On 2024-12-02 18:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

This is new commercial blah blah blah.

Traditionally, the "turbo" button meant normal speed. The other setting
was actually "slow", and it was used for games and other software
designed for slower speed.

This happened around 1985.


Go figure what they mean now.


--
Cheers, Carlos.

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On 2024-12-02 21:57, micky wrote:

In alt.comp.os.windows-10, on Mon, 2 Dec 2024 21:54:14 +0100, "Carlos
E.R." <robin_listas@es.invalid> wrote:

On 2024-12-02 18:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

This is new commercial blah blah blah.

Traditionally, the "turbo" button meant normal speed. The other setting
was actually "slow", and it was used for games and other software
designed for slower speed.

I just apologized for getting the answer from google, but I didn't get
your answer! And I know you by now and I'm sure you're right.

:-) Thanks

I remember two games on my first computer, which was an Amstrad
PC1512DD. It was around 3 or 4 times faster than the original IBM PC.
They were called "The ancient art of war" and "The ancient art of war at
sea". They used the CPU speed to time the action. On computers by 1990
the game played way too fast to be be playable. We had to use the Turbo
button to slow the computer, or run a small TSR (terminate and stay
resident program) that would waste time and make the computer slow
enough to play again :-D

https://en.wikipedia.org/wiki/The_Ancient_Art_of_War

Google has screenshots.

This happened around 1985.


Go figure what they mean now.

LOL

I posted a follow-up question, if you have time.

--
Cheers, Carlos.

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On 02/12/2024 17:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

Turbo is nowadays something that happens automatically when needed.

Yes, especially in a laptop the processor will probably not be able to
operate with all cores turbo-ing for long before it gets too hot and has
to slow itself down.

Usually on multi-core CPUs the cores turbo independently up and down to
any of many frequencies between the minimum and max turbo.

Processors typically can't run all their cores at the max turbo
frequency quoted in the specification. They may only be able to turbo
one or maybe two cores at a time to that max frequency and maybe only
for a short time.

Does it happen automatically, and then automatically revert to the lower speed later.

Yes.

Or do I have to do something, and then later undo it if I want out of
turbo?

No.

--
Brian Gregory (in England).

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: https://www.Brian-Gregory.me.uk/ (3:633/280.2@fidonet)

Carlos E.R. wrote:

On 2024-12-02 18:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

This is new commercial blah blah blah.

Traditionally, the "turbo" button meant normal speed. The other setting
was actually "slow", and it was used for games and other software
designed for slower speed.

Exactly.

This happened around 1985.

Go figure what they mean now.

If the device has lots of RGB lighting and is a "gaming" system, then
turbo might mean something (to somebody).


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

Brian Gregory wrote:

On 02/12/2024 17:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

Turbo is nowadays something that happens automatically when needed.

Yes, especially in a laptop the processor will probably not be able to operate with all cores turbo-ing for long before it gets too hot and has
to slow itself down.

Usually on multi-core CPUs the cores turbo independently up and down to
any of many frequencies between the minimum and max turbo.

Processors typically can't run all their cores at the max turbo
frequency quoted in the specification. They may only be able to turbo
one or maybe two cores at a time to that max frequency and maybe only
for a short time.

Does it happen automatically, and then automatically revert to the lower
speed later.

Yes.

Or do I have to do something, and then later undo it if I want out of
turbo?

No.

Exactly. Generally, you may be able to limit, configure, or alter some
of the cpu speed control characteristics in bios settings. These days
you might see terms like "speedstep" or other verbiage. "turbo" became meaningless many years ago.

"turbo" was mostly applicable to 80386 and 80406 processors back in the
old days. Most people pushed the "turbo" button and left it that way
forever.

Even further back in the 8080, Z80, and 6502, 6800 days. There was no
turbo. Everything always ran full balls-out speed. Whatever frequency
the quartz crystal frequency was; Generally at 1 Mhz or so.


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Mon, 12/2/2024 1:02 PM, micky wrote:

I should have said https://www.amazon.com/Business-6-Cores-i5-12450H-Fingerprint-Keyboard/dp/B0DLT2B36C/ref=sr_1_6
that the Rysen 5-6600H has a base rate of 2.9 and boost is 4.5.
--- I've heard that AMD cpus are rated higher than Intel of same speed
would be???

Cores 6 (12)
L3 cache 16 MB
Core config 1 x 6

GPU Are CPU and GPU in one package? Was that always true?
Model 660M
Clock 1.9 Ghz
Config 284:24:8 6 CUS
GFLOPS 1459.2
TDP 15-28W
Base 45W, Max. Turbo 95 W,
Released January 2022, yes 3 years ago but it's supposedly on sale, for
$500, allegedly was 900!

https://www.amazon.com/Business-6-Cores-i5-12450H-Fingerprint-Keyboard/dp/B0DLT2B36C/ref=sr_1_6

In alt.comp.os.windows-10, on Mon, 02 Dec 2024 12:48:07 -0500, micky <NONONOmisc07@fmguy.com> wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

Does it happen automatically, and then automatically revert to the lower
speed later.

Or do I have to do something, and then later undo it if I want out of
turbo?

Know anything about Nimo computers? A family business founded in
Cincinatti or New Castle, Delaware, depending on where you look. :-)

Turbo is automatic. It's demand based.

If you install AMD Ryzen Master software
and you don't use the overclocking controls (pointless),
you can watch the closed loop feedback system.

1) Frequency limit (no more than 5.000 GHz on the 5950X)
2) Voltage limit (maybe 1.41V or so, just a bad guess)
3) Power Limit (95W on some chunk of VCore circuit)
4) Temperature Limit (applies to poorly cooled items, like laptops/tablets)

The concept of closed loop, was copied from the video card (GPU)
companies.

On desktop boards, there can be overrides for the closed loop
controls. As a kind of joke, the Intel power register goes up
to 4096 Watts (a 12 bit register). This is otherwise known as
"infinity", as the regulator has coded in it, some limits
that protect it from burnout. The power supply does not
have 4096 watts to give you.

On an Intel desktop, you can draw 228W for 28 seconds or 56 seconds.
The time is a kind of "tau" value. I seem to be able to run 5GHz on one
core, all day long here on my desktop. And again, some BIOS are coded
to disable "tau" and make it infinity.

Summary: It's complicated, it's automatic, it protects the hardware, sometimes has overrides.

*******

Yes, a CPU package can have a GPU die.

+-----------+ +-----------------+
| | | GPU die |
| CPU L1/L2 | | (system memory) |
| | +-----------------+
+-----------+
+----------------+
| L3 SRAM |
| I/O die |
| DRAM interface |
+----------------+

Your proposed 6600H is a custom, with just one silicon die
and all the items placed on the single die. This is the best
way to make silicon, from the customer perspective.

The L3 cache isn't a big win, because it's only three times
faster than DRAM. The L1 and L2 are much faster.

The CPU has some communications delays. The connections
between dies, there may need to be a lot of wires in
silicon, to keep the speed up.

The speed of the GPU cores, is all over the place. And the
6600H has RDNA2 cores, not the same as the others.

These are my AMD AM4/DDR4 processors.

CPU iGPU Video card alternate (iGPU if present, disabled)
5600G Junk room. Vega 7
5700G What I'm typing on. Vega 8 (GTX 1650 video card at the moment, address decoder prob in hardware)
5950X FiveFans machine None (GTX 1050 video card at the moment)

The GPU on your laptop, is likely a little stronger than the Vega examples.

The Techpowerup site, has hardware information.

