SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Intel Corporation (INTC)
INTC 48.72+3.0%3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Tony Viola who wrote (40039)11/12/1997 3:08:00 AM
From: Petz  Read Replies (2) of 186894
 
Tony, your post is full of garbage. I said:

"If such a bug were found on a mainframe, you are right, the
microcode would be patched by replacing a PROM."

You said,
WRONGO!Microcode nowadays is stored in RAM in the big machines. In fact, in the high majority of cases, it is "concurrently maintainable", so that the customer does not even have to take an outage to install new microcode. The only reason I make a deal out of this is if you don't know what you're talking about re mainframes, why should I take credence in your knowledge of CPU chips?

1. I was talking about older mainframes since I found absolutely no evidence that modern mainframes have any undisclosed opcodes which cause an uninterruptible HALT in user mode.
2. You had better show some documentation for your claim about IBM System 390!

Then Tony objected to my claim that Pentium microcode is embedded in the silicon and cannot be changed. Here's what he said:

Are you sure there is no microcode RAM in Intel CPU's? How about being able to add MMX instructions to existing chips. How dey do dat? Also, Intel announced on chip control store that could be used to fix bugs from "outside". Sounds like we have a good use for that control store now.

1. I am sure about classic Pentiums and MMX Pentiums, less sure about the Pro or P2.

2. Tell me, how do you add MMX instructions to a Pentium classic?

3. If on-chip control store can fix bugs, and Intel could just create a floppy disk that downloads new control store code, why don't they do this for the 41 Pentium II errata contained in
ftp://download.intel.com/design/pentiumii/specupdt/24333706.pdf
In fact, they only corrected 3 bugs going from Stepping A to Stepping B, but you claim they should be able to fix most of them without even changing the mask.

4. Think about it. If the actual microcode is in RAM, where does it come from on power up? It would either have to come from a PROM or Flash RAM embedded in the processor chip, and BTW, flash is too slow to use directly. If its a flash RAM, how do you download to the flash RAM? That could only come from the motherboard, right. What motherboards do you know about that have this download feature?

5.There is a slight chance that there is a flash RAM and a RAM for microcode, but the only way to load the flash RAM with anything new is to remove the CPU from the MB and plug it into a special Intel downloader fixture. This would probably cost Intel more than just shipping them new CPU's.

Petz
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext