Barton = new K6-3? Athlon 64=new Athlon Classic?

[Spitfire_x86]

Batron is as expensive as like K6-3 to manufacture. Due to lack of clock speed, it sucks at 2800+ and 3000+ PR's. AMD is going to launch Athlon 64 in September (if they don't push it back again), Athlon classic was launched at similar time of 1999. Athlon "Classic" was not much future proof without on-die L2 cache and .25 micron manufacturing process. Athlon 64 looks similar with DDR333 and .13 micron. There was hard to find mobo/chipset for Athlon "Classic". Now it's hard to find MS-OS for x86-64. AMD was in trouble with K6-3 before launching Athlon, now AMD is in trouble with Barton.

Lots of similarities, isn't it strange?

<b> "You can put lipstick on a pig, but hey, it's still a pig!" - RobD </b>

 

[flamethrower205]

But that means within a few months of the hammer amd will begin to be king again among comp nerds

"If everything seems under control, you're just not going fast enough."
- Mario Andretti

 

[imgod2u]

Not exactly, if you recall, the P3 CuMine gave the Athlon quite a run for its money before the P4 was released. Intel's not planning on releasing a new core until 2005-2006 with Nehalem. Perhaps that core design will be moved up and released prematurely like the P7 core was and hence be hammered (pun intended) by the K8. However, this is all speculation. Processor releases and performance is not something that you can just predict based on past events. Semiconductor corporations and architects are not automatons that just exhibit the same behavior repeatedly over and over again.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.

 

[flamethrower205]

but nerds do!

"If everything seems under control, you're just not going fast enough."
- Mario Andretti

 

[baldurga]

Well, I think this is not too strange. Simply because both are the last models of one platafrom (K6-III for socket 7, Barton for socket A). Also the same "upgrade path" has been used, more L2 (or L2 at all when talking about K6). Why? Because it doesn't need to much effort to design, at least compared with a new core or a major rearrangement that allow higher clock speeds.

Also I predict that in the future top model Bartons (3200+ or higher if they are ever produced) will be hard to find. Some people will hold it's socket A systems and upgrade to Barton as a cheap alternative. It will be like the final step before changing mobo + CPU (minimum). Obviously if it's worth the cost.

In fact is the way I'm feeling with my K6-2+ system!


Still looking for a <b>good online retailer</b> in Spain :frown:

 

[POPEGOLDX]

whats good is that now SSE2 is finally getting alot of support

just in time

FOR HAMMER

its ironic

 

[juin]

expensive like A64 will be release with 1MB of L2 cache even T-bred is hard to produce at high speed 2600+ and cost more that intel conter-part.Opteron face the reverse compare to A64 or T-bred as it will be release at moderate speed close to others 64 bit chip putting less pressure to the manufacturing cost.

Performance wise

A64 will have a terrible time again Presscott as it becoming hard to believe that a 2.6 GHZ can be produce on a 0.13CU SOI.Single channel DDR 2700 that leave it with about 1.3 GB/S bandwith it good for now for a 2GHZ CPU at 2.5 it will lose scaling ability.

Opteron

Will crush many true 64 bit cpu on webserver but the market is consolidation.On any big FPU apps like dupont chemical R&D opteron is not a option as is FPU is about 1/2 of the power of a Madison or about 30/50% slower that others 64 bit CPU about the same speed that a Xeon 3.0HT GHZ (base that the opteron run at 2GHZ).All apps as to be recompile as no unix sytemes will be aviable to it no .net windows for now (this may change) a beta linux that no OEM have validate it.That leave only a linux with a slow compiler not much tool nothing as been proven also it may not be faster that a 32 conter part a intel compiler increase perf up to 20 % in spec benchmark compare to others compiler no surprise AMD have use intel compiler to run there spec benchmark for the opteron.


Software

A64 is the key to opteron strategy as A64 will bring volume to X86-64 leaving ISV with a much bigger market unlike what we see under others 64 bit chip.With a 64 bit windows mainstream it may win the desktop market with a good marketing stratergy.


Software issue

ALL driver as to be recomplie test and validate for the new os if not windows will be force to run at 32 or be non-compatible Nvidia have release driver for X86-64 but what about sound card printer network card.

The issue is no driver will be made before there a mainstrean windows 64 bit for A64 and windows will not release a mainstream windows 64 bit with no volume.

a windows 2003 advanced server is a big posibility not many driver need most will come from the systemes builder big performance good price/peformance as MS want to kill about anything in the 64 bit OS market

Just next to the lab and the bunker you will find the marketing departement.

 

[Crashman]

It's a lot like Cyrix with the MII and the Samual. We saw what happened to Cyrix and who ended up with the Samual.

<font color=blue>There are no stupid questions, only stupid people doling out faulty information based upon rumors, myths, and poor logic!</font color=blue>