Socket 940 & FX51 now or wait for Socket 939?

Status
Not open for further replies.

[k]ar|

Gawd
Joined
Jun 17, 2002
Messages
561
Thanks xonik. Is there likely to be a major difference in performance between 940 (stable server) and 939 (performance desktop)? I want the power for gaming, so I'll be throwing in a 9800XT and 1gb RAM, but what I don't want is to be left with relatively slower cpu/mobo/ram in 3 months time when socket 939 comes out.

[k]
 
Well newer faster processors are always going to be coming out so its best not to spend a ton of money on the very fastest processor at the time because then you have plenty of money to keep upgradeing as new stuff comes out thats cheaper and faster.

The performance of the FX-51 doesn't warranty its hefty price tag. I wouldn't recommend it to anyone that didn't just have money to burn on the fastest processor. Socket 939 will be out next year which will support lower latency unbuffered RAM.

For right now i would buy a Pentium 4 and overclock or an A64 3000+. Ither one will be great performers. I would wait on new motherboards for the 3000+ though because the current ones aren't that great.

The FX-51's are still mainly targeted towards servers so since almost every game is video card intensive spending $750 on a processor is not needed.

Spend the extra money on a 9800 pro or xt.
 
Originally posted by burningrave101
The FX-51's are still mainly targeted towards servers so since almost every game is video card intensive spending $750 on a processor is not needed.

The FX-51 are high-end consumer grade CPU's, they're not server chips. True that they are based off the Opteron's with 2 Hypertransports disabled, but it doesn't make them server chips.
 
Originally posted by Asian Fury
The FX-51 are high-end consumer grade CPU's, they're not server chips. True that they are based off the Opteron's with 2 Hypertransports disabled, but it doesn't make them server chips.

Yes they are consumer enthusiast chips more then their server chips. But they are built on the same technology as the Opterons which are server chips. The only difference between the two is two HyperTransport links are disabled making it impossible to run them in a dual processor configuration on an Operton board. But they are 64bit and 64bit is marketed towards servers because of its ability to address large amounts of memory.

I probably shouldn't of said it was mainly for servers though because that is incorrect but somehow i got that stuck in my head cause i know i recently read that in one of the articles on the FX-51 about it mainly being bought for small servers and gaming enthusiasts.

All a server is actually is a computer that provides client stations with access to files and printers as shared resources to a computer network. A server doesn't have to have two processors.

With the abilities of an FX-51 being so much like an Opeteron it would make it perfect for a small server or workstation. It uses ECC memory also which is used for servers and they will continue marketing 940 socket chips even after the 939 socket are in full force for quite a while i believe.
 
Well newer faster processors are always going to be coming out so its best not to spend a ton of money on the very fastest processor at the time because then you have plenty of money to keep upgradeing as new stuff comes out thats cheaper and faster.

I beleive that the 939 are still going to be expensive, but not as much. A newer chip is ALWAYS more expensive regardless of the waffer. AMD could jack up the price on them even though it costs them less to make. AMD has not yet stated a price for those chips. Since the FXs are thier highend CPUs they are bound to atleast be $400+.

The FX-51's are still mainly targeted towards servers so since almost every game is video card intensive spending $750 on a processor is not needed.

The FXs ARE NOT targeted towards severs. The opterons are targeted towards servers. The FX are targeted to gamers and the like. From my experience also is that server processors are normally more stable and can handle more of a load since they are produced to handle big servers. So it is a great chip, I am using one right now and its never sluggish. I have a friend with a 3.2C with the same hardware that cant match the performance. Sure, games are video intensive, but you also need the power to proccess to and it will increase performance, just go look at benchmarks. Sure it is not NEEDED, but who the hell wouldnt want one? And for only $300 more you get alot better upgrade path then buying the other chips. The P4 line is soon to be replaced by Prescott, which only some mobos support. The 754 line is about to be replaced with 90nm chips. Sure the FXs will also be replaced, but there will still be opterons and I believe some more FXs to come with 940 socket. Also the FX supports dual memory channel allowing alot better RAM upgrade then the normal A64s, and registered RAM costs about the same as any decent unbuffered.
 
Originally posted by Sc0rched
The FXs ARE NOT targeted towards severs. The opterons are targeted towards servers. The FX are targeted to gamers and the like. From my experience also is that server processors are normally more stable and can handle more of a load since they are produced to handle big servers. So it is a great chip, I am using one right now and its never sluggish. I have a friend with a 3.2C with the same hardware that cant match the performance. Sure, games are video intensive, but you also need the power to proccess to and it will increase performance, just go look at benchmarks. Sure it is not NEEDED, but who the hell wouldnt want one? And for only $300 more you get alot better upgrade path then buying the other chips. The P4 line is soon to be replaced by Prescott, which only some mobos support. The 754 line is about to be replaced with 90nm chips. Sure the FXs will also be replaced, but there will still be opterons and I believe some more FXs to come with 940 socket. Also the FX supports dual memory channel allowing alot better RAM upgrade then the normal A64s, and registered RAM costs about the same as any decent unbuffered.

Dude read my second response right above you lol. I wasn't really paying attention to what all i said when i said about them being server chips. They are geared towards enthusiast gamers but can also be used for small servers and workstations. There is NOTHING different from them than the Opterons except two HyperTransport links are disabled. An FX-51 would be just as stable efficient as a single Opteron for a small server. You dont have to have two processors for a server.

You dont get a better upgrade path from spending $300 more on an FX-51 because 754 pin will be around just as long as the 939 and 940 pin will. They will continue to make 940 socket using ECC RAM even after 939 pin is out.

The dual channel on-die memory controller gives a big performance boost in bandwidth benchmarks but since games can't even use but a portion of that bandwidth it doesn't help in that area. Games can't even use near the bandwidth that is provided by a Pentium 4 with dual channel enabled.

http://www.tomshardware.com/cpu/20031223/cpu-guide-16.html

In most gaming benchmarks there is only minimal differences between an overclocked Pentium 4 around 3.5-3.6Ghz and an FX-51. IMO definatley not enough to warrant its price without a 64bit OS that is going to be a while coming down the road.
 
Dude read my second response right above you lol. I wasn't really paying attention to what all i said when i said about them being server chips. They are geared towards enthusiast gamers but can also be used for small servers and workstations. There is NOTHING different from them than the Opterons except two HyperTransport links are disabled. An FX-51 would be just as stable efficient as a single Opteron for a small server. You dont have to have two processors for a server.

I was typing my reply when you posted.

In most gaming benchmarks there is only minimal differences between an overclocked Pentium 4 around 3.5-3.6Ghz and an FX-51. IMO definatley not enough to warrant its price without a 64bit OS that is going to be a while coming down the road.

Now your comparing overclocked to stock...

You dont get a better upgrade path from spending $300 more on an FX-51 because 754 pin will be around just as long as the 939 and 940 pin will. They will continue to make 940 socket using ECC RAM even after 939 pin is out.

You misread. I know they will, but what I am saying is you will be able to upgrade to more RAM due to it being dual channel. The 754 is limited in RAM. Plus the newer games keep requiring more and more RAM. I remember when I was using 128 MB just a few years ago, now minimum recommended is 512...
 
Originally posted by Sc0rched
You misread. I know they will, but what I am saying is you will be able to upgrade to more RAM due to it being dual channel. The 754 is limited in RAM. Plus the newer games keep requiring more and more RAM. I remember when I was using 128 MB just a few years ago, now minimum recommended is 512...

It being dual channel doesn't mean you will be able to upgrade to more RAM. Dual Channel has to do with giving more bandwidth. Bandwidth and the size of the memory sticks are totally different things. So yes games are requiring more and more RAM but the bandwidth speed doesn't effect the size of RAM storage.

You can have a 512MB stick of PC2100 and a 512MB of PC3200. Both are the same size but the pc3200 offers 3.2GB/s bandwidth and the pc2100 offers 2.1GB/s bandwidth.

Also why wouldn't you compare them overclocked? People say dont buy a 3200+ because a 2500+ will run at the same speed. Whats the difference? 3.5Ghz is a give me on a 3.0c processor. At least it is on all the ones i've been around.
 
Yes but your comparing stock to overclocked, when if you overclocked the FX it would probly be able to be faster again.

Scorched any luck Overclocking any more? Need to get your 3dmark up a bit.

I havnt done any overclocking today, but I did rewire my system. Also, all my 3Dmarks are 100% stock, no video/CPU overclock.
 
Originally posted by Sc0rched
Yes but your comparing stock to overclocked, when if you overclocked the FX it would probly be able to be faster again.



I havnt done any overclocking today, but I did rewire my system. Also, all my 3Dmarks are 100% stock, no video/CPU overclock.

I'm comparing a Pentium 4 @ 3.6Ghz to a FX-51 @ 2.4Ghz. Overclocks much over 200MHz are sorta rare for the A64 lineup. Tomshardare got 2.8GHz on one but with extreme cooling. I think Aces Hardware was also able to get one up to those speeds. You've got one, push it to see how far it will go lol. Getting 3.5-3.6Ghz out of a $270 3.0c Pentium 4 is a piece of cake on good cooling.
 
