Some Thoroughbred, Barton, ClawHammer, and Opteron (SledgeHammer) News.

[AGodspeed]

The Inq reports that:

Sun Microsystems will definitely use Opterons when the CPUs formerly known as Sledgehammer launch.

Better than that - Sun will sell boxes containing multiple Sledgehammers under its own brand name.

...Opteron seems to perform very well under tests carried out in the locale of Sun's HQ.

The Inq also reports (pointing to a roadmap from a Russian web site) that a ClawHammer 4000+ will arrive sometime in Q1 2003, with ClawHammer 3400+ processors arriving in October along with Barton 2800+ processors.

According to AMD, UMC will be taking off a lot of the load of the .13-micron K7 production (Tbred and/or Barton) from AMD's Fab 30. However AMD's official line is that UMC will contribute to .13-micron K7 revenue starting only in Q1 2003. We'll see how AMD spreads how their production capabilities soon enough...

So as usual, take The Inquirer with a very large grain of salt.

UPDATE: Planet3DNow claims that AMD will cut prices on May 27th and intro the first Tbreds on June 6th shipping up to 2200+ (1.8GHz).

UPDATE 2: In addition, it seems that ClawHammer is not delayed until Q1 (for availability that is) as was previously thought. According to AMD's Solomon Smith Barney conference today, the server CPU SledgeHammer (not ClawHammer) will be shipped in December of this year (which is Q4) and become available in Q1 of 2003. Knowing that AMD has said several times that ClawHammer would ship and be available before Opteron (SledgeHammer), this could give a little extra credibility to The Inq's claims that ClawHammer will be available in October of this year (unlikely IMHO).

 

[AGodspeed]

UPDATED

 

[Adul]

all i know is that i need to start saving now

 

[Rand]

If Sun was to adopt the Opteron in even limited numbers that would be a HUGE boost for AMD's presence in the entry level server market.
They're certainly not going to drop any UltraSPARC lines for their high end.

Naturally of course I'd want to see a hell of a lot more evidence then this before I put any faith in it.

I remain very doubtful that the ClawHammer will be available in October, and dubious that it will attain wide-spread availability prior to January.... nonetheless, even January for wide spread availability would be an incredibly fast pace from conception to mainstream availability let alone initial appearance in the market which should be late this year.

 

[Diable]

This is a smart move for Sun since Intel has announced they will continue production of Alpha CPU and Sun's current CPU's look like ass compared to Alpha's. This also means I'll get a few Opteron based Sun boxes to play with next year

 

[AGodspeed]

<< This is a smart move for Sun since Intel has announced they will continue production of Alpha CPU and Sun's current CPU's look like ass compared to Alpha's. This also means I'll get a few Opteron based Sun boxes to play with next year >>

Yeah, Sun's UltraSparc roadmaps seems to have slipped a bit in recent months. Maybe TI is having problems with the newer processes out there, I dunno.

I doubt Opteron will be sold (if at all to begin with) in any large quantities, and AFAIK Sun is still dedicated to UltraSparc for the vast majority of their boxes.

 

[Sid03]

very interesting stuff... taken with that huge grain of salt, of course.

but "4000+, 3400+, 2800+" really means nothing to me. i want to see some real numbers. i'm just fine with "2200mhz, but it will perform X% faster than a athlonxp".

 

[Athlon4all]

That is huge that Sun is going to be using Opteron. And that is great that CH avialability will be in Q4 2002. I cannot wait for CH!!!!!

<< all i know is that i need to start saving now >>

Good advice! I may not need CH, but I gotta try it out!!!!!

 

[AGodspeed]

but "4000+, 3400+, 2800+" really means nothing to me. i want to see some real numbers. i'm just fine with "2200mhz, but it will perform X% faster than a athlonxp".

Natually, AMD isn't going to reveal all the details about ClawHammer so many months away from launch.

 

[ST4RCUTTER]

I wouldn't be surprised if AMD launches the Barton with the 512k L2 and 166FSB. By the time Barton comes out DDR333 boards will the rule rather than the exception, and DDR333 modules will be prevalent. Stiff competition from the P4 will likely force AMD's hand this way as well, unless they decide to just wait until chips based on the Hammer core are available. SOI won't make it's AMD debut until then though...I'm really interested on seeing how much they can lower the Vcore for laptops using that technology.

 

[Sid03]

<< Natually, AMD isn't going to reveal all the details about ClawHammer so many months away from launch. >>

i don't know... intel says what mhz they are going to release a year from now. why doesn't amd? i mean, if they are ok with releasing this pr number why not mhz?

notice that with their recent press releases (like the t-bred), they don't mention mhz at all. i don't get why? it's as they are purposely avoiding it.

 

[AGodspeed]

<<

<< Natually, AMD isn't going to reveal all the details about ClawHammer so many months away from launch. >>

i don't know... intel says what mhz they are going to release a year from now. why doesn't amd? i mean, if they are ok with releasing this pr number why not mhz?

notice that with their recent press releases (like the t-bred), they don't mention mhz at all. i don't get why? it's as they are purposely avoiding it. >>

Because MHz doesn't matter all that much, it doesn't tell the whole performance picture. There's no reason to mention it.

Why doesn't Intel mention that the PIII is faster clock for clock than the Pentium 4? Because that doesn't tell you the whole story, the P4 can scale much much higher than the PIII can, enough to destroy it. There's no need to mention the PIII is faster clock for clock.

 

[Rand]

<< it's as they are purposely avoiding it. >>



They probably are, a lot of people still put tremendous faith in the clockspeed a processor is running at. If the ClawHammer is running at a lowly clockspeed they probably don't want to advertise that too much... irregardless of what's it's actual performace is.

 

[Sid03]

<< Because MHz doesn't matter all that much, it doesn't tell the whole performance picture. >>

i agree. which is why i said that they should say the mhz, and add how it's a better performer than their previous cpu.

(sometimes i wonder if my whole post gets read. )

mhz is significant. no, it's not the whole story. but it's certainly more of the story than "4000+". who knows what that is.



<< If the ClawHammer is running at a lowly clockspeed they probably don't want to advertise that too much... >>

now that makes sense. thanks, rand.

 

[AGodspeed]

mhz is significant. no, it's not the whole story. but it's certainly more of the story than "4000+". who knows what that is.

So you're saying that a 1733MHz Athlon XP would tell more of the performance story to the Average Joe than 2100+ would tell? How is Joe Average going to know that a 1733GHz AXP is faster than a 1800MHz Northwood?

i agree. which is why i said that they should say the mhz, and add how it's a better performer than their previous cpu.

(sometimes i wonder if my whole post gets read. )

Ok, so AMD should list the MHz of their processors in their press releases because it doesn't tell the whole performance story? How's that work?

I agree with Rand, AMD probably is trying to avoid using MHz as much as they can, since any Joe Average would probably think a 2GHz ClawHammer would be slower than a 2.01GHz Northwood. Certainly understandable, and a good reason for AMD to have the model number instead, don't you agree?

 

[Adul]

<< If Sun was to adopt the Opteron in even limited numbers that would be a HUGE boost for AMD's presence in the entry level server market.
They're certainly not going to drop any UltraSPARC lines for their high end.

Naturally of course I'd want to see a hell of a lot more evidence then this before I put any faith in it.

I remain very doubtful that the ClawHammer will be available in October, and dubious that it will attain wide-spread availability prior to January.... nonetheless, even January for wide spread availability would be an incredibly fast pace from conception to mainstream availability let alone initial appearance in the market which should be late this year. >>




which brings me to the point. Just how long have they beenworking on the clawhammer?

 

[AGodspeed]

<<

<< If Sun was to adopt the Opteron in even limited numbers that would be a HUGE boost for AMD's presence in the entry level server market.
They're certainly not going to drop any UltraSPARC lines for their high end.

Naturally of course I'd want to see a hell of a lot more evidence then this before I put any faith in it.

I remain very doubtful that the ClawHammer will be available in October, and dubious that it will attain wide-spread availability prior to January.... nonetheless, even January for wide spread availability would be an incredibly fast pace from conception to mainstream availability let alone initial appearance in the market which should be late this year. >>




which brings me to the point. Just how long have they beenworking on the clawhammer? >>

The very first taped-out ClawHammer (A0) silicon was produced sometime in January (SledgeHammer silicon was reached within a couple weeks of that time too). AMD has had at least 3 months and 1 week to work on ClawHammer and SledgeHammer silicon. However, since the Claw and Sledge A0 silicon were so very functional when they were conceived in January, AMD shipped off some Claw and Sledge systems to their "infrastructure partners" to test in early to mid February; this was first reported by XBitLabs and then confirmed a few weeks later by AMD.

 

[Sid03]

this thread isn't the place to debate pr ratings. so, i'll try my best to stay away from that.



<< Ok, so AMD should list the MHz of their processors in their press releases because it doesn't tell the whole performance story? How's that work?

