The History Of Intel CPUs (Archive)

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
I miss lot of important data from descriptions:

- 4004: It was a 4 bit wide CPU
- 80286: firts x86 CPU with protected mode support and non multiplexed data/address bus
- 80386: "more RAM almost always translated into a performance increase" No, the new architecture were the cause of performance increase not the amount of memory. Also you forgot to mention the most important things: virtual mode and flat memory mode.
- 80386SL: "was limited to 16-bit operations" It is not true. The SL supported 16 and 32 bit mode as well, but had 24 bit address and 16bit data bus same as 386SX. So it supported only 16 MB physical and 4GB virtual memory. The most important to mention here is the energy saving function and the integrated cache controller.

This is the point where I simply do not read forward as the quality is way below the quality I expect from Tom's hardware.
 


- 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.
 
My first computer ran a Intel 486/66 DX2 with 8MB ram and a 1MB Western Digital "Paradise" GPU. I upgraded it to a 486/133, but by then the Pentium had come out and the 486/133 couldn't even keep up with a Pentium 90mhz.
Second computer was a AMD K6II at 300mhz and my last computer under 1ghz.
 
Might add what CISC and RISC are, and what their pro's/con's are just for better understanding. And you have typos on the 8080/8086 and 80186/80188 parts.

Also you don't mention the 386DX with integrated FPU, but you list the 486SX as a budget model 486DX. Was the 386DX FPU not as integrated? I.e. both on same silicone but two dies within?

You should add a scale graphic with the size of transistors from beginning to current, that always amazes me more than anything, the insane size decrease of those things!
 
Typo page 19 second-last paragraph. ...made 'cooling it' more difficult! This article hasn't been proof-read very well, has it?
 
386 description is incorrect, it was split into 3 lines:
DX: 32-bit wide data bus. Max clock speed offered by Intel was 33 MHz. Faster, but required FPM DRAM to be delivered by sets of 4 identical sticks (FPM DRAM sticks worked in 8-bit+1-bit parity), causing most mobos to be sold with 8 stick sockets. Servers could be sold with the first 32-bit FPM DRAM sockets, though they cost an arm and a leg.
SX: 16-bit wide data bus. Max clock speed offered by Intel was 25 MHz. It could make use of FPM DRAM by sets of 2 sticks.
SL: essentially a SX with support for the HLT instruction (processor powers down when idle instead of running NOOP). Used on the first "modern" laptops.
 
As for 386SD/SL being unable to run 32-bit applications, that's not really true: it could run 32-bit instructions in a single clock cycle but it had to split every memory call in 2; as long as the software run was 16-bit, the 386DX would run as slowly as the 386SX/SL (it couldn't "glue" together 2 16-bit memory operations), but making use of extended mode (like Windows 3.1's swap file in direct disk access) would boost the 386DX a lot.
 
Lots of data skipped or erroneous: The 4004, was Intel's first. (and mentioned)... Okay the 4040 came next as an upgraded 4004. Okay then came the 8008, the first 8-bit processor for Intel. Now came the 8080, and like the 4040 was an improvement to the 4004, the 8080 was an improvement to the 8008, okay. Now.. What I don't see mentioned is these chips all required multiple voltages to run. A few Intel people seen room for improvement and left to start Zilog (the Z80, which in itself was very popular, is heavily based on the 8080, but only needed a single voltage to operate.) Intel countered quickly with the 8085.

Now, the 8086 was indeed the first 16-bit CPU from Intel, and the 8088 was pretty close to it. The only real differences between these two CPUs is the 8088 made use simpler and cheaper than the 8086 as it dropped down to an 8-bit data bus. (No other real changes.) Which came first? The 8086 in '78 and the 8088 in '79. The 80186 and 80188 were not very popular and were buggy. the [80]286 followed and was a hit as it didn't have the same bugs as the 186 and backed off slightly on problematic integrations, keeping the ones that worked and fixing the ones they could.

Now some clarifications. the 386 was a 32-bit CPU. The 386SX was the same CPU but with a pared down 16-bit data bus.(As mentioned already the 386SL was closer to the 286 bus and had features for power saving.) Shortly after the 386 took on the DX suffix to be called a 386DX. ALL Intel CPUs up through the 386 series relied on another chip for floating point from the 8086/8088 through the 386. It was similarly numbered, but the last digit was a 7.

Okay, the 486 included the FPU and had other improvements, unless there were flaws in the 486's FPU, which to save on losses, became the 486SX (FPU disabled) while the 486DX had a fully functional FPU on board. The 487 was just a re-pinned 486DX.

Then, alas, came the Pentium, so named as they couldn't trademark a CPU name that was just a series of numbers (the 486 family was still technically called a 80486, and Intel's competition was using the numbers to identify their compatible CPUs. Intel didn't like it. Tried a law suit and lost, then changed the naming convention to a word name.)
 
Very cool, but it would have been nice to see the 80386 given more detail. After all, it was still in heavy use just two years ago. A little organization called NASA still used the chip in all the space shuttles. When I asked a NASA scientist why, he replied, "It just doesn't fail. It does what we need it to do perfectly."
 
There are a lot of little oversights, but I agree with Stairmand that the lack of any mention of StrongArm and XScale seems like a big one. Intel manufactured and sold hundreds of millions of them from 1997 to 2009.
 
"The 80286 was released the same year as the 80186 and had nearly identical features"

Huh? You forgot to mention protected mode making its first appearance on Intel CPUs. While 286 protected mode was rather retarded with its 8086-esque segmentation and whatnot, it was a testing ground for that of 386 which is still used (by 32-bit OSes such as Windows 10 32-bit) nowadays.
 
Also, there's a typo in the very last paragraph of Skylake section:

"The highest-end model is known as Iris Pro Graphics 580 and it is used used inside of Skylake processors ..."

used used (twice)
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom