Can AMD salvage QFX with an in-house chipset?

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
IMHO....

AMD's 4x4 CPU pricing is very agressive. They should be able to compete on this level, but Asus and nVidia are milking cash from 4x4 on the motherboards.

The motherboard is cool, and very interesting, but over the top and overpriced. Nobody needs 20 USB2 ports, this is what hubs are for. I can't imagine many 4x4 owners will ever use all 12 SATA ports unless they are doing some massive RAID array, in which case they should be using a dedicated controller.

Were the motherboard more reasonable (and I see no reason why 1 nForce 680a wouldn't work), then plenty of people would buy an FX-70 system just because it would be the cheapest 4 core option.

There is a certain level of "prestige" in having 4 CPU cores, and this is the main market AMD and Intel are currently aiming their Quad core/4x4 offerings to right now. No games need 4 cores, and very very few home users would ever use a QX6700 or FX-74 4x4 system to its potential. Its just for show!!!

This is the same market that drops $20k on an Alienware preconfigured 4x4 Quad SLi system. The same market the 7900GX2 was aimed at before the release of the 7950GX2.

As for power usage, I care about this in only one regard : cooling however many watts from 1 tiny CPU is awkward.

The 4x4 system may be talking about 250W for the CPUs, but thats actually 125W in two different places. This is much easier for me to cool than 250W in one CPU - I use two coolers!

It seems 4x4 offers very little overclocking headroom. While this is a bad thing imho, it does mean that almost all 4x4 systems will be running very close to stock. It wouldn't surprise me if most QX6700 owners had their CPU around the 3.3-3.6GHz mark, at which point the CPU is probably approaching 175-200W - very awkward to cool from the small surface area of the IHS.

Basically my points are:

1. Yeah, C2Q is ~20%-30% faster, but as C2D was ~20-30% faster than A64 X2, what did you expect??

2. The FX-70 is very nicely priced, with a cheaper (in house?) motherboard AMD would be very competitive here on price.

3. Yeah 250W is alot, but as you cant overclock it much anyway you may as well just use the stock HSFs and its no hassle to cool. You might need a slightly better PSU, but with R600, G80, etc etc you more than likely need a new one anyway 😛

I wouldn't buy 4x4 however. QX6700 looks more appealing to me. (partly because I want to overclock and don't care about price) But that doesn't mean I don't think 4x4 looks pretty cool anyway :)
 
Heck the likely do this too to get those powers down, since they are pushing a 2.6 GHz opteron, the volume will be very low and the price they can get will be relatively high....

The point is, it is nothing special like you are so ignorantly trying to say. In fact, you are fairly dimwittedly stupid.
Click to expand...


When rational discussion fails, try name calling. You're now saying, I guess, that you didn't say they used 100nm spacing to get the power down? Just answer the question.
 
Can AMD salvage QFX with an in-house chipset?
Click to expand...
No. QFX is not salvagable
QFX is hot, noisy, expencive and pathetic


No, it's sort of like the first rev of Core 2. Remember it needing a new Rev before volume release? Remember all the problems with the BIOS'?

Stop being so negative. Anand got two chips AND a 8800GTX down to 456W, which is comparable to 955EE. He made no mention of noise, but I held my head next to the AlienWare demo machine and didn't notice excessive noise or exhaust heat.
 
- Quoting entire posts within the same scrollable region is kinda annoying.

Anyways, as I don't want people scrolling over the same repeated text over and over, here is my reply:

http://users.on.net/~darkpeace/forum_posts_long/BaronMatrix - Msg 56 - Can AMD salvage QFX with an in-house chipset _.pdf

If he was sever oriented he would be running an Opteron 2000 by now. - Something with Registered ECC DIMMs at least. :lol:
Click to expand...


I don't need PCI-x or most server features. I am a developer and I want to run several VMs for client-server tests. I don't think I'll be spending twice the money for RAM that gives no advantages over regular.
 
I don't think I'll be spending twice the money for RAM that gives no advantages over regular.
Click to expand...

Now BM, I'm not an expert so I'll leave it to someone else to explain exactly how ECC RAM is better, but c'mon - if what you say all the time is true and AMD are so increadibly wise, and let's assume for a second that ECC doesn't make a difference, then why on earth would such a fantastically wise corporation use more expensive RAM that wasn't any better?
 
The advantage of ECC Ram is reliability of data.
In fact such Ram stores redundant information and is able to auto-correct small errors (1 bit per byte) and detect more.
This is crucial for server / workstation apps, but it's typically not considered necessary for typical desktop usage.
 
