Name people who manufacture higher end processors than a Xeon?
Intel's FABs are very nice, so only really AMD & IBM I'd suspect as they are the only ones with the real drive to keep up with this sorta tech.
I think you'll find AMD's processors have far lower leakage than any 90nm Xeon though.
The combined market share of those doesn't touch the Xeon, and some of them may not even be as good, depending on what you're doing.
So? Windows has the most market share in the OS market, and I really don't think you'll fidn that is the best either.
The current Xeons are hot, slow and don't scale as well as Opterons (as the Opteron is NUMA). So why on earth would you want one?
. Intel's dual core is not a joke, it was at market first before AMD's, and it's still in its development.
Great! I'll go a buy a dual core Xeon! Oh, I can't, it isn't released till next year. Desktop only I'm afraid for the time being.
And AMD's still isn't in development? Intel's dual core is a joke too - why? Because the chips can't talk to each other directly, they talk over the FSB. Which is really slow, and blocks memory access. AMD's solution, the chips talk directly and can talk to the memory at the same time. The effectives are dramatic: See how AMD & Intel's processors scale in dual core.
Was the Pentium-III a joke? Everyone said it'd never catch on, because it was so pricey, and the 550 could scramble eggs. Boy, we were wrong.
Considering how badly the P3 did against the Athlon, yes it was a joke. It couldn't scale against it, and Intel did several paper launchs to save losing face.
Do you remember the 1.13GHz P3 they had to recall? It only came good when it became the Pentium-M, and it was redesigned.
Intel's Pentium-M is a fine chip, as it the Itanium 2 in it's own way. In the middle though, Intel has fallen behind AMD in several areas.
Are you sure? They stablised the C++ ABI back in gcc 3.2 (although 3.3 to 3.4 had a minor change to SPARC, for a good reason though).
The main purpose of the 3.2 release is to correct a couple of problems in the C++ ABI, with the intention of providing a stable interface going forward. Accordingly, 3.2 is only a small change to 3.1.1.
You dork. I said high end x86 boxen have lights out, and yes, I'm aware the HP makes some of said high end x86.
You did? Where? You said - Sun boxes are about the same cost as x86 boxes in the high end, and they have all the stuff you really need. 64-bit, lights out management - I'm sorry, but to me that implies that Sun boxes have lightsout but x86 boxes don't.
As for 8 ways and such, give me a break. Go to HP's website (http://www.hp.com/ select an 8 way system, and increase its RAM and drive configurations to the standard in the US-IV system. What's it come to? Here's a hint, it's pretty close to $150,000 depending on exactly what you consider to be precisely equivalent to the Sun boxen.
Oh I'm sorry, I thought you said - Your average 8 proc US-IV system (16 cores) from Sun costs about the same as an 8 proc (8 cores) Opteron system from HP - I think you'll find the word "Opteron" present. The most expensive Opteron box from HP is about $40k (with 8 cores). If you change that to "Intel", then of course it changes!! But that isn't what you said at the time.
Also, the US-IV systems are all dual core, so we're talking 16 cores. If you could buy a 16 core athlon machine with the sort of ram and Disks that the suns have, it wouldn't cost 39k, if you can get it at all, that is.
Sun themselves are designing one. Newisys have a chipset that'll allow 32 Opteron chips (64 cores) together. People do produce 16-way Opterons too, like IWill. No tier 1 uptake yet, but I think you'll see Sun & HP go down that route.
I don't doubt it'll cost more than $39k mind.
I never said the price for performance was equivalent, but spec out a few machines and see how close it is. It's the same in the 4-8 core realm (with the V40z), and it's about the same in the 16+ core realm due the difficulty in getting good 8+ core x86 systems. 48 cores of US-IV will be expensive, but there aren't a lot of x86 boxes out there that can touch it's whole-system performance. And yes, I too know what I'm talking about.
Oh yeah, big SMP boxes are pricy. But if your Opteron CPUs are twice the power of USIV CPU (quite likely I'm afraid, the USIV is nothing more than two of the older US3 core bolted together, and Opterons are nearly at double the clock speed. My tests show in real world stuff, Opterons are 33% quicker clock for clock than the US3+ & US3iii) then why on earth do you need to "match" the spec? You'd just save the money on a smaller Opteron box or just get a larger Opteron box for the same/less money anyway.
lights out management (you can discover problems in the hardware even after it has crashed, because it contains a little computer on a chip designed just to report the statte of the hardware, power cycle it, etc....)
