LinuxQuestions.org
Review your favorite Linux distribution.
Home Forums Tutorials Articles Register
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware
User Name
Password
Slackware This Forum is for the discussion of Slackware Linux.

Notices


Reply
  Search this Thread
Old 05-23-2022, 11:41 PM   #16
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316

It's such a pain finding this data, lol.

Gen4
https://en.wikichip.org/wiki/intel/m...itectures/gen4
the microarchitecture for Intel's graphics processing unit utilized by 3000
series GMA. Those GPUs were used in Bearlake and Crestline-based chipsets
(Santa Rosa platform)

Gen5
https://en.wikichip.org/wiki/intel/m...itectures/gen5
the microarchitecture for Intel's graphics processing unit utilized by
GMA 4500 series. Those GPUs were used in Eaglelake and Cantiga-based chipsets
(Montevina platform).

Which suggests it's actually Gen 4?, assuming the wikichip is right.

However, from what I am learning, it sounds like these gpus were not actually builtin to the cpu body, but were separate, which suggests that it's difficult in the earlier phases.

I couldn't link the gen6 with the gma 3600 however anywhere, or with any gen. I also could only find the D2xxx/N2xxx numbers listed as gen6, starting at I think n2400, I could not find anything about d2000 or n2000.

I don't know how intel was doing this, if it was a totally separate chip for gpu at earlier points, they would have just used the one they wanted, and you could judge nothing by the process size, but in later dies, the all in one SoC that is now more common, for a while now, I suspect those are going to be generally similar process nodes, I think anyway, they don't have to be, I just read while researching this that for example only Alder Lake 'performance cores' are the Intel 7 (10nm) node, the slower cores, and maybe gpu, are bigger nodes, but I don't have that data anywhere.

https://en.wikipedia.org/wiki/List_o...ocessing_units

I'll conclude that the specific Atom gpu string you have, which I have more than 10 of, all identical, is not enough data to conclude anything. But it does appear that Atom isn't using the standard intel gpu generations, so your example is good.

Note that I have not found a positive ID, and the string you have is clearly degenerated via something that the oem did to that data string since it does not contain real numbers, just the placeholders.

Intel stuff is a real pain in the butt, lol, no documentation, you hope for finding something, but in the case of that atom string, I could not find anything.

https://www.techpowerup.com/gpu-specs/gma-3600.c1797

this was a 2008 start series, and with your cpu positive match, at 2011-13, that's feasible, particularly for a low cost device like that, using previous generation gpu to save money or maybe it just fit in better.

Note that this also departs the intel naming convention, with the PowerVR SGX545 series, sigh. That's 65 nm.

So far good examples, but I knew intel would be the major issue, because they just are totally jumbled up, and mix engineering with marketing routinely, which is I think a big reason they were losing to amd zen, and had to have an internal shakeup that put an engineer back in charge, which resulted in the alder lake, which is still lagging zen from what I can see, more power hungry, but it is a start at getting back in the game so I'm not concerned for their future, but stuff like this really to me highlights how messy this really is/was internally.

This is why I hesitated trying to add any real info for intel gpus. My hope is to get it 'usually, ideally often, right', whereas with nvidia I can get it to almost always right, same for amd, but because amd doesn't publish their device name + product id lists like nvidia does, I'll miss some for sure here and there.
 
Old 05-24-2022, 12:37 AM   #17
baumei
Member
 
Registered: Feb 2019
Location: USA; North Carolina
Distribution: Slackware 15.0 (replacing 14.2)
Posts: 365

Rep: Reputation: 124Reputation: 124
According to what I read a while back, the Intel GMA 3600 was not designed by Intel, but was licensed from PowerVR. So, this video controller is not like those designed inside Intel...

Consequently, I expect the GMA 3600 will not fit in Intel's Gen3, Gen4, Gen5, &c. naming scheme.
 
