Race Sim Central
Rec.Autos.Simulators Archive

OT: CPUs- what's the deal?

23 posts by 13 people · 14 Oct 2004 09:43 UTC - 23 Oct 2004 02:38 UTC

Have CPU speeds stagnated?

Steve Simpson asked why CPU performance had stalled. Replies pointed to Intel's Pentium M architecture and AMD's Athlon 64, with upcoming dual-core chips and 64-bit Windows, and debated whether screen resolution affects CPU load in games.

hardware: PC/graphicsoff-topic

Automated summary.

Steve Simpson 14 Oct 2004 09:43 UTC#293917
Anyone else noticed that CPUs have pretty much stagnated in the last 2 years? Have AMD and Intel run out of ideas? Is there anything promising on the horizon?
"Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > years? Have AMD and Intel run out of ideas? Is there anything promising > on the horizon? Intel adopting their notebook CPU architecture will be an intersting step. My 1.6 GHz Pentium-M benchmarks faster than a 2.2 GHz P4.
"Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > years? Have AMD and Intel run out of ideas? Is there anything promisingon > the horizon? Yes I have. I just up graded from a 1.4 T-bird @ 266 fsb to a 3000XP @ 333 fsb and did not see that much of a gain. the only thing that saw a big increase in was NR2003. Granted I was all ready running a ATI 9700 pro video card. MadDAWG
"Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > years? Have AMD and Intel run out of ideas? Is there anythingpromising on > the horizon? I'm guessing everything will kick off again once Microsoft release a 64bit mainstream OS & MSOffice. imo the change to 64bit isn't as pressing as the change from 16 to 32bit was back in the Win95 days, and even that took several years. There's only so much market pressure that can be applied by computing's equivalent of hotrodders (people who have to have the latest & greatest no matter what), what's needed is a few 'killer apps' that really push 64bit (or just more powerful processing in any form) into the mainstream, and it's anyone's guess what that will be & when ;-) Malc.
"Malc" <[email hidden]> wrote in message news:416e6c21$0$44872$[email hidden]...
Quoted text (21 lines)> "Steve Simpson" <[email hidden]> wrote in message > news:[email hidden]... >> Anyone else noticed that CPUs have pretty much stagnated in the last 2 >> years? Have AMD and Intel run out of ideas? Is there anything > promising on >> the horizon? > > I'm guessing everything will kick off again once Microsoft release a > 64bit mainstream OS & MSOffice. > > imo the change to 64bit isn't as pressing as the change from 16 to 32bit > was back in the Win95 days, and even that took several years. > > There's only so much market pressure that can be applied by computing's > equivalent of hotrodders (people who have to have the latest & greatest > no matter what), what's needed is a few 'killer apps' that really push > 64bit (or just more powerful processing in any form) into the > mainstream, and it's anyone's guess what that will be & when ;-) > > Malc. >
Agreed. Given enough memory, even my current ThinkPad notebook is running my finite element analyses much faster than my old Silicon Graphics dual-processor Octane. Even for heavily non-linear analyses, the current generation of PCs is more than adequate.
Steve Simpson wrote: > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > years? Have AMD and Intel run out of ideas? Is there anything > promising on the horizon? Don't know about Intel as I am an AMD guy and really don't keep up with Intel's progress. AMD, imho, is definitely on the move. Their new 64 bit processors, from what I have seen, are kicking Intel's butt in benchmarks. We are pretty much in a transition mode, with lots of new technology on the horizon. The only reason I have not updated to a 64 bit processor, is I am waiting for socket 939 to become more readily available and cheaper. Socket 754 I understand is pretty darn good, and 64 bit amd cpu in socket 754 spanks socketA 32 bit processors pretty good. But with 939 on the horizon, I would rather wait and make the jump to it. We are at the bottom of the curve on these new 64 bit processors, and especially socket 939. Socket A indeed, has stagnated, as they have reached the upper echelon on it, and no where left to go. AMD's Barton 32 bit processors, in socket A, still have a lot of longetivy to them. My XP 2800+ running at 3200+ speeds, still gives me great performance in anything I throw at it, and scores around 6400 in 3dmark 03. We also have pci express on the horizon, to take the place of both reg pci and agp. I am just going ot wait, until the bugs get settled in some of this new technology, and upgrade then. Although, with the reasonable price on socket 754 boards and processors, it is tempting to go ahead and upgrade. If my system was choking now, I would not hesitate to do so. Hardware wise, things will be very different in a couple of years. -- Don Burnette
MadDAWG wrote:
Quoted text (15 lines)> "Steve Simpson" <[email hidden]> wrote in message > news:[email hidden]... >> Anyone else noticed that CPUs have pretty much stagnated in the last >> 2 years? Have AMD and Intel run out of ideas? Is there anything >> promising on the horizon? > > > Yes I have. > > I just up graded from a 1.4 T-bird @ 266 fsb to a 3000XP @ 333 fsb > and did not see that much of a gain. the only thing that saw a big > increase in was NR2003. Granted I was all ready running a ATI 9700 > pro video card. > > MadDAWG
I find that a little suprising. I upgraded from a XP 1600+, to an XP 2400+, then to an XP 2800+ Barton ( I know, but with the price so cheap why not), and I definitely saw difference in each 3 steps. Of course, I overclocked each one to their upper limits. I also went from 9700 Pro to 9800 Pro during this time as well. -- Don Burnette
Mitch_A 14 Oct 2004 16:47 UTC↳ reply to MadDAWG#30501
Something isnt right. You are the "exception not the rule." Nothing beats CPU cycles. Mitch "MadDAWG" <[email hidden]> wrote in message \
Quoted text (9 lines)> Yes I have. > > I just up graded from a 1.4 T-bird @ 266 fsb to a 3000XP @ 333 fsb and did > not see that much of a gain. the only thing that saw a big increase in was > NR2003. Granted I was all ready running a ATI 9700 pro video card. > > MadDAWG > >
Malc 14 Oct 2004 17:54 UTC↳ reply to MadDAWG#30509
"MadDAWG" <[email hidden]> wrote in message news:416e68fd$[email hidden]... > > "Steve Simpson" <[email hidden]> wrote in message > news:[email hidden]... > > Anyone else noticed that CPUs have pretty much stagnated in the last2 > > years? Have AMD and Intel run out of ideas? Is there anythingpromising
Quoted text (7 lines)> on > > the horizon? > > > Yes I have. > > I just up graded from a 1.4 T-bird @ 266 fsb to a 3000XP @ 333 fsb and
did > not see that much of a gain. the only thing that saw a big increase inwas > NR2003. Granted I was all ready running a ATI 9700 pro video card. I went from an XP2000 (133) to my current XP3000 (166) a while back & didn't notice much performance difference, although I was (and still am) much happier with the cooler temps. I put the XP2000 in my second PC though, and going back it feels slower & the difference is plain to see when you can compare directly. Malc.
Don, I think maybe this is your lucky day! Check out this article on anandtech: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2242 Looks like a good time to upgrade to socket 939 with a new .09 Athlon 64 3000+ for under $200 (plus new mobo) and overclock it to FX-53 speeds. I think I need to start looking at switching from my current Intel setup to a new AMD setup! :) On the Intel side of CPU's, I was also thinking development seems to have slowed down. The Intel 2.4B that I got almost two years ago for just under $200 is still around and selling for about $150. I would've thought that we'd be seeing 5+ GHz by now, but I guess they're looking at making the processors more powerful at lower clock speeds. "Don Burnette" <[email hidden]> wrote in message news:[email hidden]...
Quoted text (44 lines)> Steve Simpson wrote: >> Anyone else noticed that CPUs have pretty much stagnated in the last 2 >> years? Have AMD and Intel run out of ideas? Is there anything >> promising on the horizon? > > Don't know about Intel as I am an AMD guy and really don't keep up with > Intel's progress. > > AMD, imho, is definitely on the move. Their new 64 bit processors, from > what I have seen, are kicking Intel's butt in benchmarks. > > We are pretty much in a transition mode, with lots of new technology on > the horizon. The only reason I have not updated to a 64 bit processor, is > I am waiting for socket 939 to become more readily available and cheaper. > Socket 754 I understand is pretty darn good, and 64 bit amd cpu in socket > 754 spanks socketA 32 bit processors pretty good. But with 939 on the > horizon, I would rather wait and make the jump to it. > We are at the bottom of the curve on these new 64 bit processors, and > especially socket 939. > > Socket A indeed, has stagnated, as they have reached the upper echelon on > it, and no where left to go. AMD's Barton 32 bit processors, in socket A, > still have a lot of longetivy to them. My XP 2800+ running at 3200+ > speeds, still gives me great performance in anything I throw at it, and > scores around 6400 in 3dmark 03. > > We also have pci express on the horizon, to take the place of both reg pci > and agp. > > I am just going ot wait, until the bugs get settled in some of this new > technology, and upgrade then. Although, with the reasonable price on > socket 754 boards and processors, it is tempting to go ahead and upgrade. > If my system was choking now, I would not hesitate to do so. > > Hardware wise, things will be very different in a couple of years. > > > > > -- > Don Burnette > > >
Thanks for the link Bill. Very interesting, and, very tempting! -- Don Burnette Bill Worrel wrote: > Don, > > I think maybe this is your lucky day! Check out this article on > anandtech: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2242 > > Looks like a good time to upgrade to socket 939 with a new .09 Athlon > 64 3000+ for under $200 (plus new mobo) and overclock it to FX-53 > speeds. I think I need to start looking at switching from my current > Intel setup to a new AMD setup! :) > > On the Intel side of CPU's, I was also thinking development seems to > have slowed down. The Intel 2.4B that I got almost two years ago for > just under $200 is still around and selling for about $150. I > would've thought that we'd be seeing 5+ GHz by now, but I guess > they're looking at making the processors more powerful at lower clock > speeds. > > > "Don Burnette" <[email hidden]> wrote in message > news:[email hidden]... >> Steve Simpson wrote: >>> Anyone else noticed that CPUs have pretty much stagnated in the >>> last 2 years? Have AMD and Intel run out of ideas? Is there >>> anything promising on the horizon? >> >> Don't know about Intel as I am an AMD guy and really don't keep up >> with Intel's progress. >> >> AMD, imho, is definitely on the move. Their new 64 bit processors, >> from what I have seen, are kicking Intel's butt in benchmarks. >> >> We are pretty much in a transition mode, with lots of new technology >> on the horizon. The only reason I have not updated to a 64 bit >> processor, is I am waiting for socket 939 to become more readily >> available and cheaper. Socket 754 I understand is pretty darn good, >> and 64 bit amd cpu in socket 754 spanks socketA 32 bit processors >> pretty good. But with 939 on the horizon, I would rather wait and >> make the jump to it. We are at the bottom of the curve on these new 64 >> bit processors, and >> especially socket 939. >> >> Socket A indeed, has stagnated, as they have reached the upper >> echelon on it, and no where left to go. AMD's Barton 32 bit >> processors, in socket A, still have a lot of longetivy to them. My >> XP 2800+ running at 3200+ speeds, still gives me great performance >> in anything I throw at it, and scores around 6400 in 3dmark 03. >> >> We also have pci express on the horizon, to take the place of both >> reg pci and agp. >> >> I am just going ot wait, until the bugs get settled in some of this >> new technology, and upgrade then. Although, with the reasonable >> price on socket 754 boards and processors, it is tempting to go >> ahead and upgrade. If my system was choking now, I would not >> hesitate to do so. Hardware wise, things will be very different in a >> couple of years. >> >> >> >> >> -- >> Don Burnette
> Agreed. Given enough memory, even my current ThinkPad notebook is running > my finite element analyses much faster than my old Silicon Graphics > dual-processor Octane. Even for heavily non-linear analyses, the current > generation of PCs is more than adequate. That may be true but as far as racing simulations go, the current generation of CPUs are really dragging the chain. It's like the people that made the current crop of sims were expecting CPUs to be much faster by this stage too...
In article <[email hidden]>, [email hidden] says...
Quoted text (10 lines)> Steve Simpson wrote: > > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > > years? Have AMD and Intel run out of ideas? Is there anything > > promising on the horizon? > > Don't know about Intel as I am an AMD guy and really don't keep up with > Intel's progress. > > AMD, imho, is definitely on the move. Their new 64 bit processors, from what > I have seen, are kicking Intel's butt in benchmarks.
Yup.
Quoted text (8 lines)> We are pretty much in a transition mode, with lots of new technology on the > horizon. The only reason I have not updated to a 64 bit processor, is I am > waiting for socket 939 to become more readily available and cheaper. Socket > 754 I understand is pretty darn good, and 64 bit amd cpu in socket 754 > spanks socketA 32 bit processors pretty good. But with 939 on the horizon, I > would rather wait and make the jump to it. > We are at the bottom of the curve on these new 64 bit processors, and > especially socket 939.
What's really cool is that the current 939 motherboards are already compatible (if you believe the marketing spin) with the dual-core 939 CPUs that will be released next year. And AMD just released the 939 90nm version of the AMD64 3000 at a price point less than $200. I'm currently running an AMD XP3000 and will be upgrading to AMD64/PCI Express next summer.
Yep 64 bit is a goer but only once the game code is all changed. 64 bit CPU's in 32 bit emulation mode = slower processing for the same clock speed. regards sales1@gamesteer "Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > Anyone else noticed that CPUs have pretty much stagnated in the last 2 > years? Have AMD and Intel run out of ideas? Is there anything promisingon > the horizon? > >
On Sun, 17 Oct 2004 12:01:23 +1000, "Sales1 at Gamesteer" <[email hidden]> wrote: >Yep > >64 bit is a goer but only once the game code is all changed. 64 bit CPU's in >32 bit emulation mode = slower processing for the same clock speed. >regards AMD64 can run 32bit games/code just as fast, or faster, than a native 32bit Intel cpu.
On Sun, 17 Oct 2004 12:01:23 +1000, "Sales1 at Gamesteer" <[email hidden]> scribbled: >Yep > >64 bit is a goer but only once the game code is all changed. 64 bit CPU's in >32 bit emulation mode = slower processing for the same clock speed. >regards >sales1@gamesteer This is true, but only if the CPU is 64-bit only, which the current ones are not. They're hybrids, and perform 32 bit tasks every bit as well as anything else currently out. An Athlon64 is not going to be slower than a P4 just because it has a 64 in the name. -Slash -- "Ebert Victorious" -The Onion
MadDAWG 18 Oct 2004 10:54 UTC↳ reply to Mitch_A#30403
"Mitch_A" <[email hidden]> wrote in message news:15ybd.13063$[email hidden]... > Something isnt right. You are the "exception not the rule." > > Nothing beats CPU cycles. > > > Mitch I think its just fine. IMO the software has alot to do with it. As we all know Papy games are CPU hogs and that is the one spot where I did have a noticable differance. For most of my games I have been running a low screen res to keep cpu load down while increasing the FSAA and AF to take advantage of the video card. I don't play F1C or some of the others that seem to big CPU hogs. MadDAWG
Quoted text (9 lines)> I think its just fine. IMO the software has alot to do with it. As we all > know Papy games are CPU hogs and that is the one spot where I did have a > noticable differance. For most of my games I have been running a low > screen > res to keep cpu load down while increasing the FSAA and AF to take > advantage > of the video card. I don't play F1C or some of the others that seem to > big > CPU hogs.
Running a low resolution has no effect on the CPU. In theory if you've got a very slow CPU and a very fast video card there'll be no difference in speed between 640x480 and 1600x1200
"Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > > I think its just fine. IMO the software has alot to do with it. As weall
Quoted text (10 lines)> > know Papy games are CPU hogs and that is the one spot where I did have a > > noticable differance. For most of my games I have been running a low > > screen > > res to keep cpu load down while increasing the FSAA and AF to take > > advantage > > of the video card. I don't play F1C or some of the others that seem to > > big > > CPU hogs. > > Running a low resolution has no effect on the CPU. In theory if you've
got > a very slow CPU and a very fast video card there'll be no difference in > speed between 640x480 and 1600x1200 Thats not been my experaince. The way it was explained to me was that the CPU still has to feed the video card information, and the video card's speed only comes into play with what it does after it has that info. I don't know if thats a correct explaintaion, but I know for sure that on my sytem 640x800 was way faster than 1280x960 with the same video card. MadDAWG
Quoted text (7 lines)> Thats not been my experaince. The way it was explained to me was that the > CPU still has to feed the video card information, and the video card's > speed > only comes into play with what it does after it has that info. I don't > know > if thats a correct explaintaion, but I know for sure that on my sytem > 640x800 was way faster than 1280x960 with the same video card.
If that's the case, then your video card is the bottleneck. Maybe you're running with AA/AF and your video card can't keep up at 1280x960.
"Steve Simpson" <[email hidden]> wrote in message news:[email hidden]... > > I think its just fine. IMO the software has alot to do with it. As weall
Quoted text (10 lines)> > know Papy games are CPU hogs and that is the one spot where I did have a > > noticable differance. For most of my games I have been running a low > > screen > > res to keep cpu load down while increasing the FSAA and AF to take > > advantage > > of the video card. I don't play F1C or some of the others that seem to > > big > > CPU hogs. > > Running a low resolution has no effect on the CPU. In theory if you've
got > a very slow CPU and a very fast video card there'll be no difference in > speed between 640x480 and 1600x1200 > > Sounds wrong to me. Since the advent of 3D cards, graphics cards do handle more of the processing than b4 but those pixels still need to be generated somewhere. Graphics cards make images look smoother but dont create the image from scratch. More resolution = more cpu req.
> Sounds wrong to me. Since the advent of 3D cards, graphics cards do > handle more of the processing than b4 but those pixels still need to be > generated somewhere. Graphics cards make images look smoother but dont > create the image from scratch. More resolution = more cpu req. Nope. If the CPU is maxed out and the video card isn't, the resolution has very little, if any, impact on the frame rate. This has been shown time and again in various benchmarks on hardware sites. I suspect you may be thinking of the pre-3D accelerator days when the CPU had to generate every pixel.
"Sales1 at Gamesteer" <[email hidden]> wrote in message news:4171d273$0$10345$[email hidden]... > Yep > > 64 bit is a goer but only once the game code is all changed. 64 bit CPU'sin > 32 bit emulation mode = slower processing for the same clock speed. > regards > sales1@gamesteer If I'm not mistaken, the Athlon64 has native 32-bit processing. IOW, with 32-bit code, it actually functions as a true blue 32-bit CPU.. That's why they are performing so well NOW with existing 32-bit OS'es and software. Granted, there are other changes to the architecture that greatly improve performance in certain applications, but it's still very good at 32-bit in terms of processing alone. Alan

Sponsors

Affiliate Links

Sim Gaming Expo MOZA Racing Sim-Lab Fanatec TrakRacer Asetek Internet Privacy From NordVPN Dreamhost Capital One Credit Card Application Fanatical CDKeys Amazon Amazon UK Project Motor Racing iRacing Get your racing gloves, boots and more from Demon Tweeks. Enlist at Roberts Space Industries, developers of Star Citizen and Squadron 42


Back from the ashes since July, 2019. First created in 2001 with the merger of Legends Central (founded 1999) and simracing.dk.

A site by a sort of sim racer, for sim racers, about racing sims. News and information on both modern and historic sim racing software titles.

All products and licenses property of their respective owners.


Facebook Threads Bluesky Twitter Mastodon Instagram YouTube Tiktok Twitch Reddit Google News Apple News Subscribe via WordPress Flipboard eBay

Race Sim Central © 1999-2026 | Some links pay RSC a commission. Advertising does not equal endorsement.Cookie Policy