Intel shows better scaling in the 'VRAD Map Compilation' benchmark, but AMD shows better scalling in the 'Particle Systems Test' benchmark. Hardly an advantage to either camp in terms of pure scaling.

However, regardless of scaling, it is clear that Intel is significantly faster than AMD in both tests.

The only reason it looks like AMD scales better by looking at the graphs is because you are looking at the scaling of a 2.8GHz FX-62 vs 3GHz FX-74, whereas with Intel you are looking at a 2.93GHz X6800 vs 2.66GHz QX6700.

Notice how I calculated scaling by comparing at the same clockspeeds?

Paints a whole different picture, doesn't it Baron?

Myth debunked! Nothing to see here people, just AMD fanboy dribble. Move along now...
Click to expand...

You're a jackass. If you look at my numbers on Page 3 you will see I am comparing clock for clock 5200+ - FX70 and FX62 - FX72.
 
I don't think I'll be spending twice the money for RAM that gives no advantages over regular.
Click to expand...

Now BM, I'm not an expert so I'll leave it to someone else to explain exactly how ECC RAM is better, but c'mon - if what you say all the time is true and AMD are so increadibly wise, and let's assume for a second that ECC doesn't make a difference, then why on earth would such a fantastically wise corporation use more expensive RAM that wasn't any better?

Because it's for servers and the data consistency is more important. Do you think Intel, IBM and Sun are NOT using ECC?
 
Because it's for servers and the data consistency is more important. Do you think Intel, IBM and Sun are NOT using ECC?
Click to expand...

But if you say that are you not contradicting what you put here:

I don't think I'll be spending twice the money for RAM that gives no advantages over regular.
Click to expand...

There's no contradiction. I won't be using a server so I don't need ECC.
 
There's no contradiction. I won't be using a server so I don't need ECC.
Click to expand...

So on one hand you say "I'm not going to buy ECC RAM because it's twice as expensive and doesn't make a difference" and on the other you say that ECC has better data consistency. Apparently that doesn't fall under the category of contradiction anymore. Cheers for letting me know.

I think I know 'what you meant', but please Baron, at least give yourself half a chance by not relying on people 'reading between the lines', otherwise someone might mistake you for someone who's a lil bit stupid.
 
Intel shows better scaling in the 'VRAD Map Compilation' benchmark, but AMD shows better scalling in the 'Particle Systems Test' benchmark. Hardly an advantage to either camp in terms of pure scaling.

However, regardless of scaling, it is clear that Intel is significantly faster than AMD in both tests.

The only reason it looks like AMD scales better by looking at the graphs is because you are looking at the scaling of a 2.8GHz FX-62 vs 3GHz FX-74, whereas with Intel you are looking at a 2.93GHz X6800 vs 2.66GHz QX6700.

Notice how I calculated scaling by comparing at the same clockspeeds?

Paints a whole different picture, doesn't it Baron?

Myth debunked! Nothing to see here people, just AMD fanboy dribble. Move along now...
Click to expand...

You're a jackass. If you look at my numbers on Page 3 you will see I am comparing clock for clock 5200+ - FX70 and FX62 - FX72.

So are you claming my maths is wrong? I'm comparing the scaling of both platforms on a clock for clock basis...

Your initial argument was that AMD shows better scaling than Intel going from DC to QC. I took the time to actually calculate the scaling figures which proves that scaling is actually quite even on both platforms, with one test favoring Intel and one test favoring AMD.
 
There's no contradiction. I won't be using a server so I don't need ECC.
Click to expand...

So on one hand you say "I'm not going to buy ECC RAM because it's twice as expensive and doesn't make a difference" and on the other you say that ECC has better data consistency. Apparently that doesn't fall under the category of contradiction anymore. Cheers for letting me know.

I think I know 'what you meant', but please Baron, at least give yourself half a chance by not relying on people 'reading between the lines', otherwise someone might mistake you for someone who's a lil bit stupid.


That makes them an a-hole. If someone wants to nitpick I will lose no sleep.
 
So are you claming my maths is wrong? I'm comparing the scaling of both platforms on a clock for clock basis...

Your initial argument was that AMD shows better scaling than Intel going from DC to QC. I took the time to actually calculate the scaling figures which proves that scaling is actually quite even on both platforms, with one test favoring Intel and one test favoring AMD.
Click to expand...

What is your malfunction numbnuts? Your calculation was the same as mine. It showed better than 100% scaling for QFX and 97% scaling for C2Q.
 
This is for Baron's benefit.... When discussing scaling, you do exactly what you did.... divide the faster metric by the slower to find the multiplicative factor by which is scales up..... this is why it is called scaling.
Click to expand...

So use Windows calculator and do the comparison yourself.

2.6GHz -- 111%
2.8GHz -- 107%

for the particle tests.
 
Can AMD salvage QFX with an in-house chipset?
Click to expand...
No. QFX is not salvagable
QFX is hot, noisy, expencive and pathetic


No, it's sort of like the first rev of Core 2. Remember it needing a new Rev before volume release? Remember all the problems with the BIOS'?
4x4 is not even close to what first rev of Core2 was. 4x4 is only a marketing name for a 2P K8 system. There are no inovations in 4x4. It uses the concept of MP HTT system and 3 years old K8 architecture. Unlike Core2 which is aimed for the masses and is conquering the x86 world, 4x4 is aimed for the die-hard very rich AMD fans(0.00000001% of total x86 users on planet Earth) .
Why 4x4 sux compared to Core2 quad?
1. it is slower
2. it is hotter
3. it wastes more energy
4. it overclocks less
5. it makes more noise
6. it needs bigger case with extra cooling and powerfull PSU
7. it is more expencive
than Core2 Quad

It is pathetic try of AMD to compete agianst what Intel allready have released.

Stop being so negative. Anand got two chips AND a 8800GTX down to 456W, which is comparable to 955EE. He made no mention of noise, but I held my head next to the AlienWare demo machine and didn't notice excessive noise or exhaust heat.
Click to expand...
I am not negative. I am just being objective. I don't see the things through the green galsses of "AMD is the best".
Get back to year of 2006, so stop comparing 4x4 to 965EE. Everybody knows that Netburst was crap, but there is no Netburst chip that can waste more energy than 4x4 can. Take out your head of the sand, and face it: 4x4 is a failure.

BTW, if you can't notice the noise of 10 fans from the AlienWare dem machine, than you should visit a doctor soon.
 
Can AMD salvage QFX with an in-house chipset?
Click to expand...
No. QFX is not salvagable
QFX is hot, noisy, expencive and pathetic


No, it's sort of like the first rev of Core 2. Remember it needing a new Rev before volume release? Remember all the problems with the BIOS'?
4x4 is not even close to what first rev of Core2 was. 4x4 is only a marketing name for a 2P K8 system. There are no inovations in 4x4. It uses the concept of MP HTT system and 3 years old K8 architecture. Unlike Core2 which is aimed for the masses and is conquering the x86 world, 4x4 is aimed for the die-hard very rich AMD fans(0.00000001% of total x86 users on planet Earth) .
Why 4x4 sux compared to Core2 quad?
1. it is slower
2. it is hotter
3. it wastes more energy
4. it overclocks less
5. it makes more noise
6. it needs bigger case with extra cooling and powerfull PSU
7. it is more expencive
than Core2 Quad

It is pathetic try of AMD to compete agianst what Intel allready have released.

Stop being so negative. Anand got two chips AND a 8800GTX down to 456W, which is comparable to 955EE. He made no mention of noise, but I held my head next to the AlienWare demo machine and didn't notice excessive noise or exhaust heat.
Click to expand...
I am not negative. I am just being objective. I don't see the things through the green galsses of "AMD is the best".
Get back to year of 2006, so stop comparing 4x4 to 965EE. Everybody knows that Netburst was crap, but there is no Netburst chip that can waste more energy than 4x4 can. Take out your head of the sand, and face it: 4x4 is a failure.

BTW, if you can't notice the noise of 10 fans from the AlienWare dem machine, than you should visit a doctor soon.

I've got an idea, don't buy it. You still haven't answered why they can't increase the perf with a newer chipset and drop power with a new Rev.

The MACHINE WASN'T OVERLY NOISY. Of course fans make noise. I think it's possible but either way the scaling for the Valve tests is impressive enough to warrant serious thought.

I mean 2X the VM power is not a joke for my purposes. I'll be doing my upgrade with the timing of Vista X64 and lower priced DX10 cards. Maybe an 80nm R600 will be a good thing for QFX.
 
No, it's sort of like the first rev of Core 2. Remember it needing a new Rev before volume release?
Click to expand...

Um yeah the first revision is never released at volume you moron.

Remember all the problems with the BIOS'?
Click to expand...

Making sh!t up again I see.

Stop being so negative. Anand got two chips AND a 8800GTX down to 456W
Click to expand...

Which is great compared to 263W, no wait thats terrible.

We missed you, action! The bullie of the forum. :)
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom