41 Replies. 3 pages. Viewing page 1.
< Newer [ 1 2 3 ] Older >
 |
| 41. |
Re: Pentium 64 soon? |
Feb 10, 2004, 10:56 |
m00t |
|
|
Intel already has a 64 bit chip: Itanium.
So far it's not getting the growth Intel had hoped. It's performance has been lackluster (not enough software compiled for it, so it has to run most of it in emulated 32bit)
They're also planning on going the x86-64 route that AMD did though it appears their chip will not be compatible with AMDs meaning there might be 2 versions of WinXP64 in the future.
|
|
|
|
|
|
|
|
| |
|
| 40. |
Is English not your first language? |
Feb 9, 2004, 18:12 |
Paully |
|
|
Will you for crying out loud stop blaming the drivers!
Didn't you read that 64-Bit UT 2004 reputedly flies along?
How can there be anything wrong with the 64-bit drivers then?
That's why I'm sure the problem is the 32-64 bit thunking.
Just like 16-32 bit thunking to run 16-bit apps using Windows On Windows (wowexec) on 32-bit windows.
|
|
|
|
|
|
|
|
| |
|
| 39. |
read the shit people |
Feb 9, 2004, 10:56 |
space captain |
|
|
in case you never read my original post which was #5, here is a quote:
"it does spell lots of work for backwards compatibility tho"
_____________________________________________
Give me slack. Or kill me. |
|
|
|
|
|
|
|
|
______________________________________________
"When the bomb drops it'll be a bank holiday
Everybody happy in their tents and caravans
Everybody happy in their ignorance and apathy
No one realizes until the television breaks down..."
- SUBHUMANS |
|
|
|
|
| |
|
| 38. |
Re: No subject |
Feb 9, 2004, 10:14 |
DeaderMeat |
|
|
"this is retarded... you have to test 64 bit hardware on CODE BUILT FOR 64 bit hardware! not yesterdays games, tommorows games"
I think it's a fairly pointless article, but not for the reasons you're suggesting.
Once there's a *released* 64bit version of WinXP for A64 then eventually there will be gamers using this version of windows. For a long time there will still be far more 32bit users, and most games will be 32bit binaries (there may be some developers who provide both 32bit and 64bit binaries but I expect that this will be the minority to begin with).
Therefore, running 32bit games on a 64bit architecture with a 64bit version of Windows will be a common occurance in the future and it's something worth testing. However, I don't think it's really worth testing with a preview version of XP64 with unoptimised drivers.
I suspect (I don't know) that MS will ship 32bit Windows to home users for a long time to come, until both AMD and Intel are only building 64bit CPUs, and that 64bit will have the same sort of price premium as XP professional. So 64bit PC gaming will take a while to come to the masses.
|
|
|
|
|
|
|
|
| |
|
| 37. |
Pentium 64 soon? |
Feb 9, 2004, 09:53 |
Tumbler |
|
|
Will Intel be making a 64 bit chip soon? I'm not sure why game companies would want to program a 64 bit game when the only ppl who could play it would be folks with these AMD processors. Mabye someone will make a 64 bit benchmark demo or something like that.
This is the first time I'm hearing about 64 bit anything, is their more 64 products on the horizon?
|
|
|
|
|
|
|
|
|
VGfive.com - Game Trading site (Steam codes too!)
Kickstarter "Game Developer"! |
|
|
|
|
| |
|
| 36. |
Re: No subject |
Feb 9, 2004, 09:05 |
space captain |
|
|
""And AMD fans did nothing but say well "its NOW we are concerned about not the FUTURE". Kind of a change in decision now that its your way. ""
You fuckin idiots, I could care less about AMD CPUs, I use Pentium because my DAW requires it. I WAS SPEAKING IN GENERAL, and therefore I will repeat:
this is retarded... you have to test 64 bit hardware on CODE BUILT FOR 64 bit hardware! not yesterdays games, tommorows games
_____________________________________________
Give me slack. Or kill me. |
|
|
|
|
|
|
|
|
______________________________________________
"When the bomb drops it'll be a bank holiday
Everybody happy in their tents and caravans
Everybody happy in their ignorance and apathy
No one realizes until the television breaks down..."
