• Please review our updated Terms and Rules here

Pentium 4

hunterjwizzard said:
Yeah. We hit a time around the early 2000s where there was a massive leap forward. Since then things have pretty much plateaued.
Click to expand...

Intel was entirely responsible for that.

Things plateaued in 2008 when Intel released Nehalem and the core i series. Intel kept pleb desktop users on quad cores for the next 10 years because they had no competition. When AMD finally caught up with Ryzen, it rapidly accelerated CPU development again on Intel's side because they had quite literally nothing. They spent the next several years trying to cobble together anything that had any hope of competing, even going as far as backporting a server architecture into a larger process for desktop users, which ended up being slower than the generation before it. Then they spent even more time having to release emergency microcode fixes for processors going back to 2015 because they cut so many corners in design for speed improvements that it left massive security vulnerabilities.

Since 2018ish, things have very much changed. Both Intel and AMD have added increasing core counts to their chips to make heavy multitasking achievable to the pleb desktop users. You can now get up to 32 core CPUs for the desktop space and not have to spend tens of thousands of dollars on enterprise server gear.

Though Intel has seemed to not learn their hard lessons of the past and are no dealing with nuclear reactor levels of heat in their high end desktop parts. They could have learned from AMD's popcorn AM5 fiasco, but they just wanted one of their own so bad.
 
NeXT said:
Willamette core.
I really don't fancy the P4 because it was so many bad ideas on a chip and if you won the RAMBUS lottery, coupled with terrible memory. Even when I want to build an example P4 system I cannot help but want to make it incredibly cheesy because that generation really reminds me of people who paid way too much thinking it was the future and got smoked by a last-gen PIII.....then later on got burned by the first wave of bad capacitors.
That whole generation is something you point at and laugh whenever you see it. People running 98 on P4's are twice the fun because you can laugh at ugly strangers wearing ugly clothing.
Click to expand...
Nice write up NeXT. Makes me appreciated my 386/40 even more which could give the early P4's a run for their money. That's why the 386/40's were jerked from market. Too good.
 
GiGaBiTe said:
Things plateaued in 2008
Click to expand...
I dunno. I feel like things may be faster on paper since 2008, but that speed gets eaten up by bloatware and ad-delivery mechanisms. My modern computer today often feels slower than the one I had in 2008; it definitely doesn't complete tasks any faster.

Meanwhile I am busily scouring the net for an ancient copy of 3d studio max because it turns out, for some reason, the version from nearly 20 years ago actually works better on windows 10 than more recent builds. Go figure.
 
Agent Orange said:
Nice write up NeXT. Makes me appreciated my 386/40 even more which could give the early P4's a run for their money. That's why the 386/40's were jerked from market. Too good.
Click to expand...

Uhm… I mean, yeah, the P4, aka Pentium 4, had their problems, but I don’t think any 386/40 ever, ever gave one a “run for its money”.
 
Zare said:
Original 486DX is P4
Click to expand...

Who calls it that, and in what context? The "official" Intel name for it was "i486", aka "486" or "80486". If you can find a contemporary case of Intel *ever* referring to it as the "P4", well, the citation would be appreciated.

Obviously this is confusing because the codename for the original Pentium was "P5", with subsequent models having names like "P54C" and "P55C" and the Pentium Pro being coded "P6", but this codename (and the shipping name "Pentium") is a combination of "Penta" and the digit "5", both referring to the "fifth generation" of the x86 architecture. Clearly it's dumb that Intel *keeps calling* their processors "Pentium" when they should have been called "Hexium" (Sexium?), "Heptium" (Septium), whatever, but that doesn't mean it makes sense to retroactively slap a "P" on earlier generations.
 
Eudimorphodon said:
Who calls it that, and in what context? The "official" Intel name for it was "i486", aka "486" or "80486". If you can find a contemporary case of Intel *ever* referring to it as the "P4", well, the citation would be appreciated.

Obviously this is confusing because the codename for the original Pentium was "P5", with subsequent models having names like "P54C" and "P55C" and the Pentium Pro being coded "P6", but this codename (and the shipping name "Pentium") is a combination of "Penta" and the digit "5", both referring to the "fifth generation" of the x86 architecture. Clearly it's dumb that Intel *keeps calling* their processors "Pentium" when they should have been called "Hexium" (Sexium?), "Heptium" (Septium), whatever, but that doesn't mean it makes sense to retroactively slap a "P" on earlier generations.
Click to expand...
Given Intel's penchant for place names, how about "Hoquiam"?
 
I've heard rumors this is why Microsoft went with "Windows 10" as the follow-up to "Windows 8", and not because they cannot count.
 
Agent Orange said:
Possible Asus was the center of that fiasco,
Click to expand...

ALL motherboard vendors are to blame as well. They've been abusing CPUs for decades to try and get themselves pushed to the top of the charts. Bumping the FSB/BCLK/VCore up for stock settings has been the norm for decades, even at the detriment of the CPU.

Back in the Athlon era, Asus and others would overvolt the CPU, sometimes significantly, so they could run their +5, +10 or +15 MHz FSB bumps without causing system instability. These chips already ran smoking hot normally due to the lack of an IHS and crappy coolers of the time. They also had zero thermal protection and were known to burn/explode in the event of a severe thermal failure. I saw one run at 600F+ without a heatsink, it was very crispy and dead in seconds.
 
There use to be a video floating around where they took the heat sinks off various CPUs while playing the then-hotest video game. Was really funny watching them burn.
 
Zare said:
https://theretroweb.com/cpus/1011 see the "core" part.
BIOS of Advantech SBC with DX2/66 reports CPU as P24.
Click to expand...

The DX2 and DX4 having product designations starting with "P24" does appear in contemporary literature from Intel proper, but outside of a vague Quora post evidence that "P(a number)" was even an informal designation for x86 "generations" before the Pentium is pretty thin on the ground.

FWIW, this oral history of the creation of the 80386 processor, which is referenced by an article discussing the i960, says that codename "P4" originally referred to a "VAX-like" CPU design that was basically a do-over attempt for the team responsible for the failed iAXP432. It also mentions a separate "P7" project... so I'm going to stand corrected on "P" meaning "Penta", it seems like it means "Processor". (The i860 CPU was apparently the "N10", with "N" meaning "Numeric".) These sources also say that the "P" numbers were randomly reused, so depending on when you're talking about "Px" might refer to a different thing.

Zare said:
Btw. I was just responding to confusion about someone comparing 386 to Pentium 4 because of the acronym.
Click to expand...

Confusion made all the worse because the title of the thread is "Pentium 4" and regardless of whatever status "P4" might have as an "official" name if someone says "P4" when talking about an Intel CPU 999,999 times out of a million it's going to mean "Pentium 4". ;)
 
Agent Orange said:
I should have been more specific and said 486X/20
Click to expand...

There are references using the codename "P23" for the 486SX. Which notably doesn't have a 4 in it at all. ;)

Anyway. It's kind of amusing to ponder oddities like the claim that the 386sx was assigned the codename P9, but I think it suffices to say that any claim that codename "Px" = "Generation x" isn't particularly well supported and nobody actually does that. But it's quite common for people to say "P2/P3/P4" (or "PII/PIII/P4") when talking about the Pentium II through 4. It's sad they didn't call Core Pentium 5; it could have been the PentaPentium.
 
Eudimorphodon said:
The DX2 and DX4 having product designations starting with "P24" does appear in contemporary literature from Intel proper, but outside of a vague Quora post evidence that "P(a number)" was even an informal designation for x86 "generations" before the Pentium is pretty thin on the ground.
Click to expand...

These product designations is something I saw the first time the other day when I plugged in the Advantech board. It reported P24 as CPU. So I went on wikipedia and saw a table of internal core designations.
So it's real. In the sense that P-designation is coming out from real hardware. I wonder if it would say P4 if I put an i486DX into it. (which I don't have)

hunterjwizzard said:
There use to be a video floating around where they took the heat sinks off various CPUs while playing the then-hotest video game. Was really funny watching them burn.
Click to expand...

There was a video from early 2000s with guy powering on Athlon Thunderbird naked with thermometer on. It burned itself on fire in couple of seconds.
 
An Athlon 1400 is 72W TDP so without a heatsink of some kind to soak up the heat it will burn.

Now an i9-1400KS has a base TDP 150W up to 320W depending on the profile. Even with temp monitoring what would happen if you ran that with no heatsink at all (you can just plug a fan into the cpu fan power so there is a load for the BIOS)?
 
I thought Intel had thermal protection in the CPUs starting around the Pentium 4 era. Athlons burned because they had none.
 
You must log in or register to reply here.
Back
Top