This is the low end processor I've got.

https://www.techpowerup.com/cpu-specs/ryzen-5-5600g.c2471

It appears the 6600H is a custom thing, with an RDNA2 GPU and
yet another set of cores. Not the same cores as the iGPU
in the 7000 series CPU chips.

https://www.techpowerup.com/cpu-specs/ryzen-5-6600h.c2525

*******

This is your proposed laptop. The built-in iGPU is pretty good.

https://www.videocardbenchmark.net/gpu.php?gpu=Radeon+660M+Ryzen+5+6600H&id=4988

G3D 6286 G2D 795

It trounces the 2 core iGPU in the 7000 series. There is some variation from one 7000 series processor to the next, as the clock rate is likely capped
so the iGPU does not use too much of the power envelope.

https://www.videocardbenchmark.net/gpu.php?gpu=Radeon+Ryzen+7+7700X+8-Core&id=4651

G3D 2127 G2D 586

The iGPU which is disabled in my daily driver computer. VEGA 8

https://www.videocardbenchmark.net/gpu.php?gpu=Ryzen+7+5700G+with+Radeon+Graphics&id=4405

G3D 2795 G2D 796

The iGPU in my lowest end 5600G machine. VEGA 7. Losing one core did not hurt much.

https://www.videocardbenchmark.net/gpu.php?gpu=Ryzen+5+5600G+with+Radeon+Graphics&id=4406

G3D 2582 G2D 764

Notice how the 2D acceleration doesn't really vary all that much.

The replacement for my VEGA 8 is this thing (any card that costs $200 here...) .

https://www.videocardbenchmark.net/gpu.php?gpu=GeForce+GTX+1650&id=4078

G3D 7872 G2D 571

Your laptop then, is getting an integrated GPU that is almost as
powerful as a piss-weak 1650 :-) Which means the slow Firefox graphics
stack, will go a little faster than last week (ANGLE-D3D to WebGL translation). Note that, while the "G3D" is a gaming benchmark, actual applications
can be all over the place. The benchmark numbers are only an approximate
thing. Like, I might get "7000 worth" doing one thing, and "5000 worth"
doing something else. The purpose of the benchmark, is just a "class" indicator.

This is the strongest GPU, strong enough to burn its own power connector. ~$2000
Well worth the price, to open the box and find the power connector burned :-)

https://www.videocardbenchmark.net/gpu.php?gpu=GeForce+RTX+4090&id=4606

G3D 38579 G2D 1295

Paul





--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Mon, 12/2/2024 3:54 PM, Carlos E.R. wrote:

On 2024-12-02 18:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

This is new commercial blah blah blah.

Traditionally, the "turbo" button meant normal speed. The other setting was actually "slow", and it was used for games and other software designed for slower speed.

This happened around 1985.

Go figure what they mean now.

https://www.notebookcheck.net/AMD-Ryzen-5-6600H-Processor-Benchmarks-and-Specs.589894.0.html

They are clocked at

3.3 (guaranteed base clock) to
4.5 GHz (Turbo)

That means, on all-cores (7-ZIP compression), it
should maintain 3.3GHz.

If you run SuperPI, it runs on a single core,
and it should be able to give 4.5GHz. The CPU
should run at roughly half-power, when on turbo
of a single core.

A well designed BIOS, will have a setting to turn
Turbo off. Then the machine will vary from 0 Hz on
a core, to 3.3GHz.

When running 7-ZIP on all cores, it can manage
a bit more than 3.3GHz on all cores. It will
depend on how good a cooler they fitted, and
what thermal limit they dialed into the BIOS,
as to how much headroom there will be for fun.
Maybe your 7ZIP run is at 3.9GHz (until the
CPU gets too hot perhaps).

Some of this, you can see here.

https://www.anandtech.com/show/14873/reaching-for-turbo-aligning-perception-with-amds-frequency-metrics-/3

Paul


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Mon, 12/2/2024 3:55 PM, micky wrote:

In alt.comp.os.windows-10, on Mon, 02 Dec 2024 12:48:07 -0500, micky <NONONOmisc07@fmguy.com> wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Embarrassed to say I was able to get the answer to this from google.
(and I should have foreseen that. I apologize.). They are not the
same.

One laptop I'm looking at has a base or 1GHz, but turbo at 4.4GHz. Is it
a big weakness that it "idles" at such a low speed. Does it mean it can
only be at 4.4GHz for short periods of time. Would I be much better off
if I found one that was, say, 2.5GHz or 3 GHz, with boost of 4.4GHz?

Currently, my cpu is often +95%, often at 100%. This is because of Firefox,-- nothing else is over 8% or usually even over 2%. Firefox is
using 38% even when I'm not consciously doing anything. Like now,
afaik, I'm not doing anything now, but it's 38%. I presume loaded tabs
are running things in background, video clips or something, but that
will continue to be the case with the new computer too. Will it go up
to turbo 4.4hz because of how busy it is, and can it stay there all the
time?

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

Does it happen automatically, and then automatically revert to the lower
speed later.

Or do I have to do something, and then later undo it if I want out of
turbo?

Know anything about Nimo computers? A family business founded in
Cincinatti or New Castle, Delaware, depending on where you look. :-)

The problem with small builders, is they can be "too sharp at business"
and abuse the customer. One US maker for example, insisted there
were no user serviceable parts inside the box. If you did a RAM upgrade
with them, the warranty was voided as soon as you added RAM. This is
illegal.

A device with a base clock of 1.1GHz and a turbo of 4.4GHz, this
implies that the device is severely thermally limited, it attempts
to burn your lap, it has no fan, it's too thin for a fan. No,
you don't want one of those, unless you like "molasses slow"
general operation. It may whomp up to 4.4Ghz for a couple seconds,
and then plonk down to 1.1GHz for the remainder of the operation.

The problem, is the background loading of a modern PC. There
is too much garbage in the PC, to ever see a decent turbo
on a 2 watt tablet.

You want a better balance, a 35W or 45W processor, a cooling
system, a fan. Then your base clock rises a bit, there is
turbo, but, you don't care if the turbo isn't always running
at a super-high level. The thing gives enough performance,
the Windows Defender is happy, the Search Index is chewing
on stuff, and maybe the scroll bar works on Firefox :-)

Paul


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Mon, 12/2/2024 6:35 PM, Hank Rogers wrote:

Brian Gregory wrote:

On 02/12/2024 17:48, micky wrote:

looking to buy a new laptop and a lot give the speed as "Up to nnnghz"
and they are referring to turbo.

Does turbo = overclocking, which I thought presented heat or other
problems for cpus?

Turbo is nowadays something that happens automatically when needed.

Yes, especially in a laptop the processor will probably not be able to operate with all cores turbo-ing for long before it gets too hot and has to slow itself down.

Usually on multi-core CPUs the cores turbo independently up and down to any of many frequencies between the minimum and max turbo.

Processors typically can't run all their cores at the max turbo frequency quoted in the specification. They may only be able to turbo one or maybe two cores at a time to that max frequency and maybe only for a short time.

Does it happen automatically, and then automatically revert to the lower >>> speed later.

Yes.

Or do I have to do something, and then later undo it if I want out of
turbo?

No.

Exactly. Generally, you may be able to limit, configure, or alter some of the cpu speed
control characteristics in bios settings. These days you might see terms like "speedstep"
or other verbiage. "turbo" became meaningless many years ago.

"turbo" was mostly applicable to 80386 and 80406 processors back in the old days. Most
people pushed the "turbo" button and left it that way forever.

