-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
P4 660 couldn't beat 64 FX55 until @ 5.2ghz
Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
NightCrawler
Diamond Member
I DISAGREE, the Pentium 4 isn't much better at either video encoding or multitasking.
EightySix Four
Diamond Member
actually, last time I saw, the multitasking was better on an athlon, at least according to anand's own bench's.
VirtualLarry
No Lifer
I wonder if some of the P4's apparent advantage at media-encoding will be chipped away at, once AMD64 CPUs have SSE3 opcodes. Not to mention the new 2MB L2 P4 chips don't offer any media-encoding advantages over their prior 1MB/512KB L2 brethren.
It isn't about "much" better ...the fact is it is better...whether 1%-10% better it is still better. If someone came to me and does this sort of aplication specifically that 1-10% adds up over time and over 3-4 hour encodes depeneding on passes, bitrate, HD or not...now with HD encoding FPS are really low and so 10% of a shite load of time is still a shite load....
OK that statement was stupid and rules you out as someone having a worthwhile conversation with!!!! LOL!!!!
trust me bud!!! I had a P4 and I tested it versus this 2.6ghz and it is better in many things...many things the reviewers dont test...like running 2 cpu intensive apps like encoding and Folding.....2 instances of folding...I am not talking about taking 3 or 4 apps that all together barely use 100% cpu and see which is faster cause that negates the point...2 cpu intensive apps (intensive as a single app by themselves).....
1 AT test means nothing.....technically pcmark04 runs 3 to 4 of them and the A64 wil not win any of them and is the reason the p4 scores so much higher then the A64 in pcmark04.....
as far as I know it will have alittle effect cause there are not many of the apps that have SSE3 now.....P4 has its lead in many of those apps plain and simple due to HT and quite a few of those encoding apps being multithreaded capable....
Dual core will break into those....Since Intels dual core is rumored not to have HT it will really cut into those leads if you ask me....
superkdogg
Senior member
Does anybody feel like compressing the same large set of files with WinZip on an Intel and WinRar on an A64? I'd like to know which pair provides the best compression speed for backups, etc.
Also, a point that many people have left out of this thread-all modern computers are capable and in fact excel @ office apps and multitasking. I know that this statement is flamebait extraordinare, but IMHO it's true. I have a P4 HT system @ work and my A64 is not noticably different for running multiple apps. I am always running some sort of distributed project, playing music, maybe burning a CD and surfing the web or something else. This never gives my rig a hiccup. For this reason, I submit that benchmarks in these fields are basically synthetic because even if the number is different they amount to no real-world difference.
That leaves pretty much encoding, compiling (for those who need to), and gaming as the only real applications in which the CPU makes a huge difference. Encoding is usually won by Intel CPU's, often because of superior software optimization (SSE3, specialized code). A64 always wins @ gaming. Thus, it's basically like I said before you need to decide which is better for you. It's simply not possible to declare a winner.
The whole value, heat dissipation and future proof thing were why I went AMD basically. I didn't want to by a whole new rig in 6 months to go dual core, didn't want to have my CPU throttle @ stock speeds, and didn't want to pay exorbitant amounts for an Intel board, DDR2, and the CPU itself. Anybody else is free to make their own choices.
Also, a point that many people have left out of this thread-all modern computers are capable and in fact excel @ office apps and multitasking. I know that this statement is flamebait extraordinare, but IMHO it's true. I have a P4 HT system @ work and my A64 is not noticably different for running multiple apps. I am always running some sort of distributed project, playing music, maybe burning a CD and surfing the web or something else. This never gives my rig a hiccup. For this reason, I submit that benchmarks in these fields are basically synthetic because even if the number is different they amount to no real-world difference.
That leaves pretty much encoding, compiling (for those who need to), and gaming as the only real applications in which the CPU makes a huge difference. Encoding is usually won by Intel CPU's, often because of superior software optimization (SSE3, specialized code). A64 always wins @ gaming. Thus, it's basically like I said before you need to decide which is better for you. It's simply not possible to declare a winner.
The whole value, heat dissipation and future proof thing were why I went AMD basically. I didn't want to by a whole new rig in 6 months to go dual core, didn't want to have my CPU throttle @ stock speeds, and didn't want to pay exorbitant amounts for an Intel board, DDR2, and the CPU itself. Anybody else is free to make their own choices.
You ppl..... Trust me since I am sure you have no real testnig of this but running FH which is a cpu intensive app and another app will not likely hiccup any system...FH is programed that way to work in the background....It will take what cpu cycles are free...
I will bet you cash that the FH and same app ran will get more done in terms of percentage on the P4 versus the AMd...that is multitasking you ppl dont seem to get...It is not about lagging the system but getting the most work done using multi apps in a given time period.....LOL!!!!
I had a p4 system and 2 instances of FH garners more work done by about 30+%... 2instance of FH on AMD system garners no more extra % of work being done...Doesn't mean the system lags...It runs fine....
Most of the morons who comment on this topic of multitasking usually like to state...I can run like 4 IE wiondows opened...watching a movie, AV in background, and burning a CD or DVD!!! Big Fing deal...All of those apps together rarely use 100% cpu utilization......
Run FH 24/7 then encode an avi movie from TMPGenc to mpeg2.....Tell me how fast the movie gets done normally with no FH then with FH running and at same time how much of the WU is done??? That is a true test of HT's real world performance.....
Again it is not SSE3 that give Intel the boost... It had the leads with the p4c northwood which had no SSE3...It is HT plain and simple....Look at techreport and AT and all the other sites that used that worldbench..MOst of those apps in worldbench have no SSE3 optimizations....
SolMiester
Diamond Member
I couldnt give a toss really, however superkdogg gets my vote.
EightySix Four
Diamond Member
haha, only breezed through their last one and saw that they won the top one by fair margin.
Still they're not completely pwned. they won 2 tests v.s. 3
At least in the last anandtech runs with the new p4's
RussianSensation
Elite Member
Multi-Threaded Tasks
"This is actually the hardest test, as we have Norton AntiVirus software running in the background and the whole bunch of office applications, such as Microsoft Excel, Microsoft Project, Microsoft Access, Microsoft PowerPoint, Microsoft FrontPage and WinZip. And only in this case Intel processors supporting Hyper-Threading manage to defeat AMD competitors and become indisputable winners. This way, it is evident that AMD Athlon 64 processors can easily cope with simple multi-threaded tasks. However, when it comes to more serious workloads, Hyper-Threading technology proves highly efficient." - xbitlabs
Of course I dont know how many ppl will do all of this on every day basis. But I am going to go with Duvie and bet $$$ that P4 with 2 instances of SETI@Home will give you better frame rates at gaming than A64 with 1 instance of SETI even. (And P4 will finish a SETI unit faster while gaming too).
"This is actually the hardest test, as we have Norton AntiVirus software running in the background and the whole bunch of office applications, such as Microsoft Excel, Microsoft Project, Microsoft Access, Microsoft PowerPoint, Microsoft FrontPage and WinZip. And only in this case Intel processors supporting Hyper-Threading manage to defeat AMD competitors and become indisputable winners. This way, it is evident that AMD Athlon 64 processors can easily cope with simple multi-threaded tasks. However, when it comes to more serious workloads, Hyper-Threading technology proves highly efficient." - xbitlabs
Of course I dont know how many ppl will do all of this on every day basis. But I am going to go with Duvie and bet $$$ that P4 with 2 instances of SETI@Home will give you better frame rates at gaming than A64 with 1 instance of SETI even. (And P4 will finish a SETI unit faster while gaming too).
DrMrLordX
Lifer
Can we please not turn this into another HT discussion?
If you're stuck with a single CPU, and you frequently run multiple CPU-intensive apps, the P4 starts to shine. Whether or not these constitute "very specific conditions" varies from user to user. Some people like to encode vids in the background while playing Halflife 2 or Doom 3. I never encode anything.
That being said, a single P4 with HT is nothing compared to a good dual-cpu setup. It seems fruitless to argue about whether or not HT is useful for multitasking when you consider how "ghetto" single-cpu P4s and Athlon 64s are when compared to duallie rigs(at least where multitasking and multithreaded apps are concerned).
Think a duallie is out of your reach? MichaelD has a thread for you here
http://forums.anandtech.com/messageview...atid=28&threadid=1525940&enterthread=y
And, in some vain attempt to stay on-topic, I found the originally linked article to be rather silly. Anyone who runs an OCed P4 against a FX-55 at stock, or vice versa, is generating a worthless comparison. Any conclusions drawn from such a benchmark are equally worthless.
If you're stuck with a single CPU, and you frequently run multiple CPU-intensive apps, the P4 starts to shine. Whether or not these constitute "very specific conditions" varies from user to user. Some people like to encode vids in the background while playing Halflife 2 or Doom 3. I never encode anything.
That being said, a single P4 with HT is nothing compared to a good dual-cpu setup. It seems fruitless to argue about whether or not HT is useful for multitasking when you consider how "ghetto" single-cpu P4s and Athlon 64s are when compared to duallie rigs(at least where multitasking and multithreaded apps are concerned).
Think a duallie is out of your reach? MichaelD has a thread for you here
http://forums.anandtech.com/messageview...atid=28&threadid=1525940&enterthread=y
And, in some vain attempt to stay on-topic, I found the originally linked article to be rather silly. Anyone who runs an OCed P4 against a FX-55 at stock, or vice versa, is generating a worthless comparison. Any conclusions drawn from such a benchmark are equally worthless.
NO.....It is well documented that you need at least 2 cpu intensive apps (IE like the ones that hold 100%) to take advantage of it in MULTITASKING, or at least distinguish it from other p4's non HT and AMD systems....It is very easy to find these and ppl who Fold, run CAD, and encode (basic P4 users anyways) know this already.....You AMD fanboys are just clueless as usual....
I still Say TMPGenc (IE video encoding) which can take a considerable amount of time if one chooses multi passes, high bitrates, and motion compensation.....So instead of tieing up the computer for 4-6 hours you can feel you can use the computer effectively for other applications while still getting work done....Maybe keep the box folding and getting work units done....maybe shrink a DVD and burn it...maybe play a game and even with decent fram rates get work done on the encoding....
The advanatge is not minimal and that is a lame comment....
Do I need to rehash the HT thread of mine??? TMPGenc by itself with multithreaded enabled will gain 22% increase in encoding...Viz Renderer of my ADT2004 cad program is also SMP (multithreaded capable) and HT had also a 19-22% increase in speed. I take a complexed 3d model of a house and set up an animated walk-thru with 24-25fps and then want to create a movie of it the program has to render hundreds of frames for very little time. With heavy detailed materials, shadowing, reflectivity, and lighting effects this can take 20 seconds per each frame at say a 1024x768 resolution....now take that over say 625 frames for 25 seconds of video and that is a shite load of time....3-1/2 hours approximately...Now enable multithreaded capabilities in the render dialog box and I cut that time to 2hours and 45 minutes...
Is that real enough for you know nothings??? I am soooo tired of you shit-talkers who absoultely know nothing about which you are talking.....
Edit: I agree lets get back on topic about the absurdity of this original review.....If AMD users who obviously know very little keep their mouths shut about what they dont knwo it wouldn't turn into this...Just a huge FUD machine here constantly....
Well the trick may be because the longer pipeline of the prescott its actual poorer performance due to longer penalties due to branch misprediction HT actually speeds it up more.....Once you understand what HT enhances...a basic flaw in the architecture or consequence of the long pipeline of the P4 architecture then you can understand why the prescott actually would increase....So I dont doubt it....The comparison is too take a similar speed prescott and northwood and run each with HT off and then each with HT on...i wonder if the HT on gets them close to one another then when it was not on.....
Edit: Yes it was a northwood.....I dont do Prescott!!!
Nope I sold it about a month after I go the A64 board....I didn't have enough parts (cases, PSU, memory, fans, drives, etc...) to keep them both...i wish I could have, but I made a decision based on a few of the cad apps I run and overall system performance of this to my 3.5ghz p4.....
I wish I could but I believe the reason the prescott has rumored better HT is because I what I mentioned....
I wish I could but I believe the reason the prescott has rumored better HT is because I what I mentioned....
superkdogg
Senior member
You are proving my point. Hyperthreading only matters if somebody is running two very CPU intensive applications at the same time. This is something that most home users don't do. That's exactly why I gave the laundry list of things I can do @ once without a difference. This is basically the same old argument about HT. Is it really a needed feature? No. Is it pretty nice to have? Sure. Does it actually help? Sometimes.
Let's go back pre-HT Northwoods, shall we? Were dual-cpu rigs the most powerful for everday computing needs? Depends on what is being done. Even now that dual-core CPU's are the next generation, everyone is aware that they will offer little or no immediate performance increase.
My whole point is that if you don't need to do two CPU intensive things @ the same time(which few people do), you don't need HT. Being able to run two instances of seti is a relatively limited application, considering all the market niches and every user. HT's marketing is that it allows the user to be more effective. It does, but not in every scenario and that's why I say that in my case it would be of little or no value to me-because as you reminded everyone, the scenario that I presented is not when HyperThreading helps. It just seems like it should be a scenario that a multitasking optimization like HT would help. Which is, after all, my point. It is nice to have for the instances when people use it, but it's actual effect is minimal for most users.
P.S- If I want to use HT with my AMD system, it's simple. I press "scroll lock-scroll lock-arrow key" to change machines because with the money I saved (no DDR2, no extra cooling, $100 mobo) I built another inexpensive rig. That's really efficient multi-tasking! Flame on.
BTW, Duvie-I have a lot of respect for you. Your posts are generally well written and you've never been a jerk that I've read. I just honestly disagree with you on the value of HT in the current implementation and setting in which it's being used. It is better for two instances of seti or not dropping as much performance if you have a serious CPU app. running in the background. It just has a relatively narrow realm in which it actually helps and even when it helps it's not like it's revolutionizing the CPU world. Especially once the dual-core chips come out.
SSDD :thumbsdown: This tired debate make me feel like Bill Murray in Groundhog Day 😛
Celeron sells in bigger numbers than most other desktop CPUs put together, and until the latest Deleron incarnation was a real slug since moving to the P4 core a few years back. I can think of no more eloquent testimony to point out that any differences in performance between the higher-end offerings from either company amounts to no more than quibbling :light:
IMHO, the only real performance issue currently worth debating about new gen chips is the potential clock throttling of the higher clocked Prescotts in a typical OEM built system. The fact that many users wouldn't even know it is happening just strikes me as somewhat duplicitous. They should get every mhz performance they paid for when they need it, not be potentially experiencing reduced performance under heavy/demanding usage.
Oh, and the Russian's post is on the money IMO.
A note to anyone running more than 1 instance of F@H with Hyperthreading: Please STOP! Your damned competitive need for higher personal and/or team stats is holding up analysis of the results!
Celeron sells in bigger numbers than most other desktop CPUs put together, and until the latest Deleron incarnation was a real slug since moving to the P4 core a few years back. I can think of no more eloquent testimony to point out that any differences in performance between the higher-end offerings from either company amounts to no more than quibbling :light:
IMHO, the only real performance issue currently worth debating about new gen chips is the potential clock throttling of the higher clocked Prescotts in a typical OEM built system. The fact that many users wouldn't even know it is happening just strikes me as somewhat duplicitous. They should get every mhz performance they paid for when they need it, not be potentially experiencing reduced performance under heavy/demanding usage.
Oh, and the Russian's post is on the money IMO.
A note to anyone running more than 1 instance of F@H with Hyperthreading: Please STOP! Your damned competitive need for higher personal and/or team stats is holding up analysis of the results!
also-
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