- SUBHUMANS |
|
|
|
|
| |
|
| 35. |
Re: No subject |
Feb 9, 2004, 08:43 |
Nexus |
|
|
strangelove:
I think in the context of the discussion, he was talking about 32 bit Max in 64 bit OS, which, on the whole, won't benefit at all from the CPU being 64 bit capable, since it would just be running in Compatability mode (I have a funny feeling the speedup in the Mpeg software might have been the OS running faster without being bogged down by bad drivers like the games ones could have been, then passing the extra CPU time back to the application).
I don't doubt that 3d render engines will be making the transition to 64 bit engines at some point in the nearer future, while maintaining their 32 bit code for legacy support of older render farms and users. That'll be the interesting 64 bit benchmark for them, not if the 32 bit code runs faster in the 64 bit OS.
The Narcis:
quick repeat for you...
I'd say it's more than likely the slowdown is more of a reflection on XP64 than the A64, and XP64 is a preview version with who-knows-what state of drivers. It's already proven that the A64 can run 32 bit code quickly, and I'm not at all convinced that the differences between compatability and protected modes for the A64 are currently more significant than the OS and driver status, in which case the evaluation becomes almost entirely spurious because it can be changed completely by later driver or OS improvements.
This comment was edited on Feb 9, 09:01. |
|
|
|
|
|
|
|
| |
|
| 34. |
Re: No subject |
Feb 9, 2004, 01:10 |
m00t |
|
|
"Then again, AMD is notorious for releasing processors to the press at clock speeds they can't make yet!"
You mean like Intel did with the P3?
Another advantage (at least for the Athlon64) More registers!
Once compilers are optimized to take full advantage of them, they will be a massive boon to software in general.
|
|
|
|
|
|
|
|
| |
|
| 33. |
No subject |
Feb 9, 2004, 00:24 |
TheNarcis |
|
|
this is retarded... you have to test 64 bit hardware on CODE BUILT FOR 64 bit hardware! not yesterdays games, tommorows games
I heard this same argument come from Intel fans when AMD fans would bash the Pentium 4 on its arrival. Seems like the pot is calling the kettle black. 
And AMD fans did nothing but say well "its NOW we are concerned about not the FUTURE". Kind of a change in decision now that its your way. 
|
|
|
|
|
|
|
|
| |
|
| 32. |
No subject |
Feb 8, 2004, 23:34 |
Nonicknameforme |
|
|
Now this is simply incorrect. 64 bit refers to instruction length, and while higher memory addressing is possible with a longer instruction, that is certainly neither the only, nor the main advantage to 64 bit processing. Currently, with a maximum 32 bit address, (any byte-addressable memory) the maximum amount of memory that a 32 bit machine can access is 4 gigs, much more than we currently need/use anyway. A 64 bit processor will allow 16 exabytes (16,000,000,000 gigs) of memory... clearly enough for a while, but this is hardly the main advantage of 64 bit processing.
good god damn point... however, its kinda a perspective thing as well. I can't dissagree, but I don't like presenting those point's because stupid people assume I'm actually backing up there points.
Hey, i'm the logical sledge hammer of blues news, i'm never always right, but at least I don't beat around the bush or get offended. =P
btw, technically you need 3072 times the code to fully take advantage of a 64bit processors code pipeline... there has to be a point where the gains in detail aren't worth the increase in effort.
This comment was edited on Feb 8, 23:37. |
|
|
|
|
|
|
|
|
| "I'm too much of a narcisist to really hate stupid people." |
|
|
|
|
| |
|
| 31. |
Heh |
Feb 8, 2004, 22:43 |
Creston |
|
|
WaltC and Wazootyman, I just love watching you guys banter about this stuff...
Hang out with Carmack much? 
I recall similar predications when the PC first went 16bit, and then again when it went 32bit. It'll take some time, but 64bit is simply inevitable.