Even further back in the 8080, Z80, and 6502, 6800 days. There was no turbo. Everything
always ran full balls-out speed. Whatever frequency the quartz crystal frequency was;
Generally at 1 Mhz or so.

There are write-ups available.

https://www.anandtech.com/show/14873/reaching-for-turbo-aligning-perception-with-amds-frequency-metrics-/3

Today, the XFR behavior is missing from the device, but the other
aspects of the graph are fairly realistic of the closed-loop control
system of modern equipment.

https://images.anandtech.com/doci/14873/AMD%20Ryzen%207%20Press%20Deck-10_575px.jpg

The CPU has a Preferred Core, which is the core most likely to achieve the turbo speed on the box. The CPU may not necessarily be able to turbo on
each individual core, to the same amount, but at least one core is capable
of hitting the number on the tin.

What's not shown in the graph, is what happens when there are two CPU silicon die.
There is an eight core CCX on each die. In that case, there is a fairly
weird behavior, where the preferred CCX seems to load to 100%, and the
second CCX is a bit "slack". There are also things like this seen on the
X3D chip, where the extra L3 cache chip is laid over one CCX and not the
other, and the machine tries to "protect" the one with the L3 on top of it.

AMD processors do have some unexplained behavior. When people asked
some perfectly reasonable questions, Keller replied that the benchmark
used was defective, and he would not engage with them. I have noticed,
when doing a number-of-cores versus benchmark test, that there is a
stairstep behavior in the performance curve, that should not be there.
and this is part of SMT. The second half of SMT seems to be missing
the promised performance boost. If you have a 6C 12T processor,
you see six stairsteps in the plot, rather than 12 steps indicating
the "hyperthreading" is adding performance. The AMD also has a
weirdness, between four threads and five threads. There is a larger step
there, which suggests that some of your runs might be better if you
set -j 4 kind of thing. While there is a general description of the frequency/power/voltage, there is still some details that don't fit well.

Paul



--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Mon, 12/2/2024 6:35 PM, Hank Rogers wrote:

Exactly. Generally, you may be able to limit, configure, or alter some of the
cpu speed control characteristics in bios settings. These days you might see terms
like "speedstep" or other verbiage. "turbo" became meaningless many years ago.

"turbo" was mostly applicable to 80386 and 80406 processors back in the old days.> Most people pushed the "turbo" button and left it that way forever.

Even further back in the 8080, Z80, and 6502, 6800 days. There was no turbo. Everything always ran full balls-out speed. Whatever frequency the quartz crystal
frequency was; Generally at 1 Mhz or so.

Turbo is not meaningless. It's fully specified. And controls for it
are in the desktop motherboard BIOS screen.

https://www.intel.com/content/www/us/en/products/sku/236773/intel-core-i9-processor-14900k-36m-cache-up-to-6-00-ghz/specifications.html

Total Cores 24
# of Performance-cores 8
# of Efficient-cores 16
Total Threads 32

Max Turbo Frequency 6 GHz <===
Intel Thermal Velocity Boost Frequency 6 GHz
Intel Turbo Boost Max Technology 3.0 Frequency 5.8 GHz
Performance-core Max Turbo Frequency 5.6 GHz
Efficient-core Max Turbo Frequency 4.4 GHz
Performance-core Base Frequency 3.2 GHz <===
Efficient-core Base Frequency 2.4 GHz
Cache 36 MB Intel Smart Cache
Total L2 Cache 32 MB
Processor Base Power 125 W
Maximum Turbo Power 253 W

Turbo type, is Type 3, Closed Loop Control. Base is 3.2GHz on a P Core.
Whereas Turbo on one core, is 6 GHz, and the power rises because
VCore has been pushed up. And the Microcode feature that was
pushing the voltage up (at the instruction level),
was damaging the processor, so a new Microcode corrects that.

For that processor, you don't buy used ones, or refurbished
or reconditioned machines with that in it. You buy new stock,
and flash the BIOS immediately with a version known to hold
the new Microcode.

Each Intel flavor of turbo, would be mentioned or documented
on the Anandtech site. While the staff are all gone from there
now, the site lives on and the old articles are available.
Future Inc, shut it down functionally (no new articles).

There are multiple controls. There is Turbo as the "above normal"
feature. There is Intel SpeedStep, which is 1.2Ghz or Base, and
the CPU reduces frequency when there is no compute load.

But the BIOS also has C6 and C12 C-state control, and one
of those, shuts off a core entirely. I can see that in
Ryzen Master screen at idle. Only one core remains running.
That scheme is more than SpeedStep, in that it was proposed
to turn off cores years ago, but it took a while to find
a way to do it. It takes time to bring a core out of a low
power state.

[Picture]

https://i.postimg.cc/LXtp15jR/AMD-Ryzen-Master-machine-state.gif

To run the cores all at the same fixed frequency:

1) Turn off Turbo.
2) Turn off Enhanced Intel Speed Step (EIST).
3) Turn off all C-states (I had to do this on the Test Machine once, to get fixed speed)
4) set Windows schema to "High Performance"

Now, the above processor would be running at 3.2 base on the P-Cores.
You can examine Core 0 speed, by using CPU-Z from cpuid.com .

The laptop BIOS does not offer this level of control, and that
is just a BIOS interface issue. The BIOS hides lots of functions
that are "too hard" for the users who buy the laptops. My laptop
only has one stinking setting. The manufacturer would have used
zero settings... if they could get away with it. Having one
setting ("AHCI") was a compromise for them.

The problem with the modern BIOS, is they have gone "five layers deep"
on the menus, making it just about impossible for people to find stuff.
for example, I found an ECC setting for the other machine, and I did
not know it was there. It was down five levels. And I wasn't looking
for it at the time.

Paul

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On 2024-12-03 07:08, Paul wrote:

On Mon, 12/2/2024 3:55 PM, micky wrote:

In alt.comp.os.windows-10, on Mon, 02 Dec 2024 12:48:07 -0500, micky
<NONONOmisc07@fmguy.com> wrote:

....

A device with a base clock of 1.1GHz and a turbo of 4.4GHz, this
implies that the device is severely thermally limited, it attempts
to burn your lap, it has no fan, it's too thin for a fan. No,
you don't want one of those, unless you like "molasses slow"
general operation. It may whomp up to 4.4Ghz for a couple seconds,
and then plonk down to 1.1GHz for the remainder of the operation.

The idea is that the laptop is more of the time idling, and when you
request some thing (reformat a page in office write, or recalculate a
sheet, download and render a web page) the laptop cpu goes into high
speed for a few seconds to do the job, then goes back to idle and
dissipates the heat. It possibly has no fan so no way to keep at top
power continuously.

(mind, running at high power with a fan also wears the fan. Maybe in
that case it is worth it to place the laptop on top of a fan platform)

The problem, is the background loading of a modern PC. There
is too much garbage in the PC, to ever see a decent turbo
on a 2 watt tablet.

You want a better balance, a 35W or 45W processor, a cooling
system, a fan. Then your base clock rises a bit, there is
turbo, but, you don't care if the turbo isn't always running
at a super-high level. The thing gives enough performance,
the Windows Defender is happy, the Search Index is chewing
on stuff, and maybe the scroll bar works on Firefox :-)

Right.

--
Cheers, Carlos.

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On 02/12/2024 18:02, micky wrote:

Know anything about Nimo computers? A family business founded in
Cincinatti or New Castle, Delaware, depending on where you look. :-)

You'll be better off buying from Dell (refurbished). You can use coupon
code CYBER5320 for Latitude 5320 laptops and get 55% discount.

