Socket 3 Overdrive info required

Post new topic   Reply to topic    CPU-World.com forums Forum Index -> Information Requests (Collectible Chips only)
View previous topic :: View next topic  
Author Message
Neon_WA



Joined: 08 Nov 2008
Posts: 7146
Location: Margaret River, West Australia

PostPosted: Sat Aug 21, 2010 10:35 pm    Post subject: Socket 3 Overdrive info required Reply with quote

A while ago got to thinking about core steppings with the FDIV bug. It is known the these are affected

P5 B1
P5 C1
P54C B1
P54C B3

got me thinking about early overdrives.. especially SZ953 with a core stepping of P5T B1 and whether any were affected by the FDIV bug.

As the the SZ953 was not released until week3 or 4 1995 it appeared they were produced to late to be affected.

Looking for some other info I came across this in one of the Pentium updates
Quote:
24. Errors in Least Significant Bit of an FMUL Instruction Result with
Specific Operands

PROBLEM: The result of multiplying two operands, one of which is typically close to infinity or close to the
smallest representable normal number, may be incorrect in the Least Significant Bit (LSB) and may result in
flags that are incorrectly set. This problem is limited to a small percentage of the first production units of the
63MHz Pentium OverDrive processor (specification number SZ953, Ver#: 1.0).

All precisions: single, double and extended are affected. All rounding modes: nearest, up, down and chop
are affected. Only specific operand pairs result in errors.

IMPLICATION: Only a fraction of the Pentium OverDrive processors may produce errors in the least
significant bit of the result when using the FMUL instruction with one of these specific input operands. Any of
the exception flag bits in the FPU status word may also be incorrect.

WORKAROUND: None identified. Intel includes software with the Pentium OverDrive processor that can
detect this erratum. Please see the following Status paragraph.

STATUS: The Pentium OverDrive processors currently in production are free from this defect. Only a fraction
of the first production units at 63-MHz marked SZ953 may exhibit the defect. Owners of Pentium OverDrive
processors should run the diagnostics on the disk supplied with the Pentium OverDrive processor. The
diagnostic program tests the processor and will notify the user of pass/fail information. Should the processor
fail, the diagnostic program will report that “The Floating Point Conformance Test has Failed” and will wait for
a keystroke before continuing. Owners of the Pentium OverDrive processor should contact the Intel support
line if there are any questions.


so what are they really saying here Confused

_________________
There are 10 types of people in this world:
those who understand binary and those who don't. ~Author Unknown
http://www.x86-guide.net/Neon-WA/en/collection.html
Back to top
View user's profile Send private message [ Hidden ] MSN Messenger
CPUShack



Joined: 16 Jun 2003
Posts: 34259
Location: State of Jefferson, USA

PostPosted: Sat Aug 21, 2010 11:49 pm    Post subject: Reply with quote

just a normal bug
certain circumstances could lead to a slightly incorrect (LSB) of a floating point multiply instruction

_________________
New for 2025! The CPU Shack has a co-processor!

Visit The CPU Shack of microprocessor history and information.
Back to top
View user's profile Send private message   Visit poster's website AIM Address MSN Messenger
Display posts from previous:   
Post new topic   Reply to topic    CPU-World.com forums Forum Index -> Information Requests (Collectible Chips only) All times are GMT - 5 Hours
Page 1 of 1
Jump to:  
You can post new topics in this forum
You can reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You cannot download files in this forum

Powered by phpBB © 2001 phpBB Group