HBM2/GPU die on an MCM or Interposer for a New Kaby Lake series. So AMD must be doing APU’s on an Interposer with HBM2 and a larger Vega die and some Ryzen core complex/es. So remember even one HBM2 die with 4/8 GB of memory can act like a last level cache above a larger pool of regular DDR4 DRAM in a single/dual Channel DIMM/S and the integrated graphics will never be starved for the necessary bandwidth.
That AMD graphics IP part of the headline is debatable and it not related to actually using AMD’s current GPU IP, it’s more related to some basic IP that is already used in Intel’s graphics and Intel just needs to license that basic GPU IP from either AMD or Nvidia for Intel’s in-house graphics to remain legal. Remember that it’s hard to build a GPU from scratch like Intel did without having to license some IP from Nvidia(who Intel currently licenses from) or AMD(Who Intel may license from) just to keep Intel’s in house GPU design legal.
(1)
” “Rumored Intel Kaby Lake-G Series: Modular, Multi-Die, HBM 2, AMD Graphics IP?” ”
https://www.techpowerup.com/232067/rumored-intel-kaby-lake-g-series-modular-multi-die-hbm-2-amd-graphics-ip
It Looks like Apple is going
It Looks like Apple is going all in-house with its SOC designs(1) but will Apple still have to license some basic GPU IP from Imagination Technologies, or Nvidia(?), Or AMD(?) so Apple’s in-house GPU designs can remain legal. It looks like a messy divorce between Apple and Imagination Technologies.
(1)
“Apple will wind down its use of PowerVR IP in future SoCs”
“http://techreport.com/news/31681/apple-will-wind-down-its-use-of-powervr-ip-in-future-socs”
AnandTech really has a good
AnandTech really has a good analysis of this story and It would be great to Know Just what GPU Maker, Imagination Technologies, AMD, or Nvidia/Others, has the most important GPU IP that others are licensing. Intel designed its own In-House GPU but had to license some critical basic GPU IP from Nvidia to allow Intel to be able to legally sell its SOCs that make use of some of that critical basic GPU IP from Nvidia. And now Intel wants to License that same critical basic GPU IP from AMD, and AMD and Nvidia have a lot of shared critical basic GPU IP that is relatively the same.
So what very basic critical GPU IP will Apple still have to license from Imagination Technologies, even if Apple designs its own fully in-house GPU designs. Can Apple go to say AMD or Nvidia and get that very basic critical GPU IP licensing from another source than Imagination Technologies.
It’s impossible for anyone to design a clean sheet GPU without stepping on some GPU maker’s pre-existing IP and Imagination Technologies, AMD, and Nvidia, have a large chunk of the patents that that make up that critical basic GPU IP that many must license.
Great discussion of the
Great discussion of the scalability of Zen’s design vs Intel’s Ring Bus which can not be scaled beyond a certain point. Intel will be going to a mesh network. So AMD had its reasons(more scalable) for that 2 CCX unit “Zeppelin” modular die. Lot’s of interesting things about what Intel may do in the future. The discussing of GCN by Kanter is important and what Vega will bring, for sure Kanter is the is the expert Kampman not so much. (1)
(1)
“A moment of Zen with David Kanter: The TR Podcast 190
Digging into the whys of Ryzen”
http://techreport.com/review/31693/a-moment-of-zen-with-david-kanter-the-tr-podcast-190
P.S. Both Kanter and Kampman
P.S. Both Kanter and Kampman are still using HyperThreading(TM) instead of SMT, even when talking about Zen’s SMT. SMT(simultaneous multithreading) is the genaric computer sciences term for what Intel calls/has Branded Hyperthreading(TM)/Intel’s version of SMT.