<https://www.dellrefurbished.com/laptops?model_f[]=Latitude%205320>

Always buy the branded products to get better support online and on you
tube. You also get latest drivers and BIOS firmware from Dell website including PDF files for user and service manuals.


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: To protect and to server (3:633/280.2@fidonet)

On 03/12/2024 17:07, Jack wrote:

On 02/12/2024 18:02, micky wrote:

Know anything about Nimo computers? A family business founded in
Cincinatti or New Castle, Delaware, depending on where you look. :-)

You'll be better off buying from Dell (refurbished). You can use coupon
code CYBER5320 for Latitude 5320 laptops and get 55% discount.

<https://www.dellrefurbished.com/laptops?model_f[]=Latitude%205320>

Always buy the branded products to get better support online and on you
tube. You also get latest drivers and BIOS firmware from Dell website including PDF files for user and service manuals.

These are other coupon codes:

Cyber Week Deals:

55% off any Latitude 5320 Laptop Plus Free Ground Shipping | Coupon
Code: CYBER5320

55% off any Latitude 5420 Laptop Plus Free Ground Shipping | Coupon
Code: CYBER5420

55% off any Latitude 5520 Laptop Plus Free Ground Shipping | Coupon
Code: CYBER5520

Get up to 50% off any item in our Cyber Sale section Plus Free Ground
Shipping | No Coupon Code needed


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: To protect and to server (3:633/280.2@fidonet)

On Mon, 2 Dec 2024 21:54:14 +0100, Carlos E.R. wrote:

[snip]

Traditionally, the "turbo" button meant normal speed. The other setting
was actually "slow", and it was used for games and other software
designed for slower speed.

This happened around 1985.

I had a case with a digital speed display. It was NOT a frequency counter (maybe if was supposed to fool people). It was configured with jumpers to
show 1 of 2 numbers.

Go figure what they mean now.

Maybe there's no way to get it to run at the "turbo" speed for very long.
Its just short bursts and most of the time it runs at the normal speed.

--
22 days until the winter celebration (Wednesday, December 25, 2024
12:00:00 AM for 1 day).

Mark Lloyd
http://notstupid.us/

"auto-correct is my worst enema."

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On Mon, 2 Dec 2024 23:33:07 +0100, Carlos E.R. wrote:

[snip]

I remember two games on my first computer, which was an Amstrad
PC1512DD. It was around 3 or 4 times faster than the original IBM PC.
They were called "The ancient art of war" and "The ancient art of war at sea". They used the CPU speed to time the action. On computers by 1990
the game played way too fast to be be playable. We had to use the Turbo button to slow the computer, or run a small TSR (terminate and stay
resident program) that would waste time and make the computer slow
enough to play again :-D

I have used such programs, which were written with "do nothing" loops for timing. These would run too fast on modern machines.

[snip]

--
22 days until the winter celebration (Wednesday, December 25, 2024
12:00:00 AM for 1 day).

Mark Lloyd
http://notstupid.us/

"auto-correct is my worst enema."

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

[snip]

"turbo" was mostly applicable to 80386 and 80406 processors back in the
old days. Most people pushed the "turbo" button and left it that way
forever.

My first IBM clone had an 8088* that had turbo. It could run at 8MHz or
the original 4.77MHz.

* - actually a NEC V20 (8088 compatible). I never had a 80286, but I
remember seeing a lot of ads that mentioned turbo.

[snip]

--
22 days until the winter celebration (Wednesday, December 25, 2024
12:00:00 AM for 1 day).

Mark Lloyd
http://notstupid.us/

"auto-correct is my worst enema."

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On 12/3/2024 10:07 AM, Jack wrote:

On 02/12/2024 18:02, micky wrote:

Know anything about Nimo computers? A family business founded in
Cincinatti or New Castle, Delaware, depending on where you look. :-)

You'll be better off buying from Dell (refurbished). You can use coupon
code CYBER5320 for Latitude 5320 laptops and get 55% discount.

<https://www.dellrefurbished.com/laptops?model_f[]=Latitude%205320>

I always wonder about the battery condition on refurbished stuff. It
might seem OK but could have half its cycles used up.

Always buy the branded products to get better support online and on you
tube. You also get latest drivers and BIOS firmware from Dell website including PDF files for user and service manuals.

BTW I'm posting with a Dell Latitude 5320 Business Laptop (13.3" FHD,
Intel Core i7-1185G7 vPro, 16GB RAM, 512GB SSD, IST Card) Ultralight, Thunderbolt 4, Backlit, Fingerprint, Webcam, NFC, Latitude 5000, Win 11
Pro, Gray. Bought it new for $699 from (where else) Amazon. Love it...



--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Tue, 12/3/2024 8:59 AM, Carlos E.R. wrote:

On 2024-12-03 07:08, Paul wrote:

On Mon, 12/2/2024 3:55 PM, micky wrote:

In alt.comp.os.windows-10, on Mon, 02 Dec 2024 12:48:07 -0500, micky
<NONONOmisc07@fmguy.com> wrote:

...

A device with a base clock of 1.1GHz and a turbo of 4.4GHz, this
implies that the device is severely thermally limited, it attempts
to burn your lap, it has no fan, it's too thin for a fan. No,
you don't want one of those, unless you like "molasses slow"
general operation. It may whomp up to 4.4Ghz for a couple seconds,
and then plonk down to 1.1GHz for the remainder of the operation.

The idea is that the laptop is more of the time idling, and when you request some thing (reformat a page in office write, or recalculate a sheet, download and render a web page) the laptop cpu goes into high speed for a few seconds to do the job, then goes back to idle and dissipates the heat. It possibly has
no fan so no way to keep at top power continuously.

(mind, running at high power with a fan also wears the fan. Maybe in that case
it is worth it to place the laptop on top of a fan platform)

This pattern in Windows, plus moderately new hardware, shows we have achieved the design objective "hinted at" maybe fifteen years ago. Namely that
we would have sufficient power control, to gradually drop the hardware
to "virtual zero". This picture is getting damn close. Seven of the cores
seem to be in C12, one core is running, that one core is below 1.1GHz.
The clock on that one core, can drop lower than that.

[Picture]

https://i.postimg.cc/LXtp15jR/AMD-Ryzen-Master-machine-state.gif

Linux can't quite do this today, but that's not to say it could not
be modified to get here. There is no Ryzen Master for Linux, that I know
of, but it's pretty hard to figure out what AMD has to offer most
of the time.

I like the Ryzen Master as an alternative/adjunct to Task Manager.
If the machine is disturbed a bit, then the cores can all have relatively
low clocks applied to them. AMD has reduced the step size on the multiplier,
so it is less than the 100MHz increments popular in the past. But this
doesn't really matter, and is just more details about clock synthesis
with PLLs and divider chains.

The PC in that picture, idles at 32-33W, but because it's got a video
card in it, that pollutes the measurement. I can't pull the video
card, because of the address space bug that machine has. If I cared to
measure the machine with the 5600G in it, I could pull numbers 22W or less
from it. The 5600G just doesn't have the top clock that a more
expensive CPU would have. But that's an example of a processor
with enough cores, for modern Windows operation. You can have
a core or two for your own usage.

Paul


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Tue, 12/3/2024 12:52 PM, Mark Lloyd wrote:

[snip]

"turbo" was mostly applicable to 80386 and 80406 processors back in the
old days. Most people pushed the "turbo" button and left it that way
forever.

My first IBM clone had an 8088* that had turbo. It could run at 8MHz or
the original 4.77MHz.