Gee, 'cos it isn't like HP have lightsout, Dell have a remote access console and even Sun's own v20z/v40z have that. Of coruse, the reason Sun have it is because those boxes are Newisys reference designs, and they've put it in place.
Your average 8 proc US-IV system (16 cores) from Sun costs about the same as an 8 proc (8 cores) Opteron system from HP, for similar configurations.
But please, can I have some of what you are smoking!
This is just total crap. To get a 8-way US-IV, you need at least a V890. Which comes in at $155k!
Now, you say 8-way Opteron. No tier 1 makes them, but I presume you mean 4 seeing as you can get dual core Opterons only. Do they cost $155k? Ummm, no. They cost $39k from Sun and half configured (2 processor box) from HP cost $17k. Somehow, I can't see HP being that much different than Sun.
And when it comes down it, an Opteron is way faster than a USIV anyway so you don't even need that many processors. And yes, I do use these processors everyday so I do know what I'm talking about. Which apparently you don't.
If you really don't want to run Linux, you can of course run Solaris 10 on the v40z.
Um, the G4 core has always been close to twice as fast clock-for-clock than the Pentium 4.
Do you have any proof of that? Sure, the P4 isn't the most efficient processor ever but it isn't exactly that slow. Which revision of the P4? Williamette? Northwood? Prescott? You can't bundle them all together.
And no, photoshop benchmarks don't count.
Do you agree the G5 is more efficient that the G4? How much more efficient is the G5 than the P4?
Every newer OS X has run better than the previous version on these machines from my experience, and from what I've heard others say. Realistically, how long can that go on though until newer versions start to overwhelm older hardware?
The one thing is - does it run faster because the original code was so poor, or because the newer code is better?
Some speed-ups also come from new graphics drivers as they come out too mind (as the GUI becomes more "snappy"!).
How does what DVD Jon did reflect the 5 unique useres thing? As far as I can tell, Apple just put that in anyway. Not as if what DVD Jon did would affect it that much anyway.
The Itanium doesn't sell much, but I think you'll find 90%+ of the boxes involved run Linux. I believe some Windows stuff does exist for Itanium, but I don't know anyone who runs Windows on an Itanium.
As for SPARC - well, it does have it's own OS anyway made by Sun! I suspect Linux on Itanium will be in very high depend as it is pretty much dependent on it. (e.g. Intel will pay you lots!)
Erm, why don't you use 7.d instead where the whole process is basically automated? Run genkernel --menuconfig all if you wish to tweak which modules are available.
If all else fails, use ndiswrapper. It works fine for me.
w00t, the Mac Mini is newer therefore it's graphics card must be better. No, I disagree. The 9200 is the lowest graphics card ATi even sell these days. It may be better than the 5200, but then again that is also a budget card. The GF3 was a full blown top card when it came out.
Whether the Mac Mini does better than the X-Box or not, the point is a GF3 (Tweaked, got some GF4 tech in it for the X-box) is better than a 9200. The difference in architecture/processors is one thing, but to say the Mac Mini's graphics card is purely better is wrong, shared bus or not.
As I much as I hate to link to Tom's Hardware, I feel I must. So go here and read. The GF3 in the XBox is basically a Ti500, and the 9000 is the 9200 (9200 is 9000 + 8xAGP, like that makes any difference).
Oh look, I don't see the ATi card beat the Ti500 anywhere.
1. Charge $1 per hour of CPU time on your cluster. 2. Lower the speed of your processors. 3. Runtime of tasks increase. So your $1 does less. 4. PROFIT!!!!
I don't believe it you see. Whilst Sun can do that with SPARC, doing it was x86 is a totally different task. e.g. x86-64 has twice the number of registers as x86. Not exactly an ABI issue. This is something that is just going to require two copies of most things.
Solaris 8 does that with multiple copies of libraries, e.g./usr/lib &/usr/lib/64
I can remember reading something saying it wasn't out yet, but I can't find it now. Both you and an AC have told me otherwise, so maybe I'm wrong though?!?
The initial release of Solaris 10 is 32-bit only on x86, so you'll have to wait for release of a later version.
Likewise Project Janus also isn't included yet. ZFS wasn't included in Solaris Express, I'm not sure if it is now either. May expect a release in a few months to cover all these bases.
Slashdot Mirror
No, you fail it!
Intel's FABs are very nice, so only really AMD & IBM I'd suspect as they are the only ones with the real drive to keep up with this sorta tech.
I think you'll find AMD's processors have far lower leakage than any 90nm Xeon though.
So? Windows has the most market share in the OS market, and I really don't think you'll fidn that is the best either.
The current Xeons are hot, slow and don't scale as well as Opterons (as the Opteron is NUMA). So why on earth would you want one?
Great! I'll go a buy a dual core Xeon! Oh, I can't, it isn't released till next year. Desktop only I'm afraid for the time being.
And AMD's still isn't in development? Intel's dual core is a joke too - why? Because the chips can't talk to each other directly, they talk over the FSB. Which is really slow, and blocks memory access. AMD's solution, the chips talk directly and can talk to the memory at the same time. The effectives are dramatic:
See how AMD & Intel's processors scale in dual core.
Considering how badly the P3 did against the Athlon, yes it was a joke. It couldn't scale against it, and Intel did several paper launchs to save losing face.
Do you remember the 1.13GHz P3 they had to recall? It only came good when it became the Pentium-M, and it was redesigned.
Intel's Pentium-M is a fine chip, as it the Itanium 2 in it's own way. In the middle though, Intel has fallen behind AMD in several areas.
You'd best be careful about the iMac G5 - it has a far greater return rate than the rest of the Mac line due to overheating.
It may still work in 7 years time, then again, we've got Compaqs from 7 years ago where basically none of them have failed....
From GCC themselves.
Ok!
Here you go. It was on slashdot a while back.
You did? Where? You said - Sun boxes are about the same cost as x86 boxes in the high end, and they have all the stuff you really need. 64-bit, lights out management - I'm sorry, but to me that implies that Sun boxes have lightsout but x86 boxes don't.
Oh I'm sorry, I thought you said - Your average 8 proc US-IV system (16 cores) from Sun costs about the same as an 8 proc (8 cores) Opteron system from HP - I think you'll find the word "Opteron" present. The most expensive Opteron box from HP is about $40k (with 8 cores). If you change that to "Intel", then of course it changes!! But that isn't what you said at the time.
Sun themselves are designing one. Newisys have a chipset that'll allow 32 Opteron chips (64 cores) together. People do produce 16-way Opterons too, like IWill. No tier 1 uptake yet, but I think you'll see Sun & HP go down that route.
I don't doubt it'll cost more than $39k mind.
Oh yeah, big SMP boxes are pricy. But if your Opteron CPUs are twice the power of USIV CPU (quite likely I'm afraid, the USIV is nothing more than two of the older US3 core bolted together, and Opterons are nearly at double the clock speed. My tests show in real world stuff, Opterons are 33% quicker clock for clock than the US3+ & US3iii) then why on earth do you need to "match" the spec? You'd just save the money on a smaller Opteron box or just get a larger Opteron box for the same/less money anyway.
Gee, 'cos it isn't like HP have lightsout, Dell have a remote access console and even Sun's own v20z/v40z have that. Of coruse, the reason Sun have it is because those boxes are Newisys reference designs, and they've put it in place.
But please, can I have some of what you are smoking!
This is just total crap. To get a 8-way US-IV, you need at least a V890. Which comes in at $155k!
Sun's pricing
Now, you say 8-way Opteron. No tier 1 makes them, but I presume you mean 4 seeing as you can get dual core Opterons only. Do they cost $155k? Ummm, no. They cost $39k from Sun and half configured (2 processor box) from HP cost $17k. Somehow, I can't see HP being that much different than Sun.
v40z pricing
HP DL585 pricing
And when it comes down it, an Opteron is way faster than a USIV anyway so you don't even need that many processors. And yes, I do use these processors everyday so I do know what I'm talking about. Which apparently you don't.
If you really don't want to run Linux, you can of course run Solaris 10 on the v40z.
As we've gone through SDRAM to DDR and now to DDR-2 the voltages have actually gone down!
EDO - 5v
SDRAM - 3.3v
DDR - 2.7v
DDR-2 - 1.8v.
That is why having DDR-2 in your laptop is a good thing.
Do you have any proof of that? Sure, the P4 isn't the most efficient processor ever but it isn't exactly that slow. Which revision of the P4? Williamette? Northwood? Prescott? You can't bundle them all together.
And no, photoshop benchmarks don't count.
Do you agree the G5 is more efficient that the G4? How much more efficient is the G5 than the P4?
Apple Universal Binary Docs
It says Intel systems will not run on Open Firmware. It doesn't say it is a the "normal" PC BIOS either, or something else. Just not OF.
Ummm, you must read something different to me then.... Sun's server market dropped from 10.3% to 9.5% last quarter.
4th spot overall.
Free international shipping!
Would that be in SOVIET RUSSIA?? :)
Oh and slashdot's lameness filter is lame....
SPEC FP results.
Look, a Athlon FX-55 is nearly TWICE as fast at SPEC FP - when the clock speed is only 30% higher. SpecINT will be about equal looking at it.
The Pentium M is good, but outright performance isn't something it has.
The one thing is - does it run faster because the original code was so poor, or because the newer code is better?
Some speed-ups also come from new graphics drivers as they come out too mind (as the GUI becomes more "snappy"!).
How does what DVD Jon did reflect the 5 unique useres thing? As far as I can tell, Apple just put that in anyway. Not as if what DVD Jon did would affect it that much anyway.
Specifically, it shows two things (note, the clock throttling wasn't working on the Opteron processors mind):
The Itanium doesn't sell much, but I think you'll find 90%+ of the boxes involved run Linux. I believe some Windows stuff does exist for Itanium, but I don't know anyone who runs Windows on an Itanium.
As for SPARC - well, it does have it's own OS anyway made by Sun! I suspect Linux on Itanium will be in very high depend as it is pretty much dependent on it. (e.g. Intel will pay you lots!)
LinuxHardware
Lost Circuits
Notice how often AMD gain from running in 64-bit mode, where as Intel lose performance.
I've used WB & Tao, and yeah, they are essentially exactly the same as RH. Beyond the names, you'd be doing well to tell the difference at all.
This is off-topic, but....
Erm, why don't you use 7.d instead where the whole process is basically automated? Run genkernel --menuconfig all if you wish to tweak which modules are available.
If all else fails, use ndiswrapper. It works fine for me.
Whether the Mac Mini does better than the X-Box or not, the point is a GF3 (Tweaked, got some GF4 tech in it for the X-box) is better than a 9200. The difference in architecture/processors is one thing, but to say the Mac Mini's graphics card is purely better is wrong, shared bus or not.
As I much as I hate to link to Tom's Hardware, I feel I must. So go here and read. The GF3 in the XBox is basically a Ti500, and the 9000 is the 9200 (9200 is 9000 + 8xAGP, like that makes any difference).
Oh look, I don't see the ATi card beat the Ti500 anywhere.
I can see where this is going.....
:)
1. Charge $1 per hour of CPU time on your cluster.
2. Lower the speed of your processors.
3. Runtime of tasks increase. So your $1 does less.
4. PROFIT!!!!
Cool, no ???? step!
I agree with e40 - can you prove that it does?
/usr/lib & /usr/lib/64
I don't believe it you see. Whilst Sun can do that with SPARC, doing it was x86 is a totally different task. e.g. x86-64 has twice the number of registers as x86. Not exactly an ABI issue. This is something that is just going to require two copies of most things.
Solaris 8 does that with multiple copies of libraries, e.g.
I can remember reading something saying it wasn't out yet, but I can't find it now. Both you and an AC have told me otherwise, so maybe I'm wrong though?!?
The initial release of Solaris 10 is 32-bit only on x86, so you'll have to wait for release of a later version.
Likewise Project Janus also isn't included yet. ZFS wasn't included in Solaris Express, I'm not sure if it is now either. May expect a release in a few months to cover all these bases.