The History Of Intel CPUs (Archive)

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
- 4004: I mention that it was a 4-bit CPU. It is literally the first sentence in the article.
- 80386: Did I say that the architecture didn't improve? The article literally states that it outperformed the 80286 when both systems used the same amount of RAM. Additional RAM increased performance a considerable amount too.
- 80386SL: The 80386SL was hardware locked to 16-bit operations. The article is technically correct because the CPU is incapable of performing 32-bit operations. It could execute 32-bit instructions, but it could only do this by breaking the instructions into two 16-bit operations that were consecutively executed. This was considerably slower than running on a 32-bit processor.

I honestly question how well you read this, as you state in several places that I forgot to mention things that are already in the article.[/quotemsg]

- 4004: Yep I missed that
- 80386: That is the point you did not talk about enhancaments while it was a big step forward. The performance increase did not came from memory increase. Typical amount was between 4-16MB those days and a 286 could handle same amount, but even if I put the very same modules to a 286 or 386 the second perform much better on the same clock. Of course it is a not that simple as there was no Intel 16MHz 286, but I had (and still have) it from AMD.
- 386SL: You just repeat what I criticize:

"...but was limited to 16-bit operations. It still supported a full 4GB of RAM, so it lost only the ability to run 16-bit applications."
"The 80386SL was hardware locked to 16-bit operations. "

That is simply not true. 386SL same as SL could run 32 bit commands. The memory bus was limited, but internally and on command level it was not limitated. Good examples are Windows 3.1 386 mode or doom. They were running on the 386SX or SL nice, but was slover than a full fledged 386DX. More details were nicely described by Mintch074 and others.
 
With the advent of "Willamette", the compounding issue of super-expensive RAMBUS compounded Intel's issues with getting Pentium 4 into the mainstream. Granted they created a bridge to allow DDR(2 or 3?) RAM to work but still...
 
Wow, this type of article is a magnet for nit picks. My sympathies for the author, but that doesn't mean I won't contribute my own!

the 16-bit data bus and execution hardware allowed the 8086 to simultaneously work on two eight-bit instructions.
Click to expand...
If 8086 was a superscalar CPU, that's news to me! I think it'd be accurate to say that it could operate on twice as much data per instruction.

... the 8088 around the same time. This processor was based on the 8088, but it disabled half of the address bus, limiting it to eight-bit operations. As it still had access to up to 1MB of RAM and ran at higher frequencies
Click to expand...
Seems misleading. It was merely a performance-reduced version of the 8086, due to crippling the memory bus (as you point out). It was instruction-set-compatible with the 8086, however, and could execute all of the former's 16-bit instructions. It was in my first PC.

And no mention of 8087?

80186 was never really used in PCs, AFAIK. I think it was mostly used in control applications, for some reason. Would be interesting to know why.

As has been mentioned, the 286 included the notable introduction of support for protected mode. The importance of this cannot be overstated. It set the groundwork for all modern operating systems: protecting different processes & the kernel from each other, and enabling support for virtual memory. Without memory protection, your PC would only be as reliable as the most buggy program you ran. And a bad program crash could even result in hard drive corruption.

To segment its product line-up with a more budget-friendly offering, Intel also introduced the 80386SL. This processor was almost identical to the 32-bit 80386, but was limited to 16-bit operations.
Click to expand...
I think you mean 386SX. Maybe 386SL was a follow-on laptop chip... Anyway, again "limited to 16-bit operations" makes it sound like a 16-bit CPU. It wasn't - just had half the memory bus width. It's not clear to readers that you're talking about memory bus operations, rather than instructions.

I think it's misleading to cite the i860's failure as owing to its lack of out-of-order execution. No x86 CPUs had that until much later, and I think it was even rare in the workstation space, where the i860 was somewhat successful.

I think the 486 was the first pipelined x86 CPU. The entry also could have mentioned the decoupling of the core & memory speeds. We had the whole DX2 series of chips, with half the memory bus frequency. I even had an AMD 486 DX4 100, with a 100 MHz core clock and 25 MHz memory. Around the time of 386/486, the transition from chip RAM to SIMMs also happened.