* - actually a NEC V20 (8088 compatible). I never had a 80286, but I remember seeing a lot of ads that mentioned turbo.

[snip]

But that's an 8088. Would turbo even mean anything on a
chip like that ?

That would be, like, overclocking an 1802.

The ultimate overclockers, were the micros that were made
using 15V CMOS. They were specified to run on 5V logic,
so that the chip could interface to the other chips in
the design (5V RAM chips). But, the "speed" of the device was
proportional to that voltage. If you lifted the processor to +15V power,
it could go at 3X the speed (linear scaling). Instead of
my Celeron 50% overclock, from 300 to 450, a processor
made like that, could be 300% speed. Which I thought was
rather cool at the time, even if it didn't make a bit of
difference. As long as every bus in the computer is starving
the CPU of good things to eat, a lot of these "switches"
are worthless.

I got a taste of that on my AthlonXP processor. It was running
at 2200MHz and was known as 3200 model. So I overclocked it,
from 2200 to 2400 or 2500 or so. I benched it. The benchmarks
didn't go any faster, and they didn't go faster because the
front side bus was a "piece of work" and worthless. I immediately
turned it back down to 2200 again.

The whole history of personal computing, is just one giant sad panda :-)
It took forever, before decent interconnect was put in computers.

Paul

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On Tue, 12/3/2024 12:40 PM, Mark Lloyd wrote:

On Mon, 2 Dec 2024 21:54:14 +0100, Carlos E.R. wrote:

[snip]

Traditionally, the "turbo" button meant normal speed. The other setting
was actually "slow", and it was used for games and other software
designed for slower speed.

This happened around 1985.

I had a case with a digital speed display. It was NOT a frequency counter (maybe if was supposed to fool people). It was configured with jumpers to show 1 of 2 numbers.

Go figure what they mean now.

Maybe there's no way to get it to run at the "turbo" speed for very long. Its just short bursts and most of the time it runs at the normal speed.

In Windows, we're spoiled by the automation (and the lack of utilities
to display what is going on, and what our hardware is capable of).
You have to hunt around, to find your AMD Ryzen Master or your Intel XTU.

https://forums.opensuse.org/t/opensuse-is-not-utilizing-turbo-boost/88897

Here, they use the CPU-Z (Windows) text file report, to find the turbo table. Golly, I can't wait. But this is for a ten year old CPU, open loop control,
and they did use a hard wired turbo table back then. It could use turbo
until it overheated, but the whack-a-mole controls of ten years ago, are
not quite as smooth as the closed loop control today (achieves same effect,
but maybe without squirting NOOPs into the decoder).

https://forums.anandtech.com/threads/what-is-the-real-turbo-boost-for-i7-4930k.2380370/

The info for AMD (and for Intel newer generations) could be different. 5700G

( https://www.techpowerup.com/cpu-specs/ryzen-7-5700g.c2472 # 3.8GHz 4.6GHz BaseClock is 100MHz )

Package Power Tracking (PPT) 88.0 W (current) 65W processor
Thermal Design Current (TDC) 65.0 A (current)
Thermal Design Current SoC (TDC SoC) 50.0 A (current)
Electrical Design Current (EDC) 95.0 A (current)
Electrical Design Current SoC (EDC SoC) 75.0 A (current)

Temperature 0 34 degC (93 degF) (Package)
Temperature 1 27 degC (81 degF) (Cores (Max))

Temperature 2 27 degC (81 degF) (Core #0)
Temperature 3 27 degC (80 degF) (Core #1)
Temperature 4 26 degC (79 degF) (Core #2)
Temperature 5 26 degC (78 degF) (Core #3)
Temperature 6 26 degC (79 degF) (Core #4)
Temperature 7 26 degC (79 degF) (Core #5)
Temperature 8 26 degC (78 degF) (Core #6)
Temperature 9 27 degC (80 degF) (Core #7)

Temperature 10 27 degC (80 degF) (L3 Cache)
Temperature 11 26 degC (78 degF) (GFX) <=== iGPU is OFF
Voltage 0 1.36 Volts (VID)
Voltage 1 1.36 Volts (CPU VDD)
Voltage 2 1.10 Volts (SoC VDD)
Power 00 12.39 W (Package)
Power 01 2.62 W (Cores)
Power 02 1.62 W (Core #0)
Power 03 0.88 W (Core #1)
Power 04 0.06 W (Core #2)
Power 05 0.12 W (Core #3)
Power 06 0.01 W (Core #4)
Power 07 0.29 W (Core #5)
Power 08 0.01 W (Core #6)
Power 09 0.49 W (Core #7)
Power 10 0.38 W (L3 Cache)
Clock Speed 0 2992.30 MHz (Core #0) # Not sure what these bogus values mean
Clock Speed 1 2992.30 MHz (Core #1) # Ryzen Master does not show these particular values
Clock Speed 2 2992.30 MHz (Core #2)
Clock Speed 3 2992.30 MHz (Core #3)
Clock Speed 4 2992.30 MHz (Core #4)
Clock Speed 5 2992.30 MHz (Core #5)
Clock Speed 6 2992.30 MHz (Core #6)
Clock Speed 7 2992.30 MHz (Core #7)
Core 0 max ratio (effective) 46.50
Core 1 max ratio (effective) 44.75
Core 2 max ratio (effective) 46.50
Core 3 max ratio (effective) 44.75
Core 4 max ratio (effective) 44.75
Core 5 max ratio (effective) 44.75
Core 6 max ratio (effective) 37.20 <=== A core that can't turbo ? No. Did a load test next.
Core 7 max ratio (effective) 46.50

With 7ZIP compression run. Cooler is after-market, blow-down type (Noctua).

Temperature 0 56 degC (131 degF) (Package) <=== toasty warm, no skin burns today.
Temperature 1 53 degC (127 degF) (Cores (Max)) Unlike a laptop, there is no reason to throttle
Temperature 2 49 degC (120 degF) (Core #0) or to act in an adverse manner. At a guess.
Temperature 3 50 degC (122 degF) (Core #1) Even the stock cooler, used on this thing, is enough.
Temperature 4 53 degC (126 degF) (Core #2) This after-market cooler is a reject from another
Temperature 5 50 degC (122 degF) (Core #3) project, and hey, why not use it for this. I didn't
Temperature 6 52 degC (124 degF) (Core #4) buy this to replace the stock cooler, just why use
Temperature 7 49 degC (120 degF) (Core #5) the stock cooler when you have left-overs to eat.
Temperature 8 50 degC (122 degF) (Core #6)
Temperature 9 48 degC (119 degF) (Core #7)
Temperature 10 40 degC (104 degF) (L3 Cache)
Temperature 11 35 degC (94 degF) (GFX)
Voltage 0 1.34 Volts (VID)
Voltage 1 1.34 Volts (CPU VDD)
Voltage 2 1.09 Volts (SoC VDD)
Power 00 64.23 W (Package)
Power 01 46.59 W (Cores)
Power 02 5.99 W (Core #0)
Power 03 6.23 W (Core #1)
Power 04 6.33 W (Core #2)
Power 05 5.37 W (Core #3)
Power 06 5.70 W (Core #4)
Power 07 5.03 W (Core #5)
Power 08 6.42 W (Core #6)
Power 09 5.51 W (Core #7)
Power 10 1.64 W (L3 Cache)
Clock Speed 0 4612.87 MHz (Core #0) <=== one or two fractional multipliers below max
Clock Speed 1 4612.87 MHz (Core #1) "Meets description on tin" even if not all that exciting.
Clock Speed 2 4637.81 MHz (Core #2) Would do this for the entire 7ZIP compression run.
Clock Speed 3 4612.87 MHz (Core #3)
Clock Speed 4 4612.87 MHz (Core #4)
Clock Speed 5 4612.87 MHz (Core #5)
Clock Speed 6 4612.87 MHz (Core #6)
Clock Speed 7 4612.87 MHz (Core #7)
Core 0 max ratio (effective) 46.50
Core 1 max ratio (effective) 46.50
Core 2 max ratio (effective) 46.50
Core 3 max ratio (effective) 46.50
Core 4 max ratio (effective) 46.50
Core 5 max ratio (effective) 46.50
Core 6 max ratio (effective) 46.50 <=== seems OK here, we are turbo on all cores on a gutless processor. Yay.
Core 7 max ratio (effective) 46.50

*******

Try a second machine, for a change of complexion.

