T
I couldn't help smile, either.
Funny how Intel marketing loves cache sizes and bandwidth but forgets
latency.
George Macdonald said:Yes, as noted in the article and mentioned by me here this is a "very
un-Intel way".
With respect to recent postings by someone called "pigdos", to do with the
two hop AMD graphics connection, I'm suspicious that this Intel doc was
maybe his err, reference for the "question" he umm, "posed" and then argued
about. The question in my mind is whether Intel is sending (encouraging ?)
advocates (sales droids) to post nonsense/FUD like this in technical
forums. Over the past few weeks/months we've had quite a few similar
postings defending/pushing Intel's "technologies" from people we've never
heard from before... usually anonymously and usually a single shot deal.
Intel is getting desperate!
But Intel is laughing all the way to the bank. ;p
But Intel is laughing all the way to the bank. ;p
Some years ago (when I was a totally newbie) I read Intels FUD and was
sold on Intel, long story short... I was very disappointed with the
price/performance and stability, they fooled me once.
Some years ago, if you go back far enough, Intel was an honest,
enterprising, interesting and innovative company. The first sign I got
that things were going awry was when they scrapped "Solutions" magazine...
well before the days of electronic publishing.
Looks like Intel may get the last laugh on this one
Looks like Intel may get the last laugh on this one
http://www.anandtech.com/printarticle.aspx?i=2713
Looks good.... the AMD AM2 FX-62 2.8GHz 125W chip just dropped $500!
The said:Looks like Intel may get the last laugh on this one
http://www.anandtech.com/printarticle.aspx?i=2713
Yeah, that looks like the good thing (for consumers) is AMD's going to
price back into bargain levels again.
I seriously don't see how AMD is going to overturn a 40% performance
deficit of their fastest available against the Conroe within 6 months.
AM2 is just going to give more bandwidth and compatibility with DDR2.
So unless they have been keeping a canine ready in the bag waiting for
a moment like this to pounce forth.
Could be, but I don't think it'll get as bad as it did in the past...
Fastest CPU available / Price
09/20/1996
AMD K5 P-100 $82.00
CYRIX 6X86-P166 $256.00
INTEL PENTIUM - 200MHz $627.00
06/30/1997
AMD K6 PR2-233 $495.00
Cyrix 6x86-PR-200+ $103.00
Intel P-ll 266/512k(Box ) $865.00
Intel P-Pro 200Mhz-512k $1055.00
01/02/1998
AMD-K6-233 Box $250.00
CYRIX 6x86MX-PR200+ $109.00
Intel P-II 300/512 Box $790.00
10/30/1998
AMD-K62-350/100MHz $147.00
CYRIX MII-300GP/66MHz $76.00
Intel P-II 450/512 Box $635.00
03/08/1999
AMD-K62-400/100MHz OEM $174.00
CYRIX MII-333GP/83MHz $57.00
Intel P-III 500/512 Box $745.00
I did percentages for hexus's and 40% seems a bit of a stretch, maybe
Intel is including synthetic benchmarks in that 40% figure?
http://www.hexus.net/content/item.php?item=4843&page=1
AMD Athlon FX-60 @ 2.8GHz vs. Intel Conroe @ 2.67GHz
iTunes 6.0.1.3, Intel wins by 14.71%.
Windows Media Video 9, Intel Wins by 12%.
XMPEG, Intel Wins by 12%.
- FPS - Games @ 1024*768 / Medium settings.
Unreal Tourn 04 Bot Patch, Intel Wins by 17.11%.
Quake 4, Intel Wins by 22.84%.
Half Life Source, Intel Wins by 24.27%.
FEAR (CPU Max, Graphics – Medium), Intel Wins by 45.60%.
Whats AMD going to get out of DDR2, 5% tops?
Ed
Ed said:I did percentages for hexus's and 40% seems a bit of a stretch, maybe
Intel is including synthetic benchmarks in that 40% figure?
The interesting thing is that the CoreDuo benchmarks show a big jump inhttp://www.hexus.net/content/item.php?item=4843&page=1
AMD Athlon FX-60 @ 2.8GHz vs. Intel Conroe @ 2.67GHz
iTunes 6.0.1.3, Intel wins by 14.71%.
Windows Media Video 9, Intel Wins by 12%.
XMPEG, Intel Wins by 12%.
- FPS - Games @ 1024*768 / Medium settings.
Unreal Tourn 04 Bot Patch, Intel Wins by 17.11%.
Quake 4, Intel Wins by 22.84%.
Half Life Source, Intel Wins by 24.27%.
FEAR (CPU Max, Graphics – Medium), Intel Wins by 45.60%.
Whats AMD going to get out of DDR2, 5% tops?
I'm comparitively young. The earliest I remember is the 8086....
presumably you're talking about some time before that?
Aren't all games synthetic benchmarks? ;-)
The interesting thing is that the CoreDuo benchmarks show a big jump in
FP performance, that seems to be good for some games used as benchmarks.
Get them at the SPEC site, I followed a link here, didn't benchmark it.George said:Where have you seen FP performance benchmarks? For the game tests run on
Intel's supplied systems, I think there's far too much other stuff going on
to conclude that the "40%" is mainly due to FP gains... though one would
expect *some* FP gain from a single-cycle SSE execution unit.