Don't compare Stock to overclocked. If you're going to compare cpu's...

A64 FX-51 vs. P4 3.2EE
A64 3200+ vs. P4 3.2C
A64 3000+ vs. P4 3.0C

Whats the use of comparing an oc'd 3.0C to stock A64 FX-51. They are two different processors targeted for different categories. You might as well compare a Athlon XP 2500+ @ 2.5ghz to a P4 3.2EE.
 
Originally posted by burningrave101
I'm comparing a Pentium 4 @ 3.6Ghz to a FX-51 @ 2.4Ghz. Overclocks much over 200MHz are sorta rare for the A64 lineup. Tomshardare got 2.8GHz on one but with extreme cooling. I think Aces Hardware was also able to get one up to those speeds. You've got one, push it to see how far it will go lol. Getting 3.5-3.6Ghz out of a $270 3.0c Pentium 4 is a piece of cake on good cooling.

Go read www.xtremesystems.org forums.
 
Originally posted by pduan87
Go read www.xtremesystems.org forums.

I hope you realize alot of those crazy overclocks are fakes. This one is from a japanese guy off that site that supposedly overclocked a Pentium 4 over 5Ghz. People from here were boasting about it when i posted the link to the tomshardware 5.25Ghz overclock.

http://members.jcom.home.ne.jp/3233730401/WS000123.JPG

Its pretty easy to photochop benchmark results with paint shop or photoshop.
 
Ok, I experimented some with the overclocks. I was going by 5s up the FSB and got up to 215 stable, windows wouldnt boot at 220. I didnt go back down until it worked again though, got to lazy. Now ASUS needs to release a BIOS that gives me ability to change my multipyers so I can do higher.
 
Originally posted by [k]ar|
I was reading this link about the current AMD lineup and the anticipated socket 939 mobos.

Does anyone know when these will be coming out? More importantly, should I buy now or wait?

Thanks in advance; interested to hear some opinions.

[k]

The FX line is for exotic hardcore enthusiasts and WILL be phased out for the consumer line of 64chips which will be socket 939. So if you're looking for a piece of history get the FX otherwise, upgrade options are more likely with socket 939.
 
The FX line will continue after the Socket 939 transition. The distinguishing features, I presume, will still be the dual channel controller and the user adjustable multiplier. Just think about it; the only things that are changing during the transition will be one pin and the dropping of registered RAM as a requierment.
 
Originally posted by DaveX
Don't compare Stock to overclocked. If you're going to compare cpu's...

A64 FX-51 vs. P4 3.2EE
A64 3200+ vs. P4 3.2C
A64 3000+ vs. P4 3.0C

Whats the use of comparing an oc'd 3.0C to stock A64 FX-51. They are two different processors targeted for different categories. You might as well compare a Athlon XP 2500+ @ 2.5ghz to a P4 3.2EE.

And why not?

"Before we pass over to the actual benchmarks results, we decided to undertake a few overclocking attempts to figure out the overclocking potential of the new AMD Athlon 64 3000+ processors. The thing is that the CPUs of this price range are usually considered a good buy for overclocking purposes. That is why it would make a lot of sense to add the performance rates for the overclocked AMD Athlon 64 3000+ processor to the benchmark results."

http://www.xbitlabs.com/articles/cpu/display/athlon64-3000_4.html

Tomshardware and X-bit Labs both do this and so do countless other hardware sites. The only time someone has a problem with it is when the fanboy's cpu is on the losing side because of it.

The distinguishing features, I presume, will still be the dual channel controller and the user adjustable multiplier. Just think about it; the only things that are changing during the transition will be one pin and the dropping of registered RAM as a requierment.

"At first I would like to say a few words about the overclocking friendly features of this CPU. Since Athlon 64 3000+ is based on the same core as Athlon 64 3200+, it doesn’t allow increasing the clock frequency multiplier beyond the nominal 10x, just like its elder brother. However, you can set the clock frequency multiplier to a lower value, though it hardly makes much sense for overclockers."

You'll be able to lower the multiplier on any of the A64's but the ability to raise it past its nominall value is only available on the more expensive FX line.
 
First up, thanks for all the responses, guys. I think I'm going to take the plunge when my DD Maze4-64 arrives next week.

Sc0rched - I think you have the exact system that's been on my shopping list - Asus SK8v, FX51, 1GB Corsair RAM & 9800XT. How are you finding it generally? It will be a pretty good step up (I hope!) from my current rig, and should get a useful low-cost mid-life boost when MS finally get WOW64 out the door.

I was considering a "low cost" option of k8v deluxe & A64-3200, but I don't want to get locked into the 'Duron64' track with socket 754, and the full-fat FX upgrade will only be £250 more expensive all told. Besides, I don't like doing things by halves :D

Now, what to do with my current system...

[k]
 
I love my system, only complaint I have with it really is that the SK8V is only on 1001 BIOS which doesnt have the option to change multipliers. Im still tweaking it though trying to get it to do even better. You should be really set with the DD watercooling. With your current system you could use it as a file server, you could fold with it, or you could hook it up to a TV and use it as a media center. Good Luck!
 
Originally posted by burningrave101
You'll be able to lower the multiplier on any of the A64's but the ability to raise it past its nominall value is only available on the more expensive FX line.
Isn't that common knowledge? Maybe I'll be more specific next time.
 
Originally posted by burningrave101
I hope you realize alot of those crazy overclocks are fakes. This one is from a japanese guy off that site that supposedly overclocked a Pentium 4 over 5Ghz. People from here were boasting about it when i posted the link to the tomshardware 5.25Ghz overclock.

http://members.jcom.home.ne.jp/3233730401/WS000123.JPG

Its pretty easy to photochop benchmark results with paint shop or photoshop.

You post links from toms and then say extream posts fake pics :rolleyes:
 
Originally posted by skratch
You post links from toms and then say extream posts fake pics :rolleyes:

More nOOb fanboy comments :rolleyes: .

Tomshardware is just as good of hardware site as all the rest and they put out more articles then all the rest.

And if you guys want to be ignornant and believe all those scores and overclocks on extream systems then go ahead lol. But if you knew what you were doing you would know a photochop when you seen it.

Sometimes they can be edited pixel by pixel perfectly ands its near impossible to tell.

I think i'd believe something off a reputable hardware site as being true and not a fake photochop before i'd believe everything i seen on a forums based site with a bunch of end users trying to make their epenis look bigger then it really is.
 
Originally posted by burningrave101
Tomshardware is just as good of hardware site as all the rest and they put out more articles then all the rest.

Of course Toms Hardware puts out way more articles because most of them aren't even computer hardware related. I for one wouldn't want to depend on articles from a site that does a lot of stuff on digital cameras, printers, scanners and junk. Whenever they do put out a computer hardware review, the results are always predictable. Intel beats AMD, ATI and nVidia are "dead even".
 
I have to agree with some of the things about Toms hardware. There are many sites that I would put before it with terms of reliability. Such as the Athlon64 vs. 3.2C review he did. Now ask me how the 3.2C whoops up on the FX-51 when every other magazine, site, and computer company gets it opposite. I just find that alot of his reasonings and conclusions are opposite from the rest of the crowd, and it isnt just this review either.
 
Originally posted by burningrave101
More nOOb fanboy comments :rolleyes: .

Tomshardware is just as good of hardware site as all the rest and they put out more articles then all the rest.

And if you guys want to be ignornant and believe all those scores and overclocks on extream systems then go ahead lol. But if you knew what you were doing you would know a photochop when you seen it.

Sometimes they can be edited pixel by pixel perfectly ands its near impossible to tell.

I think i'd believe something off a reputable hardware site as being true and not a fake photochop before i'd believe everything i seen on a forums based site with a bunch of end users trying to make their epenis look bigger then it really is.

Stop being ignorant. Toms Hardware is shit. They are extremely biased and they are not professional at all. Did you see the ss that was posted on Toms but then taken off? It showed a P4 clocked at 13k mhz. I must say, I have no respect for Toms with his constant AMD bashing and Intel whoring. Tom's o/c was shit anyways. He didnt do even ONE BENCHMARK. As for the people over at xtreme, they are very pro. I'm assuming they make up all their benchmarks as well right? Superpi, sandra, 3dmark, CPUZ, Pifast, etc. etc. All of those benchmarks are made up right? Just so they have bigger e-penises? Have you even seen some of their rigs? Seriously, stop talking out of your ass and go over their and learn.
 
One of them recently broke 30k in 3dmark... How can you photoshop an ORB on 3dmarks site?
 
Originally posted by burningrave101
More nOOb fanboy comments :rolleyes: .

Tomshardware is just as good of hardware site as all the rest and they put out more articles then all the rest.

And if you guys want to be ignornant and believe all those scores and overclocks on extream systems then go ahead lol. But if you knew what you were doing you would know a photochop when you seen it.

Sometimes they can be edited pixel by pixel perfectly ands its near impossible to tell.