( https://www.techpowerup.com/cpu-specs/ryzen-9-5950x.c2364 5950X 2 silicon die, 3.4 GHz 4.9 GHz (lower base, higher turbo)

TDP Limit 105.0 Watts # The report format is different. Weird.
Tjmax 90.0 C

Package Power Tracking (PPT) Current : 142.0 W, fused : 142.0 W
Thermal Design Current (TDC) Current : 95.0 A, fused : 95.0 A
Electrical Design Current (EDC) Current : 140.0 A, fused : 140.0 A

Temperature 0 31 degC (88 degF) (Package)
Temperature 1 29 degC (84 degF) (CCD #0)
Temperature 2 28 degC (82 degF) (CCD #1)
Temperature 3 28 degC (82 degF) (Cores (Max))

Temperature 4 28 degC (82 degF) (Core #0) # We are Idle for this run.
Temperature 5 28 degC (82 degF) (Core #1)
Temperature 6 27 degC (80 degF) (Core #2)
Temperature 7 27 degC (80 degF) (Core #3)
Temperature 8 27 degC (80 degF) (Core #4)
Temperature 9 26 degC (79 degF) (Core #5)
Temperature 10 27 degC (81 degF) (Core #6)
Temperature 11 26 degC (79 degF) (Core #7)
Temperature 12 27 degC (80 degF) (Core #8)
Temperature 13 27 degC (80 degF) (Core #9)
Temperature 14 26 degC (79 degF) (Core #10)
Temperature 15 26 degC (78 degF) (Core #11)
Temperature 16 26 degC (78 degF) (Core #12)
Temperature 17 26 degC (78 degF) (Core #13)
Temperature 18 26 degC (78 degF) (Core #14)
Temperature 19 26 degC (78 degF) (Core #15)

Temperature 20 28 degC (81 degF) (L3 Cache #0)
Temperature 21 27 degC (80 degF) (L3 Cache #1)
Temperature 22 28 degC (81 degF) (SoC)
Voltage 0 0.95 Volts (VID)
Voltage 1 1.04 Volts (CPU VDD)
Voltage 2 0.99 Volts (SoC VDD)
Power 00 30.66 W (Package)
Power 01 1.94 W (Cores)
Power 02 0.81 W (Core #0) \___ Background load skulking on low cores.
Power 03 0.55 W (Core #1) /
Power 04 0.11 W (Core #2)
Power 05 0.15 W (Core #3)
Power 06 0.03 W (Core #4)
Power 07 0.02 W (Core #5)
Power 08 0.60 W (Core #6)
Power 09 0.02 W (Core #7)
Power 10 0.02 W (Core #8)
Power 11 0.02 W (Core #9)
Power 12 0.02 W (Core #10)
Power 13 0.01 W (Core #11)
Power 14 0.01 W (Core #12)
Power 15 0.01 W (Core #13)
Power 16 0.02 W (Core #14) <=== The power is not OFF, but it is low
Power 17 0.01 W (Core #15)
Power 18 0.96 W (L3 Cache #0)
Power 19 0.43 W (L3 Cache #1)
Clock Speed 0 5048.82 MHz (Core #0) <=== Turboed, while the report is being generated ? Clever.
Clock Speed 1 3599.16 MHz (Core #1)
Clock Speed 2 3599.16 MHz (Core #2)
Clock Speed 3 3599.16 MHz (Core #3)
Clock Speed 4 4039.06 MHz (Core #4)
Clock Speed 5 4039.06 MHz (Core #5)
Clock Speed 6 3599.16 MHz (Core #6)
Clock Speed 7 4039.06 MHz (Core #7)
Clock Speed 8 3599.16 MHz (Core #8)
Clock Speed 9 3599.16 MHz (Core #9)
Clock Speed 10 3599.16 MHz (Core #10)
Clock Speed 11 3599.16 MHz (Core #11)
Clock Speed 12 3599.16 MHz (Core #12)
Clock Speed 13 3599.16 MHz (Core #13)
Clock Speed 14 3599.16 MHz (Core #14)
Clock Speed 15 3599.16 MHz (Core #15)
Clock Speed 16 3599.16 MHz (L3 Cache #0)
Clock Speed 17 3599.16 MHz (L3 Cache #1)
Core 0 max ratio (effective) 50.50
Core 1 max ratio (effective) 50.50
Core 2 max ratio (effective) 50.50
Core 3 max ratio (effective) 48.75
Core 4 max ratio (effective) 49.75
Core 5 max ratio (effective) 49.75
Core 6 max ratio (effective) 50.50
Core 7 max ratio (effective) 49.75
Core 8 max ratio (effective) 48.75
Core 9 max ratio (effective) 39.40
Core 10 max ratio (effective) 39.40
Core 11 max ratio (effective) 39.40
Core 12 max ratio (effective) 49.25
Core 13 max ratio (effective) 39.40
Core 14 max ratio (effective) 39.40
Core 15 max ratio (effective) 39.40

With 7ZIP compression run. Cooler is after-market, two-fan, sideways, 250W cooling rating, DeepCool AK620
Same file is being compressed as in the other example. Files on RAMDisk. Five case fans o.O Groan.

