Re: <Can we drop this now?>
No Elmer,
We can not drop this when you post messages that ask for information, then deny the picture before your eyes, contradict your previous posts, then claim your criteria for information has changed. You, as a member of Silicon Investor have a duty to post honest and accurate information. If you post a challenge to other members who then meet it, it is a besmirch upon your honor when you deny the facts or change the basis of your challenge.
Your challenge:
To: Brian Hutcheson (5971 )
From: Elmer
Tuesday, May 5 1998 12:50AM ET
Reply # 5977 of 6042
If a Celeron 266mhz = a K6-200 then a Celeron 333mhz most likely
will be similar to a K6-266mhz .>
Brian, I posted a bunch of links yesterday showing the Celeron
performed better that the claims posted here. How about backing up
your notion that the 266mhz Celeron = a K6-200. You may be right but
we'll never know if all you do is talk and never prove. I posted 11
benchmarks. How about posting 2?
EP
Then like an oaf, you chose to taunt when he did not have time to search:
To: Brian Hutcheson (6002 )
From: Elmer
Wednesday, May 6 1998 12:33AM ET
Reply # of 6042
<PS sorry no benchmarks >
That's OK, I knew you couldn't find any. That's why I asked, so you
could see that you couldn't find any.
EP
Upon receiving numerous posts to the contrary, including my post:
A quick search on Alta Vista of "celeron benchmark" returns on the
first 3 pages:
Graph showing Celeron266 Win95 Bus. Winstone performance is 91%
of PMMX233 (e.g. ~= PMMX200).
zdnet.de
"In fact, the Celeron could barely keep pace with PCs running on Intel's
PMMX-200, which is used in systems that cost as little as $799."
pcworld.com
You change your criteria to performing worse:
To: Kevin K. Spurway (6008 )
From: Elmer
Wednesday, May 6 1998 10:56AM ET
Reply # of 6042
<I guess this one can be included in addition to the three benchmarks you yourself posted which show Celeron's lack of performance.>
On the contrary. Brian claimed the Celeron performed worse than a 200mhz Pentium. This goes on the list with the other 9 showing Celeron equal to or faster than a 233mhz Pentium MMX. I'm glad you guys keep making my point for me. You (AMDites) keep claiming poor performance for the Celeron and then you can't come up with anything
to back up your claims. I even gave you a headstart.
Don't worry, profits are just around the corner.
EP
To: Ed Sammons (6029 )
From: Elmer
Friday, May 8 1998 12:04AM ET
Reply # of 6042
<However, there is no man so blind as one who refuses to see.>
Ed I tried twice to drop this but if you insist...perhaps you have trouble
with the English language. I asked someone for a benchmark showing
the Celeron performing worse than a Pentium 200, as that person
claimed. He was unable or unwilling to produce one. So be it. You
posted one showing it to be exactly equal to a 233mhz Pentium MMX.
In our language, "exactly equal to a 233mhz Pentium MMX" does not
mean worse than a 200mhz Pentium. You posted another showing it to
be 91% of a 233mhz Pentium MMX. A 200mhz Pentium runs at 85%
the speed of a 233mhz Pentium so this doesn not constitute a
benchmark showing it running worse than a 200mhz P. You posted
another quote saying "it could barely keep up with a 200mhz Pentium.
In our language "it could barely keep up" means it could keep up, just
barely. It does not mean that it was worse. All I wanted to do was see if
that person could back up what they were claiming. I expected them to
be able to but apparently they can't.
Can we drop this now?
EP
Then in a flash of brilliance, you repeat a link tomshardware.com which shows that the Celeron266 is 1% slower than the K6-200 (in other words, C266 ~= K6-200), proving Brians's original claim and the premise of your challenge, that the 266mhz Celeron = a K6-200. This is a cpu that is barely even in production anymore, from a second-rate (from the spirit of your posts) chipmaker. What does this say when Intels's newest and "low-cost" CPU can't beat a CPU that cost 1/2 and is being phased out of production?
I spent two minutes on a search of "celeron benchmark" and in the first 30 hit,s of over 200,000 hits, were the pages I posted. I find it difficult to believe there will not be more such pages, if the search were continued. You state "You posted one showing it to be exactly equal to a 233mhz Pentium MMX." Elmer, read the graph. That is a WindowsNT benchmark. Celeron is for the home user. They will Windows95, not WinNT (for instance, you will have a slight problems running DOS games). The graph shows that for that Win95 benchmark, C266 is 9% slower than PMMX233.
I make no claim to being a computer guru, but your statement, "A 200mhz Pentium runs at 85% the speed of a 233mhz Pentium so this doesn not constitute a benchmark showing it running worse than a 200mhz P." shows a lack of rudimentary understanding of computer architecture. Computer system speed does does not scale linearly with CPU frequency (200/233/ = .858). This only occurs when the other component speeds in the system also scale the same. Even Intel claims in the iCOMP benchmark that PMMX200=182 and PMMX233=203 => .897 or 10%). However, this is an idealized benchmark, since no disk, video, or other peripheral access is taken into account, so no useful work is performed. What that benchmark really shows is the CPU is stalled about 1/3 of the time waiting on L1 cache misses. A good rule of thumb for today's PC and applications is the CPU is busy 1/3 - 1/2 the time, memory and peripherals have the bus 1/2 - 2/3 time. In other words, you can expect, depending on applications, 5-7% speedup going from PMMX200 to PMMX233, as evidenced by the 5% speedup shown by your beloved graph at tomshardware.com. That is well below the 9% dropoff going from the PMMX233 to the C266, or below the PMMX200 for that benchmark.
You open your mouth freely, yet you speak with no authority. You come to this tread like an unwelcome house guest, partake of its hospitality, disrespect its members, post opinions of questionable accuracy, and then, like a boor, deny or change the subject when other members turn hostile to you.
Elmer, be a man. Choose your words with consideration, then stick by them. Admit with humility when you you are wrong, and graciously accept praise when others acknowledge you you are right.
SI keeps a history of posts, and inaccuracies may come back to bite the poster. I suggest you read George Orwell's 1984. In it he shows how to successfully change history. If you do not like Orwell, I suggest, in its place, a book by Miss Manners. |