I think i'd believe something off a reputable hardware site as being true and not a fake photochop before i'd believe everything i seen on a forums based site with a bunch of end users trying to make their epenis look bigger then it really is.


Your kidding me right im a noob to this site.Pease just give it up with this crap about extream just go look at the orb and see how many are in the top 10.

This wouldnt even be talked about on there forum.Your telling me extreme posts fakes when you read toms the biggest intel paid tabloid online.

And when you spend 10k on cascade cooling maybe youll start posting over there.
 
Originally posted by Sc0rched
One of them recently broke 30k in 3dmark...

Good ol' Oppainter using his FX-51 @ 2.8ghz w/ Mach II cooling...
 
Originally posted by DaveX
Good ol' Oppainter using his FX-51 @ 2.8ghz w/ Mach II cooling...

Tomshardware just did an overclock on the FX-51 and got over 2.8GHz also on it. Tomshardware isn't bias, AMD fanboys are bias. I personally know several of the guys that does the reviews and i can tell you their not bias. If they were Intel and Nvidia bias then ATI wouldn't of just won every benchmark in their new vga roundup.

If you guys would look at the stock benchmarks in their new Intel vs AMD then the A64's win nearly every time. It wasn't toms fault that most of the AMD A64 boards right now are shit and chipset limited. Most people are not going to get over a 200MHz overclock out of these processors and thats a fact. Look at the majority of the reviews on these processors out there. All the A64 3000+ is is a crippled 3200+ that couldn't perform at that processors rating so they decide to shut off half the cache to make it work and sell it.

Not every post on extreme is fake but most of the outrageous ones are. This japanese guy thats suppost to of gotten over 5ghz on several intel chips is the biggest fake out of all of them. And he was posted numorous times on extreme because it was a big deal.

http://members.jcom.home.ne.jp/3233730401/WS000123.JPG

If you guys can't tell thats a fake your morons because you dont even have to look twice at that one. A 1.6ghz willamette running at over 5ghz? lol I dont hardly think so :rolleyes: .

Tom's o/c was shit anyways. He didnt do even ONE BENCHMARK. As for the people over at xtreme, they are very pro.

Its impossible to benchmark at those speeds . The top completely stable speed was 4.7GHz which is still more then i've seen completely stable and legit. And it wasn't a article to see what the top stable speed was, it was to see how far you can push one.

You guys obviously dont know how easy it is to photoshop any results you could possibly want into a benchmark or overclock. And if your good at it you can't even tell its been done.

Theres a full video on the tomshardware 5Ghz overclock.

Tomshardware reports the truth that noone wants to believe. Nvidia 's video cards are great choices now. They've completely fixed the image issues with their new compiler and the 5900SE which is the same price range as the ATi 9600XT kicks the living shit out of it for a mid-range card. The 5950 also competes very well against the 9800XT and wins in several benches. ATi's top of the line 9800Xt is definately the top video card out right now though and you can see that in the nvdia bias benchmarks on toms :rolleyes:.

Most of you are going off word of mouth and have no idea what your talking about. I'm not saying tomshardware hasn't screwed up in some of their reviews cause i know they have but to say their intel and nvidia bias because you dont want to believe the results is bull shit. Maybe the site your going off of for results is bias or doesn't know what their doing for the intel side of the benchmarks.

And toms is the biggest intel paid tabloid? Last time i checked there was alot more A64 adds on the site then Intel.

I'm assuming they make up all their benchmarks as well right? Superpi, sandra, 3dmark, CPUZ, Pifast, etc. etc. All of those benchmarks are made up right? Just so they have bigger e-penises? Have you even seen some of their rigs? Seriously, stop talking out of your ass and go over their and learn.

I think your the one that needs to go learn lol. I never said those benchmarks are fake, i said any moron can learn how to take the results with those benchmarks and edit the pic perfectly in photoshop. Most of them are nothing but photochops especially those 5GHz Intel overclocks.

I have to agree with some of the things about Toms hardware. There are many sites that I would put before it with terms of reliability. Such as the Athlon64 vs. 3.2C review he did. Now ask me how the 3.2C whoops up on the FX-51 when every other magazine, site, and computer company gets it opposite. I just find that alot of his reasonings and conclusions are opposite from the rest of the crowd, and it isnt just this review either.

When did the 3.2c whoop up on the FX-51? The only one that was whooping up on the FX-51 was the P4EE and it is the best 32bit processor out there. Especially when its an easy overclock to 3.6Ghz with good cooling. 2.8GHz on an FX-51 is a crap shoot with extreme cooling involved. The Pentium 4's are a much better choice for all around prcoessors. Sure your gaming performance will go up a little on an FX-51 but load and boot times for the OS and applications will stay the same. Why do you think all the major OEM's like Dell and Gateway wont use these new A64's? AMD used to have a great performance/price ratio but with these new A64's its no longer there. The Pentium IV's are a much better performance/price ratio. AMD can't even manage to get out their flagship chips on their own chipset. They have to have someone else do everything for them.

Of course Toms Hardware puts out way more articles because most of them aren't even computer hardware related. I for one wouldn't want to depend on articles from a site that does a lot of stuff on digital cameras, printers, scanners and junk. Whenever they do put out a computer hardware review, the results are always predictable. Intel beats AMD, ATI and nVidia are "dead even".

Almost every single review is computer hardware related. Sure they put out a couple reviews now and then on games and other such things but they almost always have a new hardware review every couple of days.
 
i wonder how that guy got a 2 meg L3 cache on his williamette...?
And i have to say, although i do agree with burnin, many of the other people's points are vaild too. I am also annoyed when i go to toms and all their new articles are printer roundups:eek: , but most of their other stuff is pretty valid and not biased. They did screw up big time on the ee vs fx review, so that is a fine reason for them to loose some cred, if you go to the toms forum there was a huge thread complaining about that.
In plently of their articles, i have seen them show ati on top, (although i also noticed how when ati won the comment would be something to the effect of "this prog was optomized for ati cards as we can see" but when an nvidia card won, they would say "here we can see the true potential of x" or "fx... is nearly 50% faster than radeon... in the benchmark".)

But i still visit their site, and read and 99% trust their reviews.
 
Originally posted by burningrave101
bullshit bullshit bullshit...

You are an idiot. STFU. The 1.6 P4 is as bug. It was really a P4 EE but since the WCPUID version was old, it can't read that. You wouldn't know that would you?

Tom's is biased. Period. As the person said above, every single review, Intel beats AMD. However, the rest of the internet says the other way around.

None of the oc's are fake. You are just a moron. Do you have any proof that they are fake or are you just jealous that you can't ever reach those speeds? So you are telling me, that the o/c's are fake even though they have benchmarks to back it up? Are you seriously talking out of your ass?

Tom's oc was shit considering that he got a modded ASUS board and a HAND PICKED 3.4 P4EE. The gap between stable and highest o/c should be less considering they were using LN2 cooling. Did you see how he applied the thermal paste? Like FUCKING GLUE ON A 5 YEAR OLDS CRAFT PROJECT. People over at xtreme have gotten 5ghz+ with benchmarks.

A freaking 2.8 Athlon FX would KILL a P4EE 3.6. You yourself is a damn Intel fanboy and you cant admit that your beloved P4EE has been beaten.

No the P4's are not a better choice. Ever heard of the Athlon 64 3000+?

Only good card right now for Nvidia is the 5900nu. However, 9800np will still kill that card. That card is on par with maybe a 9700. The reason why the benchmarks are so close between 5950u and 9800xt is because Tom is biased.

Seriously dude, stop making yourself look like a moron. I dare you to go over to xtremesystems.org forums and call all their benchmarks fake. You will get your ass flamed to oblivion.
 
Guys, guys, guys... calm down! :D

I only came to ask whether I should buy or wait, but it seems I accidentally started a flame war concerning (so far as I can tell) Intel v AMD, nV v ATI and THG & Extreme v everyone :eek: . Lets try to keep the discussion mature instead of descending into flames and insults; however, seeing as we're already there, I guess I'll have to join in before we get locked.

My unwanted 2c:

This is the Strictly AMD forum, and we're discussing whether an FX-51 & socket 940 is preferable (performance wise) to waiting for socket 939 and the FX that will go with it.
  • If you're pimping Intel, get lost. I've never bought a Pentium and I'm not starting now
  • If you're getting into nV vs ATI, get lost. The 9800XT is King ATM so that's what I'm getting
  • If you're going to bash/pimp other sites, get lost. I come to the [H] for reasoned informed opinions and a knowledgeable user base. I don't care about Extreme forum and their benchmarks (real or not). I don't like/respect THG so I don't go there any more. It's as simple as that. I don't bash - I just don't click.
So please can we stop the flames and get back on-topic - namely whether the performance improvement in socket 939 over 940 is worth a 3-month wait.

Thanks

[k]
 
IMO, There wont be a performance gain on the new ones other then the pure Mhz speed improvement. All they are doing is getting rid of the registered memory and putting it on a smaller die. Unless I am missing somthing the 3 month wait will just improve clock speeds. With that happening the price will go down, except I dont see it slipping to much since that will still be thier top of the line processor and they are in it for a profit. So I see the new FXs to be about $400 and the A64s to be about $250 with the new die and socet 939. However that is just my best guess from experiences, and since AMD has not realeased a price yet, to my knowledge, then I would say go for it. Thats what I did after months of deciding. I have had no problem with mine, I just need to get a Raptor for it to get the HD speed up. It plays all games with high res, all settings on high with no lag, including Halo, one of the slower games. I just cant wait for 64bit apps, OS, and drivers to see this thing really shine. I should be installing Mandrake soon to see how 64bit linux works. I got the files sitting on my computer just need to get some CDs to burn it to, luckly im going to best buy today to return a cold cathode(stupid RED cathode being bright PINK, BAH) so I might pick up some CDs and tell you how it goes.
 
You forgot the other big advantage to dropping registered RAM. The dropping of the registered RAM requirement will result in up to a 5-10% performance gain in some benchmarks. Some benchmarks show the Athlon 64 3200+ performing close to or better than the Athlon 64 FX-51. This is indicative of the registered RAM's increased latency. When this limitation is dropped, you will automatically see a marginal to significant increase in performance without even considering clock frequency increases.
 
Originally posted by pduan87
Some serious AMD fangirl trolling bullshit...


You are an idiot. STFU. The 1.6 P4 is as bug. It was really a P4 EE but since the WCPUID version was old, it can't read that. You wouldn't know that would you?

He was claiming that over 5Ghz overclock over the summer around August. Was there a 3.2EE then? lol

None of the oc's are fake. You are just a moron. Do you have any proof that they are fake or are you just jealous that you can't ever reach those speeds? So you are telling me, that the o/c's are fake even though they have benchmarks to back it up? Are you seriously talking out of your ass?

Oh i can reach the same overclocks as they did and the same way too. I'm pretty good with photoshop :D.

http://www.overclockers.com/articles444/

Thats just one of the articles on the net talking about benchmark faking today. Its easy to do and lots of people do it. So dont be a noob and believe everything you see just because you dont know any better.

Tom's oc was shit considering that he got a modded ASUS board and a HAND PICKED 3.4 P4EE.

Got a modded Asus board? They didn't GET anything. They added some MOSFOTS and modded it slightly themselves so it would have more phases so they could get more voltage to the processor. The Asus board is only a 3 phase board while some of the other higher end boards like the Max 3 are up to 4 phase.

The 3.4EE was an engineering sample. The same sample Intel sends to everyone else that gets them like x-bit Labs and Anadtech, and Ace's Hardware. They were just lucky and got ahold of a 3.4EE a little early.

A freaking 2.8 Athlon FX would KILL a P4EE 3.6. You yourself is a damn Intel fanboy and you cant admit that your beloved P4EE has been beaten.

No the P4's are not a better choice. Ever heard of the Athlon 64 3000+?

Like i said getting 2.8GHz out of an FX-51 is not easy and you have to have really good cooling. You can get 3.6Ghz out of a P4EE any day of the week on normal aircooling.

The 3000+'s are handicaped 3200's that couldn't perform at 3200+ specs so they got half their die cut to make them able to run. So in other words they have very little overclocking potential. Around 2.2GHz is the most people are able to get out of them. A 3000+ overclocked doesn't even have a chance at performing better then a Pentium 4 at 3.5Ghz.

At stock, the majority of benchmarks shows the 3000+ performs on par with a Pentium 4 @ 3.2GHz.

http://www.anandtech.com/cpu/showdoc.html?i=1937&p=5

At stock speeds the Pentium 4 3.2 beat the 3000+ in 10 out of 20 of
the benchmarks.

http://www.xbitlabs.com/articles/cpu/display/athlon64-3000_5.html

At stock speeds the Pentium 4 3.2 beat the 3000+ in 15 out of 28 benchmarks.

Its pretty 50/50 as far as performance goes between these two processors. The thing everyone needs to take into account is that a 2.8c or 3.0c can easily do 3.5Ghz and higher. The A64's on the other hand are extremely hit and miss so you can't expect anything on any of them over 200MHz.

Has AMD Castrated Overclocking?

If you're an overclocking freak, yes. With its latest release of the 64bit processors, mainly the Athlon 64 and the Athlon 64 FX, AMD is not allowing much room to the 'push for more' world. It's quite sad for a lot of us who enjoy looking for our CPUs upper limits--hitting them feels so darn cool, ya know.

AMD's latest processors do have some new architecture to handle certain instructions differently, that is a given. But have they actually put a stop to the happy overclocking madness? Well, more or less, and a lot of people are quite disappointed.

http://www.ipkonfig.com/Articles/AMDCastrated/

Only good card right now for Nvidia is the 5900nu. However, 9800np will still kill that card. That card is on par with maybe a 9700. The reason why the benchmarks are so close between 5950u and 9800xt is because Tom is biased.

I guess you haven't heard of the 5900SE(XT) then huh. It overclocks to 5950 speeds. And on par with a 9700? The 5950 beats 9800 pro's in alot of the benchmarks with their new driver compiler for directX 9. The reason the benchmarks are so close isn't because toms is biased. You need to read more video card reviews and up to date ones at that. I suggest you go to the video card forum here and catch some of the reviews being talked about so you can get up to speed on things.

Seriously dude, stop making yourself look like a moron. I dare you to go over to xtremesystems.org forums and call all their benchmarks fake. You will get your ass flamed to oblivion.

Why would i bother going to a photochop forum when i can post right here and get all the flames i need from AMD fangirls like yourself :p ?
 
Originally posted by burningrave101
He was claiming that over 5Ghz overclock over the summer around August. Was there a 3.2EE then? lol



Oh i can reach the same overclocks as they did and the same way too. I'm pretty good with photoshop :D.

http://www.overclockers.com/articles444/

Thats just one of the articles on the net talking about benchmark faking today. Its easy to do and lots of people do it. So dont be a noob and believe everything you see just because you dont know any better.



Got a modded Asus board? They didn't GET anything. They added some MOSFOTS and modded it slightly themselves so it would have more phases so they could get more voltage to the processor. The Asus board is only a 3 phase board while some of the other higher end boards like the Max 3 are up to 4 phase.

The 3.4EE was an engineering sample. The same sample Intel sends to everyone else that gets them like x-bit Labs and Anadtech, and Ace's Hardware. They were just lucky and got ahold of a 3.4EE a little early.



Like i said getting 2.8GHz out of an FX-51 is not easy and you have to have really good cooling. You can get 3.6Ghz out of a P4EE any day of the week on normal aircooling.

The 3000+'s are handicaped 3200's that couldn't perform at 3200+ specs so they got half their die cut to make them able to run. So in other words they have very little overclocking potential. Around 2.2GHz is the most people are able to get out of them. A 3000+ overclocked doesn't even have a chance at performing better then a Pentium 4 at 3.5Ghz.

At stock, the majority of benchmarks shows the 3000+ performs on par with a Pentium 4 @ 3.2GHz.

http://www.anandtech.com/cpu/showdoc.html?i=1937&p=5

At stock speeds the Pentium 4 3.2 beat the 3000+ in 10 out of 20 of
the benchmarks.

http://www.xbitlabs.com/articles/cpu/display/athlon64-3000_5.html

At stock speeds the Pentium 4 3.2 beat the 3000+ in 15 out of 28 benchmarks.

Its pretty 50/50 as far as performance goes between these two processors. The thing everyone needs to take into account is that a 2.8c or 3.0c can easily do 3.5Ghz and higher. The A64's on the other hand are extremely hit and miss so you can't expect anything on any of them over 200MHz.



http://www.ipkonfig.com/Articles/AMDCastrated/



I guess you haven't heard of the 5900SE(XT) then huh. It overclocks to 5950 speeds. And on par with a 9700? The 5950 beats 9800 pro's in alot of the benchmarks with their new driver compiler for directX 9. The reason the benchmarks are so close isn't because toms is biased. You need to read more video card reviews and up to date ones at that. I suggest you go to the video card forum here and catch some of the reviews being talked about so you can get up to speed on things.



Why would i bother going to a photochop forum when i can post right here and get all the flames i need from AMD fangirls like yourself :p ?

WOOOOOW.... just stfu at this moment. They did get a handpicked p4 from intel and the asus was especially modded. Dude stop arguing here and get your ass handed to you over at xtremesystems.
 
Originally posted by pduan87
WOOOOOW.... just stfu at this moment. They did get a handpicked p4 from intel and the asus was especially modded. Dude stop arguing here and get your ass handed to you over at xtremesystems.

You have nothing to back up what you have said but your fanboy witts lol.

Wheres the law against moddifying the motherboard anyways? All they did was give it an extra phase to make a 4 phase mobo like some of the other top overclocking boards. Anyone with the intelligence in that area could do the same thing to get better overclocks.

The only thing you got against tomshardware is your personal opinions so if you wanna have a PERSONAL OPINION about something go right ahead but since your personal opinions have no basis of facts their worthless.

LOL @ http://www.xtremesystems.org/forums/showthread.php?threadid=25901
 
Status
Not open for further replies.
Back
Top