Temperature 0 69 degC (156 degF) (Package)
Temperature 1 66 degC (149 degF) (CCD #0)
Temperature 2 61 degC (141 degF) (CCD #1)
Temperature 3 64 degC (147 degF) (Cores (Max))
Temperature 4 60 degC (140 degF) (Core #0)
Temperature 5 60 degC (140 degF) (Core #1)
Temperature 6 62 degC (144 degF) (Core #2)
Temperature 7 64 degC (146 degF) (Core #3)
Temperature 8 62 degC (143 degF) (Core #4)
Temperature 9 64 degC (147 degF) (Core #5)
Temperature 10 59 degC (138 degF) (Core #6)
Temperature 11 62 degC (143 degF) (Core #7)
Temperature 12 56 degC (132 degF) (Core #8)
Temperature 13 54 degC (129 degF) (Core #9)
Temperature 14 60 degC (140 degF) (Core #10)
Temperature 15 56 degC (132 degF) (Core #11)
Temperature 16 61 degC (142 degF) (Core #12)
Temperature 17 56 degC (132 degF) (Core #13)
Temperature 18 60 degC (139 degF) (Core #14)
Temperature 19 56 degC (132 degF) (Core #15)
Temperature 20 45 degC (112 degF) (L3 Cache #0)
Temperature 21 44 degC (111 degF) (L3 Cache #1)
Temperature 22 39 degC (101 degF) (SoC)
Voltage 0 1.27 Volts (VID)
Voltage 1 1.28 Volts (CPU VDD)
Voltage 2 1.00 Volts (SoC VDD)
Power 00 144.87 W (Package)
Power 01 108.10 W (Cores)
Power 02 7.41 W (Core #0)
Power 03 8.28 W (Core #1)
Power 04 7.28 W (Core #2)
Power 05 7.98 W (Core #3)
Power 06 7.43 W (Core #4)
Power 07 8.06 W (Core #5)
Power 08 7.27 W (Core #6)
Power 09 8.20 W (Core #7)
Power 10 5.11 W (Core #8) \
Power 11 5.12 W (Core #9) \
Power 12 5.92 W (Core #10) \
Power 13 5.40 W (Core #11) \___ Second CCX always has lower loading
Power 14 6.70 W (Core #12) / 7ZIP cannot be set to 64 threads, to really load these
Power 15 5.16 W (Core #13) /
Power 16 6.38 W (Core #14) /
Power 17 5.72 W (Core #15) /
Power 18 5.27 W (L3 Cache #0)
Power 19 3.20 W (L3 Cache #1)
Clock Speed 0 4573.93 MHz (Core #0) <=== Could not make it to 4.9GHz. (fused) power limited
Clock Speed 1 4573.93 MHz (Core #1) <=== Will run like this all day long
Clock Speed 2 4573.93 MHz (Core #2) <=== 70C operating is less than TJMax of 90C
Clock Speed 3 4573.93 MHz (Core #3)
Clock Speed 4 4573.93 MHz (Core #4) This is "a" form of Turbo, but not max-Turbo on all cores.
Clock Speed 5 4598.93 MHz (Core #5) I'm not complaining. Bench is OK. Meets expectations.
Clock Speed 6 4498.95 MHz (Core #6)
Clock Speed 7 4598.93 MHz (Core #7)
Clock Speed 8 4573.93 MHz (Core #8)
Clock Speed 9 4623.92 MHz (Core #9)
Clock Speed 10 4523.95 MHz (Core #10)
Clock Speed 11 4573.93 MHz (Core #11)
Clock Speed 12 4498.95 MHz (Core #12)
Clock Speed 13 4498.95 MHz (Core #13)
Clock Speed 14 4523.95 MHz (Core #14)
Clock Speed 15 4623.92 MHz (Core #15)
Clock Speed 16 4573.93 MHz (L3 Cache #0)
Clock Speed 17 4573.93 MHz (L3 Cache #1)
Core 0 max ratio (effective) 46.00
Core 1 max ratio (effective) 46.00
Core 2 max ratio (effective) 46.00
Core 3 max ratio (effective) 46.00
Core 4 max ratio (effective) 46.00
Core 5 max ratio (effective) 46.00
Core 6 max ratio (effective) 46.00
Core 7 max ratio (effective) 46.00
Core 8 max ratio (effective) 46.00
Core 9 max ratio (effective) 46.25
Core 10 max ratio (effective) 46.00
Core 11 max ratio (effective) 46.00
Core 12 max ratio (effective) 46.00
Core 13 max ratio (effective) 46.00
Core 14 max ratio (effective) 46.00
Core 15 max ratio (effective) 46.25

A little SuperPI, single thread for a hero-run. This run is "contrived", in the sense that I did a first run, to see which core the hardware selected automatically
for the computation. On the second run, I used the Task Manager "affinity" option,
to assign the executable to that single good core. This lets the processor "put on a good
show for the folks", as this is the only core that hits just over 5GHz. some of the
others hit, like 4950, which isn't as impressive.

Temperature 0 60 degC (139 degF) (Package)
Temperature 1 59 degC (138 degF) (CCD #0)
Temperature 2 33 degC (90 degF) (CCD #1)
Temperature 3 56 degC (132 degF) (Cores (Max))
Temperature 4 37 degC (99 degF) (Core #0)
Temperature 5 35 degC (94 degF) (Core #1)
Temperature 6 36 degC (96 degF) (Core #2)
Temperature 7 31 degC (87 degF) (Core #3)
Temperature 8 37 degC (98 degF) (Core #4)
Temperature 9 30 degC (85 degF) (Core #5)
Temperature 10 56 degC (132 degF) (Core #6) <=== This one carries the hot potato. Affinity points here. Affinity=Core13 for Core#6
Temperature 11 29 degC (84 degF) (Core #7)
Temperature 12 32 degC (89 degF) (Core #8) <=== The second silicon die is ~5C cooler than the hot potato die
Temperature 13 31 degC (87 degF) (Core #9)
Temperature 14 29 degC (84 degF) (Core #10)
Temperature 15 28 degC (82 degF) (Core #11)
Temperature 16 29 degC (83 degF) (Core #12)
Temperature 17 28 degC (81 degF) (Core #13)
Temperature 18 28 degC (83 degF) (Core #14)
Temperature 19 28 degC (81 degF) (Core #15)
Temperature 20 33 degC (91 degF) (L3 Cache #0)
Temperature 21 30 degC (86 degF) (L3 Cache #1)
Temperature 22 29 degC (84 degF) (SoC)
Voltage 0 1.46 Volts (VID)
Voltage 1 1.48 Volts (CPU VDD)
Voltage 2 1.00 Volts (SoC VDD)
Power 00 69.15 W (Package)
Power 01 19.49 W (Cores)
Power 02 2.84 W (Core #0) <=== System activity is on here, apparently.
Power 03 0.34 W (Core #1)
Power 04 1.62 W (Core #2)
Power 05 0.13 W (Core #3)
Power 06 0.24 W (Core #4)
Power 07 0.19 W (Core #5)
Power 08 13.55 W (Core #6) <=== Is it warm in here, or is it just me ?
Power 09 0.17 W (Core #7)
Power 10 0.09 W (Core #8) \
Power 11 0.06 W (Core #9) \
Power 12 0.06 W (Core #10) \
Power 13 0.05 W (Core #11) \___ Second die is having a nap.
Power 14 0.05 W (Core #12) /
Power 15 0.06 W (Core #13) /
Power 16 0.05 W (Core #14) /
Power 17 0.08 W (Core #15) /
Power 18 4.37 W (L3 Cache #0)
Power 19 1.90 W (L3 Cache #1)
Clock Speed 0 4039.06 MHz (Core #0)
Clock Speed 1 3979.07 MHz (Core #1)
Clock Speed 2 4039.06 MHz (Core #2)
Clock Speed 3 4039.06 MHz (Core #3)
Clock Speed 4 3979.07 MHz (Core #4)
Clock Speed 5 3979.07 MHz (Core #5)
Clock Speed 6 5048.82 MHz (Core #6) <=== My best turbo on this pig
Clock Speed 7 4039.06 MHz (Core #7)
Clock Speed 8 3599.16 MHz (Core #8) <=== These aren't really running at 3600, not at 60mW.
Clock Speed 9 3599.16 MHz (Core #9) But if you attempt to probe a core, you "wake it up"
Clock Speed 10 3599.16 MHz (Core #10) which is the "pollution problem" with this work.
Clock Speed 11 3599.16 MHz (Core #11) That's why we should not be slapping the dev around
Clock Speed 12 3599.16 MHz (Core #12) when seeing anomalous numbers like this. There's a reason
Clock Speed 13 3599.16 MHz (Core #13) these things happen.
Clock Speed 14 3599.16 MHz (Core #14)
Clock Speed 15 3599.16 MHz (Core #15)
Clock Speed 16 4998.83 MHz (L3 Cache #0)
Clock Speed 17 3599.16 MHz (L3 Cache #1)
Core 0 max ratio (effective) 50.50 The info in this portion of the table
Core 1 max ratio (effective) 50.25 seems mostly bogus.
Core 2 max ratio (effective) 50.50
Core 3 max ratio (effective) 40.40 You are sometimes suspicious that the
Core 4 max ratio (effective) 50.00 core numbering on modern hardware, is
Core 5 max ratio (effective) 50.00 "virtualized", because many times the
Core 6 max ratio (effective) 50.50 pattern on Task Manager is not what it should be.
Core 7 max ratio (effective) 40.40
Core 8 max ratio (effective) 36.00
Core 9 max ratio (effective) 36.00
Core 10 max ratio (effective) 36.00
Core 11 max ratio (effective) 36.00
Core 12 max ratio (effective) 36.00
Core 13 max ratio (effective) 36.00
Core 14 max ratio (effective) 36.00
Core 15 max ratio (effective) 36.00

You can tease out the behaviors, with a little work, and the right report.