Old 05-24-2022, 12:38 AM   #18
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316
I suspect that the entire ID name string is a placeholder for several series, I have a lot of those in my list, all different model product ids. The products IDs are all in a sequence, 16 of them, from obe0 to ebef.

That's the only information I have.

It's possible this is not match-able to any specific series I suspect unless I can get more solid information on it.

The product IDs come between a Haswell and a Crystal Well. There's one other Atom:

Atom Processor Z36xxx/Z37xxx Series Graphics & Display 0f31

which came before the broadwell based gpus.

Haswell is gen7.5, broadwell gen8

Atom/Celeron/Pentium Processor x5-E8000/J3xxx/N3xxx Integrated Graphics Controller
was the gen8 series as far as I could determine.

gen1 > gen4 don't use these full 4 digit numbers.

I don't have any 5 or 5.75 matches. 6 has an Atom Z2760

https://www.techpowerup.com/cpu-specs/atom-z2760.c1437

https://www.techpowerup.com/cpu-spec...ort=generation

The first N2xxx or D2xxx listed there are Cedarview/atom, 2011, which would put it right around gen6

https://www.techpowerup.com/cpu-specs/atom-n2600.c1407

so that's the cpu, cedarview, which means the cpu ID is right

IT doesn't list which gma it is, so it may be right they used an old gpu in this, it could be gen 4, but very difficult to say.

I found a new list of intel gpus:
https://www.techpowerup.com/gpu-spec...cedarview.g383
Intel GPU Architecture History
1998-2002 Generation 1.0
1998 i740
2002-2003 Generation 2.0
2004-2005 Generation 3.0
2005-2006 Generation 3.5
2006-2007 Generation 4.0
2008-2010 PowerVR SGX535
2008 Generation 5.0
2008-2010 PowerVR SGX545
2010 Generation 5.75
2010-2013 Knights
2011 Generation 6.0
2012-2013 Generation 7.0
2013 Generation 7.5
2014-2015 Generation 8.0
2015-2016 Generation 9.0
2016-2020 Generation 9.5
2019-2021 Generation 11.0
2020-2021 Generation 12.1
2021-2022 Generation 12.2
2021-2022 Generation 12.5
2022 Generation 12.7

As you can see, the insanity continues. If it's GMA 3600, it's gen4, but it might be PowerVR SGX545

It looks like the wikichip thing might be wrong, and Arctic Sound is gen12.5, and there's a new one wikichip didn't even list, Alchemist.

Last edited by h2-1; 05-24-2022 at 12:44 AM.
 
1 members found this post helpful.
Old 05-24-2022, 01:08 AM   #19
baumei
Member
 
Registered: Feb 2019
Location: USA; North Carolina
Distribution: Slackware 15.0 (replacing 14.2)
Posts: 365

Rep: Reputation: 124Reputation: 124
Yes, the PowerVR SGX545.

According to Intel's documentation, the N2600 is actually a 64-bit processor.

If I correctly remember the story I read back in 2019 before I bought this little laptop, the reason ASUS set a flag to force the N2600 into 32-bit mode is because Microsoft refused to write a 64-bit driver for this rather rare video controller. I saw there was a 32-bit Linux driver (gma500) for the controller, and I figured it would be good enough for my purposes. I put Slackware 14.2 on the little laptop, and I have been using it every day since (uptime --- 01:59:44 up 29 days, 15:31, 22 users, load average: 0.18, 0.16, 0.18).

One day soon, I intend to find out whether this little laptop will run well with Slackware 15.0.

Last edited by baumei; 05-24-2022 at 01:12 AM.
 
Old 05-24-2022, 01:08 AM   #20
linuxdaddy
Member
 
Registered: May 2022
Location: New Mexico, USA
Distribution: Slackware 15.0 & 64 bit-current, antiX
Posts: 118

Rep: Reputation: 29
You're welcome,

Here is the output from my Intel 865G --gen2 aka Springdale
motherboard. Glad my old hardware is useful.

Code:
$ ./pinxi --gpu -C --zv
CPU:
  Info: model: Intel Pentium 4 bits: 32 arch: Netburst Northwood
    process: Intel 130nm family: 0xF (15) model-id: 2 stepping: 9
    microcode: 0x2E
  Topology: cpus: 1x cores: 1 smt: <unsupported> cache: 512 KiB note: check
  Speed (MHz): 2400 min/max: N/A core: 1: 2400 bogomips: 4799
  Flags: ht pae sse sse2
  Vulnerabilities: <filter>
Graphics:
  Device-1: Intel 82865G Integrated Graphics
    vendor: ASUSTeK P5P800-MX Mainboard driver: i915 v: kernel arch: Gen2
    process: Intel 130nm ports: active: VGA-1 empty: none bus-ID: 00:02.0
    chip-ID: 8086:2572 class-ID: 0300
  Display: x11 server: X.Org v: 1.20.11 driver: X: loaded: intel gpu: i915
    display-ID: :0.0 screens: 1
  Screen-1: 0 s-res: 1280x1024 s-dpi: 96 s-size: 338x270mm (13.31x10.63")
    s-diag: 433mm (17.03")
  Monitor-1: VGA-1 mapped: VGA1 model: Samsung SyncMaster
    serial: HVFA103063 built: 2006 res: 1280x1024 hz: 60 dpi: 86 gamma: 1.2
    size: 380x300mm (14.96x11.81") diag: 482mm (19") ratio: 5:4 modes:
    max: 1280x1024 min: 720x400
  OpenGL: renderer: Mesa DRI Intel 865G x86/MMX/SSE2 v: 1.3 Mesa 20.3.5
    direct render: Yes
 
Old 05-24-2022, 02:36 AM   #21
aus9
LQ 5k Club
 
Registered: Oct 2003
Location: Western Australia
Distribution: Icewm
Posts: 5,842

Rep: Reputation: Disabled
Quote:
Your minimalist graphics Device-1 name, AMD, is a pretty good example SNIP
I swapped to Xorg Debian for the moment, and seeing last poster command I get

Code:
pinxi --gpu -C --zv  # on Debian Xorg Icewm
SNIP
Device-1: AMD Picasso/Raven 2 [Radeon Vega Series / Radeon Mobile Series]
I hope to reboot and try that command in sway shortly and edit this post
reboot to TC sway
Code:
pinxi --gpu -C --zv  # On TC sway 
  Device-1: AMD driver: amdgpu v: kernel arch: GCN 5 code: Vega
Does below help?
Code:
cat /proc/cpuinfo
processor	: 0
vendor_id	: AuthenticAMD
cpu family	: 23
model		: 24
model name	: AMD Ryzen 3 3200G with Radeon Vega Graphics
stepping	: 1
SNIP

Last edited by aus9; 05-24-2022 at 02:44 AM.
 
Old 05-24-2022, 03:07 PM   #22
baumei
Member
 
Registered: Feb 2019
Location: USA; North Carolina
Distribution: Slackware 15.0 (replacing 14.2)
Posts: 365

Rep: Reputation: 124Reputation: 124
Hi "h2-1",

According to my understanding, this computer has the "Intel GMA 4500HD" integrated into the chipset.

