| 1 |
Nikos Chantziaras <realnc@×××××.de> posted h0qlib$fs8$1@×××××××××.org, |
| 2 |
excerpted below, on Thu, 11 Jun 2009 13:16:45 +0300: |
| 3 |
|
| 4 |
> Intel re-introduced HyperThreading with Core i7 a while back. |
| 5 |
|
| 6 |
Well, if you'd continued your quote... |
| 7 |
|
| 8 |
>> That's where we are today. On a modern CPU, hyper-threading provides |
| 9 |
>> very little real performance gain, one that actually may be a loss if |
| 10 |
>> one considers what else that same transistor budget could have |
| 11 |
>> otherwise been used for, but the market, once programmed for it, now |
| 12 |
>> continues to demand it, so Intel continues to provide it. |
| 13 |
|
| 14 |
IOW, when I said |
| 15 |
|
| 16 |
>> Happily Intel has moved beyond that stage now, |
| 17 |
|
| 18 |
"That stage" referred to the old netburst architecture where hyper- |
| 19 |
threading originally became popular, not hyper-threading itself, which |
| 20 |
they continue to offer, now more due to demand than for any technical |
| 21 |
reason. |
| 22 |
|
| 23 |
-- |
| 24 |
Duncan - List replies preferred. No HTML msgs. |
| 25 |
"Every nonfree program has a lord, a master -- |
| 26 |
and if you use the program, he is your master." Richard Stallman |
Archives