Pentium was Intel's first superscalar x86, meaning it could execute two instructions simultaneously. It was the first x86 CPU with support for multi-processor SMP, and dual-CPU Pentium mobo's weren't so uncommon. The PCI bus also appeared with the Pentium, but this was a chipset/mobo feature. If you don't count the Pentium MMX, I think the fastest Pentiums were just 200 MHz.

One interesting thing about the Pentium Pro is that its performance improvements owed almost entirely to its out-of-order execution (also, the FPU was probably faster). It was possible to hand-craft code that ran as fast on a normal Pentium as a Pro, but not usually the case.

The Pentium II entry shows the board, but no mention is made of the "Slot 1" socket they used.

Intel also introduced the new LGA 775 interface that featured support for DDR2 memory and improved the quad-pumped FSB.
Click to expand...
I had a socket-475 P4 Prescott with quad-pumped FSB. I think even some Northwoods had it.

The LGA 775 chipset introduced PCIe support.

Prescott was also Intel's first 64-bit x86 processor, allowing it to access more RAM.
Click to expand...
Prescott was rumored to have 64-bit support, but it was disabled. I actually wonder how much of its heat/performance problems came from all the baggage of 64-bit.

Pentium D was actually the first desktop Intel CPUs with x86-64 support enabled. Maybe a Prescott Xeon had it first... not sure.

Core was a separate CPU iteration than Core 2. It supported slightly different instructions than Core 2, and came about 1 year earlier. Intel confused the matter by using Duo vs. Solo to distinguish the number of CPU cores, and I don't recall any Core Duo's - only Core 2 Duo's. I don't consider Core 2's still viable, BTW. It's theoretically possible, but I pity any poor souls doing it.

Atom brought back Hyperthreading, BTW. That was supposed to compensate for lack of out-of-order, but the reality was that it didn't.

Nehalem was the first & only generation (so far) to support triple-channel memory, which was pretty bizarre. Also, PCIe was not on-die for at least some of the CPUs. It went over Quickpath to the chipset, which implemented the PCIe.

With Ivy Bridge, Intel switched from soldering desktop CPU's IHS to using heat sink compound. This, coupled with Ivy's smaller die size are common blamed for the Ivy's inferior overclockability.

Silvermont also dropped hyperthreading.

Conspicuously absent: Itanium & IA64. This important piece of Intel's history tells a lot about how AMD was able to beat them to 64-bit support, on the desktop, and possibly even explains PAE. It might even explain why Intel continued to push x86 into low-power and mobile, in spite of the inherent advantages of a RISC ISA like ARM. Basically, every time Intel has ever tried to deviate from x86, they've gotten burned. I think they might've learned that lesson too well.
 
I find it pretty incredible that an article in Tom's Hardware doesn't even mention the (then) King of Overclockers: the Celeron 300A (Mendocino core) that could overclock to 450 MHz (then the fastest clock speed available from Intel) and in some cases, beat the crap out of the Pentium II 450 which, while it enjoyed 4 times the amount of L2 cache (512 Kb VS. 128 Kb) ran it at 1/3rd its clock speed... When Mendocino ran it at full speed, making software that ran small data sets much faster on Celeron than on P2.
 
It's funny to call that a GPU, though. For me, GPU implies programmability, which we really didn't see until around 2000 (honorable mention for Rendition's Verite, which had like a 33 MHz integrated ARM core, in 1997).

No, the 387 was always in a separate socket. 486 DX was first to integrate it.

Yes, a comparison of the physical core sizes would've been nice. Also, a graphic showing the relative sizes, if all were manufactured on the same node.

Some graphs would be quite reavealing, showing such things as x86 CPUs by feature size, clock speed, memory bandwidth, TDP, or max core count, over time.

Some benchmarks, such as this, would've been the icing on the cake: http://www.tomshardware.com/charts/x86-core-performance-comparison/benchmarks,128.html No need to normalize by clock speed, I think.

Agreed.

Except that's wrong, as I and others have pointed out.

Well, that's a bit simplistic. For use in space, there's a natural advantage of larger feature size, since it's less sensitive to radiation. Also, space-hardened ICs lag for other reasons, mostly having to do with it being a smaller market that's far more risk adverse. There are other ways to skin this cat, such as by running multiple cores in lockstep, but I'm not up to date on what most current satellites and space probes use.
 