I'm kinda wondering though when we'll stop keeping everything compatible with all our old crap. I'm sure Microsoft could actually create an OS that runs 10 times better than anything it currently has if it didn't have to be largely compatible with all those old apps.
Creston
|
|
|
|
|
|
|
|
| |
|
| 30. |
Re: i can't wait |
Feb 8, 2004, 22:04 |
dubfanatic |
|
|
oh yeah, anyone ever use actionsniper to win stuff on ebay? god, it's such a beautiful thing, so is emailing the guy who was in the lead up until the last 2 seconds of the auction, when you then come in and outbid bid a lot! HAHAHAHAHA! its so fun taunting them!
I hope you get scammed, preferably for a few thousand dollars. eBay sniping is only a rung higher than virus writing on the Pathetic Internet Activities ladder.
This comment was edited on Feb 8, 22:05. |
|
|
|
|
|
|
|
| |
|
| 29. |
Re: No subject |
Feb 8, 2004, 21:44 |
strangelove |
|
|
The 64-bit 'revolution', as with all major changes that depend on a hardware/software synergy, has been a bit of the chicken and the egg scenario. People don't buy the hardware because the software doesn't exist, but the software doesn't get written because the hardware isn't popular. This impasse has existed for years with Itanium. The beauty of the Opteron and Athlon64 is IMO the fact that they are capable of extremely good 32-bit performance, which makes the architecture useful without 64-bit apps. This makes it more likely that the hardware will proliferate, and then more software developers will release 64-bit versions of apps that can truly benefit from 64-bit architecture. Our company is now planning 64-bit releases, since our customer base is starting to purchase A64s and Opterons in significant enough numbers to justify the dev time (yay finally.)
I am also curious to see if rendering complex scenes in say 3dsmax is quicker in a 64-bit windows envirnment.
Rendering complex scenes is exactly the type of application that benefits from a 64-bit environment. The speedup is sometimes dramatic.
|
|
|
|
|
|
|
|
| |
|
| 28. |
Re: Drivers.... |
Feb 8, 2004, 21:38 |
Michael Bolton's hair dresser |
|
|
My god and I thought Shacknews readers were stupid.
|
|
|
|
|
|
|
|
|
| Dr. Michael Bolton's Hairdreser PhD, MD, LLP, DDS |
|
|
|
|
| |
|
| 27. |
Drivers.... |
Feb 8, 2004, 19:54 |
Lilricky |
|
|
What the article failed to mention, was that of the device drivers used, almost none were 64-bit, and those that were were not optimized.
|
|
|
|
|
|
|
|
| |
|
| 26. |
Re: No subject |
Feb 8, 2004, 19:23 |
Wazootyman |
|
|
WaltC - All I was talking about in my original post were various advantages and disadvantages of 64 bit processing. My point was not to claim that 64 bit processing was a bad move, but rather, that this evolution of the hardware also carried disadvantages. I never claimed those disadvantages outweighed the advantages.
Also...
I don't think it's question of "if" anymore... Yes, yes... As I said in my original post... 64 bit is the future... the question is how long it will take to become the standard. Personally, I would suspect it will not be a quick transition, considering the present investment in 32 bit hardware/software. Then again, I have been suprised by the speed of shifts in hardware standards in the past...
|
|
|
|
|
|
|
|
| |
|
| 25. |
Re: results??? |
Feb 8, 2004, 19:13 |
vacs |
|
|
Did you even read the quote blue posted?
here it is again.
"
The Athlon 64 quickly distinguished itself as the Gaming CPU with its outstanding performance in almost every 32-bit game
...[but] under XP64 Preview [scores]were 4% to 51% slower than 32-bit,
"
So, as you can see, YOU got it completly wrong.
ok, now read the quote you gave me again more slowly and try to understand the meaning behind it and not the words in it!
The first part was about the Arhlon 64 with a 32-bit OS running 32-bit games were the Athlon indeed marginally surpasses P4s of roughly the same speed.