>>

that makes absolutely no sense. mhz doesn't tell the whole story... that is true. but "3400+" tells even LESS OF THE STORY.

what's wrong with wanting to know the mhz? are you telling me that you aren't curious? wouldn't you rather know the pr AND the mhz? you don't agree that knowing both would give an even more complete picture, than just one or the other?

btw, i don't think joe average reads theinquirer.

 

[Bluga]

W00t!

 

[Goi]

This is a bad year for people who want to upgrade...too much waiting for the next big thing!
Sigh, I guess I'll just settle for the Thoroughbred when it comes out...I just can't wait any longer...

 

[BD231]

This is a bad year for people who want to upgrade...too much waiting for the next big thing!
Sigh, I guess I'll just settle for the Thoroughbred when it comes out...I just can't wait any longer...

Same, waiting SUCKS!

 

[AGodspeed]

Sine your pm's aren't on, I guess I'll just post here until you activate them...

To clarify one very important thing, we wouldn't be here if it weren't for the everyday average Joe Blow customer who goes to his nearest Best Buy to pick up "a really dudical computer!" Companies like Intel and AMD get a very large chunk of their revenue from Joe Blow, and therefore a lot of their marketing decisions (like the PR rating or like MHz) are geared towards Joe Blow. So...

that makes absolutely no sense. mhz doesn't tell the whole story... that is true. but "3400+" tells even LESS OF THE STORY.

Would you say that 3400+ tells even less about performance if a ClawHammer 3400+ were, on average, faster than a 3.4GHz Northwood (assuming AMD's PR rating stays conservative in any way like it is now)? Would you then say that 2.2GHz (the approx. clock speed of a 3400+ CH) would tell more about overall performance? Are you going to sit here and tell me Joe Blow is going to know that a 2.2GHz ClawHammer is equivalent to a 3.4GHz Northwood? No, and so AMD invented the PR rating (not perfect, but can you suggest something better or more realistic?).

what's wrong with wanting to know the mhz? are you telling me that you aren't curious?

It's listed on AMD's web site. All you have to do is search beyond the press releases, not all that hard. Or you can always go to the million or so hardware sites online or just CNET/ZDNET.

For the record, Intel can call their 2.53GHz processor the "Intel Pentium 4 7000" for all I care. It's up to Intel (and the industry in general) to come up with something better than MHz to help Joe Blow (99% of PC buyers) figure out how processors perform in relation to each other. Get the point? AMD could just call their Athlon XP processors the "AMD Athlon XP 5500", completely avoiding MHz altogether. It wouldn't change the performance of their processor, but it would confuse the hell out of Joe Blow (a very important person) and they would lose money. Therefore, AMD has to come up with something that Joe Blow might be more easily able to relate to. That's right, MHz!...but not really. It's the PR rating. Lets market "2000+" processors so that it'll look like "2000MHz", but it has to perform like one too (which has been verified too many times to count).

It's all about Joe Blow and marketing.

Once you activate your pm's I'll send this to you so that I don't clutter up this thread. It's off topic.

 

[HardWareXpert]

AGodspeed!

That nForce mobo you have, I was thinking of getting that same MSI one, any good?

 

[HardWareXpert]

<< mhz is significant. no, it's not the whole story. but it's certainly more of the story than "4000+". who knows what that is.

So you're saying that a 1733MHz Athlon XP would tell more of the performance story to the Average Joe than 2100+ would tell? How is Joe Average going to know that a 1733GHz AXP is faster than a 1800MHz Northwood?

i agree. which is why i said that they should say the mhz, and add how it's a better performer than their previous cpu.

(sometimes i wonder if my whole post gets read. )

Ok, so AMD should list the MHz of their processors in their press releases because it doesn't tell the whole performance story? How's that work?

I agree with Rand, AMD probably is trying to avoid using MHz as much as they can, since any Joe Average would probably think a 2GHz ClawHammer would be slower than a 2.01GHz Northwood. Certainly understandable, and a good reason for AMD to have the model number instead, don't you agree? >>



Agreed, thats the hole reason why AMD have the PR system on there CPU's, People after all do think a 2Ghz P4 is faster than a AthlonXP 2000+(1.73Ghz) and as for missleading us, no not at all, as it's performance than counts not Mhz.

 

[Degenerate]

I see sometimes dealer false advertise. they put xp 2000G as well as P4 2000G.... It is rather confusing...
But anyhow..... How does Pr rating incease by 600 over a change of the year? interesting.