Code:
user1@darkstar:~$ ./pinxi -V
pinxi 3.3.16-10 (2022-05-23)
user1@darkstar:~$
user1@darkstar:~$ ./pinxi --gpu -C --zv
CPU:
  Info: model: Intel Core2 Quad Q8300 bits: 64 type: MCP arch: Yorkfield
    process: Intel 45nm family: 6 model-id: 0x17 (23) stepping: 0xA (10)
    microcode: 0xA07
  Topology: cpus: 1x cores: 4 smt: <unsupported> cache: L1: 256 KiB
    desc: d-4x32 KiB; i-4x32 KiB L2: 4 MiB desc: 2x2 MiB
  Speed (MHz): avg: 2000 min/max: 2003/2499 scaling: driver: acpi-cpufreq
    governor: ondemand cores: 1: 2000 2: 2000 3: 2000 4: 2000 bogomips: 19998
  Flags: ht lm nx pae sse sse2 sse3 sse4_1 ssse3
  Vulnerabilities: <filter>
Graphics:
  Device-1: Intel 4 Series Integrated Graphics vendor: Dell 4 driver: i915
    v: kernel arch: Gen4 process: Intel 65n ports: active: VGA-1
    empty: DP-1,HDMI-A-1 bus-ID: 00:02.0 chip-ID: 8086:2e22 class-ID: 0300
  Display: server: X.Org v: 1.20.14 with: Xwayland v: 21.1.4
    compositor: xfwm v: 4.16.1 driver: X: loaded: modesetting unloaded: vesa
    alternate: fbdev gpu: i915 display-ID: :0.0 screens: 1
  Screen-1: 0 s-res: 1440x900 s-dpi: 96 s-size: 381x238mm (15.00x9.37")
    s-diag: 449mm (17.69")
  Monitor-1: VGA-1 model: Dell SE198WFP serial: GM7787941TES built: 2007
    res: 1440x900 hz: 60 dpi: 90 gamma: 1.2 size: 408x255mm (16.06x10.04")
    diag: 481mm (18.9") ratio: 16:10 modes: max: 1440x900 min: 720x400
  OpenGL: renderer: Mesa DRI Intel G45/G43 (ELK) v: 2.1 Mesa 21.3.5
    direct render: Yes
 
Old 05-24-2022, 04:07 PM   #23
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316
baumei, great example, another problematic one, I had assumed that '4 series' was generation 4, hahah, this is why I was getting no gen5 results, as you correctly located, it's part of the gma 4500 series.

That explains the missing gen5. Now all the gens except 10, which was cancelled, and the newest ones, have data, which is what I'd expect given how far back the pci ids datasets I'm using go.

I've added 'built:' as well, I opted for the general time period. I have to still tweak a bit the current ones, as with the nvidia active series, I will end the years with + to indicate it's still active, but I have to still confirm which ones are active.

Code:
pinxi --gpu -C --zv
CPU:
  Info: model: AMD Ryzen 5 2600 bits: 64 type: MT MCP arch: Zen+ gen: 2
    process: GF 12nm family: 0x17 (23) model-id: 8 stepping: 2
    microcode: 0x800820D
  Topology: cpus: 1x cores: 6 tpc: 2 threads: 12 smt: enabled cache:
    L1: 576 KiB desc: d-6x32 KiB; i-6x64 KiB L2: 3 MiB desc: 6x512 KiB
    L3: 16 MiB desc: 2x8 MiB
  Speed (MHz): avg: 2329 high: 3762 min/max: 1550/3400 boost: enabled
    scaling: driver: acpi-cpufreq governor: ondemand cores: 1: 1523 2: 3528
    3: 1580 4: 3412 5: 3762 6: 1824 7: 1762 8: 3359 9: 1413 10: 1560 11: 1544
    12: 2690 bogomips: 81598
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
  Vulnerabilities: <filter>
Graphics:
  Device-1: AMD Cedar [Radeon HD 5000/6000/7350/8350 Series] vendor: XFX Pine
    driver: radeon v: kernel alternate: amdgpu arch: TeraScale 2
    code: Evergreen process: TSMC 32-40nm built: 2009-15 pcie: gen: 1
    speed: 2.5 GT/s lanes: 16 link-max: gen: 2 speed: 5 GT/s ports:
    active: DVI-I-1,VGA-1 empty: HDMI-A-1 bus-ID: 0a:00.0 chip-ID: 1002:68f9
    class-ID: 0300
The old Intels you've found really helped, I didn't have any data on some of these at all so had to guess. There's 2 more I can't match:

82852/855GM Integrated Graphics Device 3582
82854 GMCH Integrated Graphics Device 358e

These are very early, from what I can find:
gen 1: 827xx/8281x
gen 2: 82865
gen 3: 8291xx
gen 3.5: 8294x
gen 4: 8296x

I'm going with gen2 based on this:
https://www.techpowerup.com/gpu-spec...ort=generation

which lists i852GM Graphics Montara as gen 2. I'll assume i850 series is same as 8285x series ids.