Paul

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

On 2024-12-04 04:54, Paul wrote:

On Tue, 12/3/2024 12:52 PM, Mark Lloyd wrote:

[snip]

"turbo" was mostly applicable to 80386 and 80406 processors back in the
old days. Most people pushed the "turbo" button and left it that way
forever.

My first IBM clone had an 8088* that had turbo. It could run at 8MHz or
the original 4.77MHz.

* - actually a NEC V20 (8088 compatible). I never had a 80286, but I
remember seeing a lot of ads that mentioned turbo.

[snip]

But that's an 8088. Would turbo even mean anything on a
chip like that ?

Yes (without overclocking).

You could find IBM PC clones, using the 8088 or equivalents at various
speeds. The normal speed was called turbo. The non turbo speed was
actually putting the CPU at slow speed. Just the reverse.

--
Cheers, Carlos.

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On Tue, 3 Dec 2024 22:54:20 -0500, Paul wrote:

On Tue, 12/3/2024 12:52 PM, Mark Lloyd wrote:

[snip]

"turbo" was mostly applicable to 80386 and 80406 processors back in
the old days. Most people pushed the "turbo" button and left it that
way forever.

My first IBM clone had an 8088* that had turbo. It could run at 8MHz or
the original 4.77MHz.

* - actually a NEC V20 (8088 compatible). I never had a 80286, but I
remember seeing a lot of ads that mentioned turbo.

[snip]

But that's an 8088. Would turbo even mean anything on a chip like that ?

It wasn't overclocking. This board actually had an 8MHz CPU. That was considered "turbo" compared to the 4.77MHz clock of the original IBM-PC,
which it could work at for compatibility. The 80286 and later systems used
a higher base frequency,

That would be, like, overclocking an 1802.

Was that the one from RCA?

[snip]

--
21 days until the winter celebration (Wednesday, December 25, 2024
12:00:00 AM for 1 day).

Mark Lloyd
http://notstupid.us/

"I do not consider it a sign of divine love to consign to hell people
who live good lives but make an honest mistake in belief" [Moshe
Shulman]

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: ---:- FTN<->UseNet Gate -:--- (3:633/280.2@fidonet)

On 05/12/2024 17:49, AJL wrote:

These newer LTs with the glued in batteries scare me.

You will find all laptops these days will come with things glued
together. They use very few screws to save few cents and to make it
harder for users to open the box and upgrade. They want people to throw
away their laptops and buy new one. We are living in a throw-away
society. No wonder landfill sites are full of toxic materials.


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: To protect and to server (3:633/280.2@fidonet)

On Thu, 05 Dec 2024 17:49:50 -0500, micky <NONONOmisc07@fmguy.com> wrote:

In alt.comp.os.windows-10, on Thu, 5 Dec 2024 20:27:30 +0000, New
Gadgets Skeptic <gtyhjuikj@mkjuipkikjh.com> wrote:

On 05/12/2024 17:49, AJL wrote:

These newer LTs with the glued in batteries scare me.

HOw come when I look at AJL's post from this time and date, I don't find
the line above? I don't see it anywhere in the thread!!!

Because New Gadgets Skeptic uses a different date format and a different time zone.

If you have Agent configured appropriately, you can just dbl-click the following
MID to open the message:
Message-ID: <visp2h$1ok38$1@dont-email.me>


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: Newshosting.com - Highest quality at a great p (3:633/280.2@fidonet)

On Thu, 12/5/2024 7:38 PM, Char Jackson wrote:

On Thu, 05 Dec 2024 17:49:50 -0500, micky <NONONOmisc07@fmguy.com> wrote:

In alt.comp.os.windows-10, on Thu, 5 Dec 2024 20:27:30 +0000, New
Gadgets Skeptic <gtyhjuikj@mkjuipkikjh.com> wrote:

On 05/12/2024 17:49, AJL wrote:

These newer LTs with the glued in batteries scare me.

HOw come when I look at AJL's post from this time and date, I don't find
the line above? I don't see it anywhere in the thread!!!

Because New Gadgets Skeptic uses a different date format and a different time zone.

If you have Agent configured appropriately, you can just dbl-click the following
MID to open the message:
Message-ID: <visp2h$1ok38$1@dont-email.me>

Just in case, look at it on Howard.

http://al.howardknight.net/?STYPE=msgid&MSGI=%3Cvisp2h%241ok38%241%40dont-email.me%3E

Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

User-Agent: PhoNews/3.13.3 (Android/11)

Even when headers can't be parsed, a news reader is still
expected to try to show the body. The User-Agent isn't
necessary for rendering the post, and is merely a decoration
for human consumption. The content declaration might upset
Agent, as I don't really know whether Agent has any of the
newer "features" supported.

Does Forte Agent handle 7bit and 8bit ?

As another example:

<vitdn3$1tgad$1@dont-email.me>

http://al.howardknight.net/?STYPE=msgid&MSGI=%3Cvitdn3%241tgad%241%40dont-email.me%3E

Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit

User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.3.1

That's an old-style code page that Agent should be able to understand.

Paul


--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: A noiseless patient Spider (3:633/280.2@fidonet)

Char Jackson <none@none.invalid> wrote:

On Thu, 05 Dec 2024 17:49:50 -0500, micky <NONONOmisc07@fmguy.com> wrote:

In alt.comp.os.windows-10, on Thu, 5 Dec 2024 20:27:30 +0000, New
Gadgets Skeptic <gtyhjuikj@mkjuipkikjh.com> wrote:

On 05/12/2024 17:49, AJL wrote:

These newer LTs with the glued in batteries scare me.

HOw come when I look at AJL's post from this time and date, I don't find >the line above? I don't see it anywhere in the thread!!!

Because New Gadgets Skeptic uses a different date format and a different time zone.

If you have Agent configured appropriately, you can just dbl-click the following MID to open the message:
Message-ID: <visp2h$1ok38$1@dont-email.me>

AJL posted his message only to alt.comp.os.windows-11 (probably a
newsreader ("PhoNews/3.13.3 (Android/11)") problem/limitation).

The rest of this thread is mostly crossposted to
alt.comp.os.windows-11 and alt.comp.os.windows-10, so if micky was
reading alt.comp.os.windows-10 first, he wouldn't (yet) see AJL post (I
didn't either).

So two 'errors': AJL's post not being crossposted and New Gadgets
Skeptic posted a crossposted response to a non-crossposted post, without
saying so.

--- MBSE BBS v1.0.8.4 (Linux-x86_64)
* Origin: NOYB (3:633/280.2@fidonet)