The second part now is about the same Athlon 64 but with a 64-bit version of Windows running again the same 32-bit games as before. This setup is going to be important and will be widespread as soon as the 64-bit version of Windows ships... and this is the currently the disappointing part of the story in that those 32-bit games can run slower than before.
See, it's not that difficult to understand 
Again, the only thing which will matter this year is 32-bit performance for average Joe and the eager hardcore overclocker. The standard 64-bit desktop is going to be important next year at the earliest but for those you can't wait to run their 64-bit Windows should be prepared for what can happen -> AT article linked above.
This comment was edited on Feb 8, 19:15. |
|
|
|
|
|
|
|
| |
|
| 24. |
Re: No subject |
Feb 8, 2004, 18:57 |
ReDeeMeR |
|
|
In other words, AMD is the king, Intel sux ass while trying to keep up and that article is out of place, end of story.
|
|
|
|
|
|
|
|
| |
|
| 23. |
No subject |
Feb 8, 2004, 18:47 |
Nexus |
|
|
It's interesting to wonder what the difference would be if the A64s weren't also damn fast at 32 bit. If that were the case, the early adopters of high end hardware wouldn't have been going with it until there were 64 bit games that ran faster on it, which in turn there wouldn't have been any where near as quickly because the platform wouldn't have much market penetration at all.
Instead of looking at the chicken and the egg and wondering which should come first, we're all looking at the lovely shiny golden egg and considering how nice it is, while also pondering how nice it could be to have a chicken later. Since getting an egg is attractive, other people can start thinking about what happens when all those eggs turn into chickens.
Or something....
|
|
|
|
|
|
|
|
| |
|
| 22. |
Re: No subject |
Feb 8, 2004, 18:28 |
WaltC |
|
|
But does it double transfer rate? Of course, not everything the processor will be fetching from memory will by 64 bit instructions (x86 allowing variable length instructions... not to mention data), but, simply put, longer instructions = more memory transfer needed.
Biggest bottleneck there is latency, and the A64 memory controller was integrated into the cpu for just that purpose. Relative to latency and current core-logic, off-die memory controllers, it screams. My point was that AMD did that precisely with 64-bit processing in mind. It's *why* they moved the controller to the cpu, in other words.
Which makes it no less of a disadvantage... CPUs are getting more and more complex and that trend seems unlikely to reverse itself, but that doesn't mean increased complexity cannot be considered a disadvantage.
Well, it's like anything else, it's a pro-con situation. If the advantages to "greater complexity" outweigh the disadvantages, you gain when you "increase the complexity," right?...;)
Yes I'm aware of micro-ops, but that still doesn't change the fact that more steps will be required to reduce a 64 bit instruction with 4 arguments to micro-ops.
Well, speaking of "complexity," how about the requirement for pure RISC and/or VLIW for greatly increased compiler complexity and developer involvement on the software side...? I've never seen that moving the compexity from the hardware to the software is any great shakes in terms of an advantage. From a developer's point of view it could be a big disadvantage, depending on what you need to do. It's just not the black & white issue it was 10 years ago, is the point I wanted to make. The CISC-RISC controversy of a decade ago is no longer relevant, IMO.
Regardless of chip avaliability, 64 bit processing won't become a standard untill software developers jump on board. Like I said: 64 bit is the future, but how close is that future? I'm not really sure.
Chicken and egg problem, right? Before 64-bit computing can really take off developers need to see an established base of 64-bit platforms. That's the beauty of A64, and is why AMD did it this way. The fact that M$ is directly supporting the standard virtually guarantees its success--it's not only M$, but Intel is also jumping on the x86-64 bandwagon (although the specifics of Intel's approach are yet to be revealed.) Producing cpus that are natively both IA-32 and x86-64 in hardware is the very best way to ensure the standard will succeed. I don't think it's question of "if" anymore...;)
|
|
|
|
|
|
|
|
|
| It is well known that I do not make mistakes--so if you should happen across a mistake in anything I have written, be assured that I did not write it! |
|
|
|
|
| |
41 Replies. 3 pages. Viewing page 1.
< Newer [ 1 2 3 ] Older >
|
|
Twitter 
