Ok, the opteron would be pretty cool for my quantum comp stuff, but then for everything else....I mean rendering would be ok, but games and shite prolly wouldn't like it much. damn.
According to the xbit labs game benchmarks i think it would be an excelent gaming machine, albeit expensive. Amd has expressed concern that the opteron might be looked at as a gaming machine though, since corporations might think if its good for games it cant be such a great server machine. and im not too sure about rendering scores, its way too early to tell final opteron performance with opteron since it was done in 32 bit os with 32 bit apps, and also there is no word as to if the programs even identify the opteron as having sse2 enabled, and im sure that [-peep-] trident chip had atleast a little something to do with it.
What game company would be in the right mind to write a 64-bit game. NEO-GEO? It is not cost affective. It takes a long time and $$$ to port software to a new platform. Just to sell an extra 2% seats? It would do pooly as a 32-bit game machine. If your lucky they'll put a ATI Rage Fury built into the MOBO and omit the AGP controller. AGP would mess up HyperTransport timing really good. Consumer price performance would be horrible too.
It's a server processor, let's see it compared to the Itanium2. Opteron is a fat bloated Athlon. Better used as a rice cooker.
First of all, it is not that hard to port a game to 64-bit. The makers of UT2k3 have already stated it is a rather harmless switch to port the code to 64 bit, only requiring little amounts of time.
There is absolutely no reason for the Opteron or athlon64 to handle 32 bit codes poorly, havent you been reading the reviews? Opteron and Athlon64 still retain all 32-bit functionality of the XP, there are no cutbacks, and the Athlon runs 32 bit code faster clock for clock.
Agp does not mess up Hypertransport at all. Agp will go straight through the Hypertransport line and requires no special optimizations, just that motherboards with no agp contain no agp tunneling chip. Opteron mainboards contain no agp right now because the version they are selling now is strict sevrer work, servers have no need for anything but onboard video. There are numberous chipsets already announced which contain agp ports and have special optimizations for CAD cards.
How do you about price/performance? Do you own an opteron or Athlon64? Do you have an opteron board with an agp slot? Do you know how Athlon64 will be priced?
And in regards to the heat comment, see paragraph above, nobody has given any straight information in regards to opteron temperatures. The only regards to opteron cooling we know is Newisys saying it is cool and asking customer to touch the cpu to feel that it is cool as to help win them over. ace has also hinted to the Opterons cool temperature in his Opteron article
You must understand that recompiling for x86-64 does not necessarily mean using 64-bit addressing and data, but more using the extended register previously 8 now 16 for x86 and xmm (sse). A simple recompile should get a significant boost.
They integrated a copper heat spreader, and the core is larger, thus a larger area to disperse heat. Also, running at 1.6gzh, I doubt it gets more than warm with good stock cooling.
<font color=red>*</font color=red><font color=white>*</font color=white><font color=blue>*</font color=blue>
... And I'm proud to be an American, where at least I know I'm free, and I won't forget the men who died, who gave that right to me.
That's what I expected also- boosts w/ a simple recompile. Let's see then, assuming I wait some more till opteron is maybe 2.6Ghz, and price remains at ~$750, 4 of those puppies will cost me $3000, then comes the 4GB RAM (dammit, I was hoping opteron would sport a spiffy 800Mhz DDR/QDR bus or better)- $1000 maybe. $1500 for the quadro and $500 for hard drives and sound card comes out to a nice $6000, and then we need mobo hehe.
Hilbert space is a big place.<P ID="edit"><FONT SIZE=-1><EM>Edited by flamethrower205 on 04/22/03 07:22 PM.</EM></FONT></P>
Still another thing to remember, dual setups are not the best when it comes to games, but hopefully more games will allow SMP support so we can actually have benchies with 4 opteron setups
The real mark of Opteron's success will be how much market-share AMD can claim in the coming months and years in the server segment of the market. I know Sun Microsystems has said that IA-64 will never be a success, and they're interested in X86-64 specifically because of its compatibility with 32-bit apps. Don't underestimate the impact of that one, single feature. Sure, maybe Opteron's benchmarks didn't blow everybody away, but they're still very good, and when you add to that the fact that 32-bit and/or 64-bit apps can run on the same platform, you end up with a product that may be very successful.
Those who live in glass houses shouldn't take showers.
And thats the server parts, if the chips can be adequately cooled by having an open air duct from a fan inches away from the side of the heatsink, imagine the little amount of cooling it will need when not in rack form or when athlon64 comes out
what im worried about is, sinse its based on teh Athlon ALOT (the review says its merely an AXp with ondie memory controller, the new 64bit and sse2 support, and more cache)
wont it be Mhz limited? because the Athlon has pretty much met its limit now...
<font color=red>*</font color=red><font color=white>*</font color=white><font color=blue>*</font color=blue>
... And I'm proud to be an American, where at least I know I'm free, and I won't forget the men who died, who gave that right to me.
But it also has SOI and the longer pipelines which indicate higher attainable clockspeed, but the exact numbers of how much that helps is not introduced right now
O, I understand about the 32 bit/ 64 bit. I dunno, I guess I'm just grouchy cause it won't be like going from my 200Mhz -> current rig. Eh, maybe I'll wait another 2 years for an upgrade.....
Like the first Wilamette, these chips are relatively new, their die is significantly different in what they feature (the on-die memory controller circuitry, the HT implementation, the 64-bit, etc.) and how they look, and at first they will not be that much scalable. But look over time where we have gotten since the 1.3GHZ Pentium 4.
--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden.
I thought that the P4 was initially released at 1.5GHz and reduced in speed afterwords...
<font color=blue><pre>I'm proud to be an American,
who served my country in the US Air Force,
to protect the rights of my fellow Americans,
to hold protests against others like me.</pre><p></font color=blue>
what im worried about is, sinse its based on teh Athlon ALOT (the review says its merely an AXp with ondie memory controller, the new 64bit and sse2 support, and more cache)
wont it be Mhz limited? because the Athlon has pretty much met its limit now...
No doubt the longer pipeline will help a bit. But all the extra circuitry for the on-die memory controller could also be limiting clock speed (lots of new potential critical paths). I suspect once Prescott is out on 90nm and starts ramping clock speeds, Hammer is gonna be in deep kim chee. Nonetheless, I gotta give props out to the engineers who can make a 1.8GHz processor perform as well as it does.
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.