This covers most of the intel pci ids I have, there's a few embedded devices left but in any practical sense, no inxi user will ever be running those so not worth worrying about.

Last edited by h2-1; 05-24-2022 at 04:13 PM.
 
1 members found this post helpful.
Old 05-24-2022, 04:21 PM   #24
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316
aus9, I suspect tinycore simply strips out some data from the kernel to get the kernel smaller. But nevertherless it is a good example of why product IDs are much better to ID stuff than strings, all I need is 1 correct match on a pci id and all devices no matter what name they might have are going to be correctly IDed.

linuxdaddy your gen2 is a good sample, confirms ID I think.

The intel is now way better, but of course, way more granular, than before.

I have a slight confidence in most of the results, there's still a few that might be off, but I'm also limited by the data in the pci dbs I'm using, so far everything here has been in the dbs, so that's encouraging, though they did not include any cpu info of course so it was hard to really confirm guesses.

I believe the gen5 '4 series' that I had taken to be referring to gen 4, was actually referring to the gma 4xxx series.

thanks for digging up this old hardware, it's helping lock down the intels.

I think most old amd will work ok unless I made a mistake, and the nvidia's I know are roughly correct.

Last edited by h2-1; 05-24-2022 at 05:09 PM.
 
Old 05-24-2022, 06:04 PM   #25
linuxdaddy
Member
 
Registered: May 2022
Location: New Mexico, USA
Distribution: Slackware 15.0 & 64 bit-current, antiX
Posts: 118

Rep: Reputation: 29
I have one of these I can run later tonight and post. I also have other 845 -865 desktop
boards. and 1 945 board also.

Code:
82852/855GM Integrated Graphics Device 3582

Last edited by linuxdaddy; 05-24-2022 at 06:07 PM.
 
Old 05-24-2022, 06:20 PM   #26
linuxdaddy
Member
 
Registered: May 2022
Location: New Mexico, USA
Distribution: Slackware 15.0 & 64 bit-current, antiX
Posts: 118

Rep: Reputation: 29
Also the Intel 865 chipset was the last to offer AGP slot on the motherboard.
The 915-945 had the first PCI-express slot for graphics.
 
Old 05-24-2022, 06:33 PM   #27
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316
Sounds good, the goal here is to lock down as much as possible the old chip ids from amd/intel, so there's less to maintain going forwards.

I've been chipping away at the amd gpu/apu chip ids too, so those have most ids matched now.

I don't have these matched:
Code:
Atom Processor E6xx Integrated Graphics Controller	4108
Atom Processor E6xx Integrated Graphics Controller	4109
Atom Processor E6xx Integrated Graphics Controller	410a
Atom Processor E6xx Integrated Graphics Controller	410b
Atom Processor E6xx Integrated Graphics Controller	410c
Atom Processor E6xx Integrated Graphics Controller	410d
Atom Processor E6xx Integrated Graphics Controller	410e
Atom Processor E6xx Integrated Graphics Controller	410f
Atom Processor E6xx Integrated Graphics Controller	8182
Atom Processor Z36xxx/Z37xxx Series Graphics & Display	0f31
E7221 Integrated Graphics Controller	258a
Graphics Media Accelerator 500 Graphics	2e5b
HD Graphics	1606
LE80578 Graphics Processor Unit	5002
Moorestown Graphics and Video	4100
UHD Graphics	46a1
Xeon E3-1200 Processor Family Integrated Graphics Controller	010a
Xeon E3-1200 v3 Processor Integrated Graphics Controller	040a
Xeon E3-1200 v3 Processor Integrated Graphics Controller	041a
I think the E6xx is maybe PowerVR SGX535 but the E means embedded device so it doesn't really matter.

I think the xeon v3 are gen 8 because Xeon E3-1200 v2/3rd Gen are gen 7. But I don't know for sure so I'll leave them unhandled for now. That's very few unhandled device ids overall for Intel compared to what it started with.

It's also possible xeon 3rd gen isn't even referring to the same as a core 3rd gen is referring to, you simply can't know with intel stuff.

Last edited by h2-1; 05-24-2022 at 06:35 PM.
 
Old 05-24-2022, 06:49 PM   #28
Gerard Lally
Senior Member
 
Registered: Sep 2009
Location: Leinster, IE
Distribution: Slackware, NetBSD
Posts: 2,177

Rep: Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761Reputation: 1761
Quote:
Originally Posted by h2-1 View Post
As usual:
Code:
# if you have pinxi installed:
pinxi -U

# if you don't:
wget -O /usr/local/bin/pinxi && chmod +x /usr/local/bin/pinxi
There must be a URL missing here, surely?
 
Old 05-24-2022, 07:19 PM   #29
h2-1
Member
 
Registered: Mar 2018
Distribution: Debian Testing
Posts: 556

Original Poster
Rep: Reputation: 316Reputation: 316Reputation: 316Reputation: 316
Gerard Lally, ooops, red face, sorry. smxi.org/pinxi == shortcut to github download url.
 
1 members found this post helpful.
Old 05-24-2022, 07:24 PM   #30
baumei
Member
 
Registered: Feb 2019
Location: USA; North Carolina
Distribution: Slackware 15.0 (replacing 14.2)
Posts: 365

Rep: Reputation: 124Reputation: 124
Hi "h2-1",

According to what I dug up about five years ago, this little laptop has an "Intel GMA950" integrated video controller, with two ports; and the chip-set is "Intel 945GME Express".
Code:
user1@darkstar:$ ./pinxi -V
pinxi 3.3.16-11 (2022-05-24)
user1@darkstar:$
user1@darkstar:$ ./pinxi --gpu -C --zv
CPU:
  Info: model: Intel Atom N270 bits: 32 type: MT arch: Bonnell
    process: Intel 45nm family: 6 model-id: 0x1C (28) stepping: 2
    microcode: 0x218
  Topology: cpus: 1x cores: 1 tpc: 2 threads: 2 smt: N/A cache: L1: 56 KiB
    desc: d-1x24 KiB; i-1x32 KiB L2: 512 KiB desc: 1x512 KiB
  Speed (MHz): avg: 1200 high: 1600 min/max: 800/1600 scaling:
    driver: acpi-cpufreq governor: ondemand cores: 1: 800 2: 1600
    bogomips: 6719
  Flags: ht nx pae sse sse2 sse3 ssse3
  Vulnerabilities: <filter>
Graphics:
  Device-1: Intel Mobile 945GSE Express Integrated Graphics vendor: ASUSTeK
    driver: N/A alternate: i915 arch: Gen3.5 process: Intel 90nm built: 2005-06
    bus-ID: 00:02.0 chip-ID: 8086:27ae class-ID: 0300
  Display: server: X.org v: 1.18.3 driver: X: loaded: vesa
    unloaded: modesetting alternate: fbdev,intel gpu: N/A tty: 80x27
  Message: GL data unavailable in console. Try -G --display
 
  


Reply



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
Testers for inxi/pinxi redone -C CPU logic... huge internal changes h2-1 Slackware 353 02-24-2022 08:51 PM
Huge inxi/pinxi upgrade, new features, Logical volumes, raid rewrite, beta testers? h2-1 Slackware 12 12-17-2020 05:04 PM

LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware

All times are GMT -5. The time now is 03:36 PM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration