• Home
• Reviews
• News
• Downloads
• CNET TV

About me

• Member since: February 11, 2006

My posting summary

1 to 8 of 8

My comments

• Yes, it is
  You have to understand that Cell is a design that is two issue, in-order, and lacks branch prediction. This is a far, FAR more simpler design compared to 970. This is why the clock rate is so high.
  
  For single-threaded general computing, Cell is a huge step backwards. It may have the same ISA as PowerPC, but inside, it is entirely different.

  February 28, 2006

  0 replies

• Yes, it is
  You have to understand that Cell is a design that is two issue, in-order, and lacks branch prediction. This is a far, FAR more simpler design compared to 970. This is why the clock rate is so high.
  
  For single-threaded general computing, Cell is a huge step backwards. It may have the same ISA as PowerPC, but inside, it is entirely different.

  February 28, 2006

  0 replies

• Embarrassing?
  If it were embarrassing, Intel wouldn't have told the media about it.
  
  Clearly, it's something else.

  February 11, 2006

  0 replies

• Ha.
  AMD will ship quad-core in mid-2006? Ha.
  
  Even their own roadmap says 2007.
  
  http://www.amdcompare.com/techoutlook/
  
  But hey, who am I to argue with a Droid?

  February 10, 2006

  0 replies

• Per Processor
  Until MS updates their page to say "per-die," I'm not going to trust some random unverified info on someone's blog.
  
  http://www.microsoft.com/licensing/highlights/multicore.mspx

  February 10, 2006

  1 reply

• Nope
  Hell, if AMD didn't show it to the public, what's the point?
  
  You might as well have said AMD showed a 12-core processor 5 years ago.

  February 10, 2006

  1 reply

• Nope.
  Here comes the cluebat.
  
  MS charges on per-processor basis, not per-core.

  February 10, 2006

  1 reply

• .
  "Microsoft counts each die as a CPU when computing license fee."
  
  LOL. Here comes the cluebat.
  
  MS charges on per-processor basis, not per-core.

  February 10, 2006

  0 replies