Hi there, thanks for putting all this together. Could you, please mention when we started to overclock the Intel CPUs and when these CPUs officially started to require active/passive cooling?

Otherwise, I like the new version of the THW web site: there is a lot less crap to be blocked by my script/add blocker but I greatly miss the print feature which could put on a single page a whole article. It was particularily usefull with the slideshow type articles and for reading articles offline on my phone/tablet. Could you please pass the message to add back the print feature?

Thanks & regards
 
Sadly we're missing any way of really comparing these processors to see how they improved. I assume there must be some benchmark that can be applied to all these processors (Whetstone?). This could then be shown as for single and multiple threads. It would be interesting to also have comparison figures for AMD processors, perhaps a Cray 1 and a modern smart phone.
 
These sideways scrolling photo-articles are really frustrating to read. Every time you move to the next photo the advert at the top of the screen causes the photo and text to jump up and down 2-3 times.
Also, a very minor point, it should say "See less detail" or "See fewer details".
 
My first "mainstream" computer was a 286 I bought from a friend. The second, a 486DX33 and a 486DX50 (for my wife). The Pentium 90 was next and on through several more until I switched to AMD. Intel's i5 2500K was my first CPU after the happy fling with AMD. IMHO, the 15 2500K was arguably Intel's best CPU ever when considering processing-power-increase. I still have my old 2500K lovingly resting on my desk in the original box. I've come now to the i7 6700K (a very worthy to the 2500K) and an i-7 5960X. Computer-wise, life just gets better and better.
 
I'll bet overclockers have been at it since day 1. I've heard stories of people overclocking their 386 DX 25's to 50 MHz! Imagine doing a 2x overclock, today!

My 386 DX 25 had a passive heatsink, but I think every 486 I saw had a heatsink/fan.

You know they make their money from ads, right? And more page views = more ads. But some of their articles (not this slide show, I think) are using a more mobile-friendly format.
 
@BIT_USER you are correct, I was thinking of my 486's motherboard when I was reading the 386 part and for some reason completely forgot the co-processor's socket on my 368 board. Sorry!
 


Lol, looks like Flying-Q has flew too fast and missed the AMD version. While we are here, adding a word about other x86 chip makers would make sense. Two decades ago we had Cyrix making these x86 compatible CPUs then we had VIA if I recall correctly and even Transmeta licensed the x86 technology, I think.

Cheers
 
VIA is still at it, I think. They acquired Cyrix, which is the other one I was going to mention (my dad once had one of their 387-compatible FPUs, which was supposedly much faster, and they made a strong Pentium competitor, as I recall).

There were even some Soviet x86 clones, from what I've seen. But this stuff is pretty obscure and I'm not really sure how many readers care. I mean, I thought the x86 retrospective was nostalgic, but I'm not sure I'd even read a history of the various clones.

I remember seeing this, years ago. It's worth a click, IMO: A Humble Russian Man Presents His CPU Collection (gizmodo.com)

You can probably find a lot more, here: CPU / chip collectors (cpu-world.com)
 
The 808x series of CPU's included the 8085, a much improved 8080, the 8086 as mentioned in the article, the 8087, the floating-point math coprocessor for the 8086 and the 8089, a coprocessor for I/O that wasn't widely used outside of mainframe terminals.
 


Yea, but it didn't seem like an issue here. It is was a direct clone. Of course I would have preferred a picture of Intel's 4004, but due to the age of these parts that isn't readily available.

A few notes to users on this thread. I appreciate your discussion over the history of Intel's CPUs and the picture stories. I wanted to take a moment to reply to some of your comments.

First, as some of you have noticed, there are details about essentially every processor on the list that isn't in this piece. This is due to the type of article this is. It is designed to be a brief overview of Intel's processors, give you a look at them, and address the key points and changes from one CPU to another. The bulk of these CPUs have been reviewed on Tom's Hardware before, and if you wish to read a more in-depth article over select processors, I encourage you to read those stories. Although I would like to have all of that detail in this one piece, it simply isn't practical. All of that information would fill a text book.

In regards to the mention of other competing processors that we could write about, I shouldn't go into too much depth here, but we do have more articles of this nature in the works, so you might see pieces over those products later this year.
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom