<< Imagine...AMD Clawhammer at 2666MHz while Intel is at 4.4GHz on the exact same .09 process...yet both offer identical performance.
>>AMD will be laughing their ars off all the way to the bank
Ausm
<< Imagine...AMD Clawhammer at 2666MHz while Intel is at 4.4GHz on the exact same .09 process...yet both offer identical performance.
>>AMD will be laughing their ars off all the way to the bank
Ausm
very insightful post, kenAF. some interesting info there. intel vs. amd war seems to be heating up a little bit. hopefully amd will come out on top...
t4t3r
t4t3r
This is great news for all! My next will probably go towards that 64bit platform since I can run Linux and XP 64-bit. To accompany that, I will want to get a 64-bit, 66mhz SCSI card. I want a REAL upgrade this time, not a minor speed bump -- the Hammer will be it!
vash
vash
vash, the 64-bit PCI cards have nothing to do with 64-bit OS code. That just translates into bandwidth of the PCI bus when said and done.
Imagine...AMD Clawhammer at 2666MHz while Intel is at 4.4GHz on the exact same .09 process...yet both offer identical performance.
Cool stuff. When's Intel's next revision of the Pentium 4 due? I heard mid 2003, but I'm not sure.
Cool stuff. When's Intel's next revision of the Pentium 4 due? I heard mid 2003, but I'm not sure.
<<
<<
<< Also AMD official confirmed that mass Hammer processors will be provided with an HIS metal lid improving the heat dissipation from the processor die and protecting it against mechanical damages. >>
finally. >>
Agreed. >>
I wonder is Sun is going to pick up the hammer line. If you haven't heard - Sun's getting friendly with AMD, See this thread.
i also have to say ken, well done..
cept i haven't seen a hammer mobo, don't even know what the socket will be called..... i think were more then 5 months from me be able to have a hammer. i think 2nd half 2002 more then early 2002..
<< i also have to say ken, well done..
cept i haven't seen a hammer mobo, don't even know what the socket will be called..... i think were more then 5 months from me be able to have a hammer. i think 2nd half 2002 more then early 2002.. >>
ClawHammer processors and motherboards won't be out any earlier than October most likely. December is more realistic....
Umm it is kinda of redundant to compare hammer to p4 since AthlonXP already owns it and hammer has a 30 percent perfortmance gap over AthlonXP. Also hammer has 86-64 and p4 is only 32bit. Hammer is just so much better than any pentium 4 could ever be without major changes to the p4 architecture. I heard hammer will use a 800mhz to 1200mhz hypertransport FSB! WHOA! 533mhz fsb doesnt seem so great anymore.
I still want to know what kinda of socketing the hammer will use? It clearly wont use socket a! I dont think AMD or Intel will ever go back to those dumb secc cartridges thats for sure! Come on AMD start sampling those things already!
I still want to know what kinda of socketing the hammer will use? It clearly wont use socket a! I dont think AMD or Intel will ever go back to those dumb secc cartridges thats for sure! Come on AMD start sampling those things already!
maj,
<< Hammer is just so much better than any pentium 4 could ever be without major changes to the p4 architecture. I heard hammer will use a 800mhz to 1200mhz hypertransport FSB! WHOA! 533mhz fsb doesnt seem so great anymore. >>
Well, remember that the P4 scales [in MHz] much better than the Athlon and Clawhammer, for the reasons I listed above. If you think the differences in clock speed between the Athlon and P4 are high now, you ain't seen nothing yet, because the MHz differences will only get larger as both companies move to newer and better processes.
Certainly, Clawhammer would lay utter waste to the P4 at comparable clock speeds...but they won't be running anywhere near the same clock. AMD will debut the Clawhammer at just over 2.0GHz (3400+), and it will still outperform the 3.0GHz P4 that Intel will release at about the same time. However, by the time Intel and AMD move on to .09, Intel will be able to match, or nearly match the Clawhammer's performance with massive clock advantages (i.e. 2.66GHz Clawhammer = 4400MHz P4; at same time on market). AMD should have a performance advantage over the P4 until mid 2003, when Intel makes the move to .09 (aaarg typos!))...Intel should then be able to clock the P4 high enough to compete with the Clawhammer or at least get close. Of course, the next P4 (mid 2003) will also make other performance improvements as well.
Ultimately, in two years, I expect we'll be about where we were several months ago....both companies offering comparable performance, AMD doing it with great IPC, and Intel doing it with high frequencies...with AMD delivering the best price/performance.
<< Hammer is just so much better than any pentium 4 could ever be without major changes to the p4 architecture. I heard hammer will use a 800mhz to 1200mhz hypertransport FSB! WHOA! 533mhz fsb doesnt seem so great anymore. >>
Well, remember that the P4 scales [in MHz] much better than the Athlon and Clawhammer, for the reasons I listed above. If you think the differences in clock speed between the Athlon and P4 are high now, you ain't seen nothing yet, because the MHz differences will only get larger as both companies move to newer and better processes.
Certainly, Clawhammer would lay utter waste to the P4 at comparable clock speeds...but they won't be running anywhere near the same clock. AMD will debut the Clawhammer at just over 2.0GHz (3400+), and it will still outperform the 3.0GHz P4 that Intel will release at about the same time. However, by the time Intel and AMD move on to .09, Intel will be able to match, or nearly match the Clawhammer's performance with massive clock advantages (i.e. 2.66GHz Clawhammer = 4400MHz P4; at same time on market). AMD should have a performance advantage over the P4 until mid 2003, when Intel makes the move to .09 (aaarg typos!))...Intel should then be able to clock the P4 high enough to compete with the Clawhammer or at least get close. Of course, the next P4 (mid 2003) will also make other performance improvements as well.
Ultimately, in two years, I expect we'll be about where we were several months ago....both companies offering comparable performance, AMD doing it with great IPC, and Intel doing it with high frequencies...with AMD delivering the best price/performance.
AMD should have a performance advantage over the P4 until mid 2003, when Intel makes the move to .11...
I think you mean .09u.
I think you mean .09u.
<< vash, the 64-bit PCI cards have nothing to do with 64-bit OS code. That just translates into bandwidth of the PCI bus when said and done. >>
Yeah, I know that, I just mentioned that I wanted a 64-bit card to accompany the 64-bit processor and 64-bit OS. I know 64-bit cards are out there today on systems, but I want a real upgrade next time.
vash
<< Is this going to be compatible with my Epox 8kha+ motherboard? >>
Nope, sorry.
ClawHammer will use a completely different socket design.One thing that has not been mentioned here is the inclusion of SOI. We don't yet know what kind of impact this will have on the required Vcore and ultimately the heat dissipation. If AMD has anywhere near the success IBM and Motorola have it should be considerable. Once expitaxial wafers are available Intel should join the SOI masses. AMD may even have the jump on Intel with SOI that they did with copper interconnects.
As far as multiple cores, I expect the Clawhammer to have a single core...at least initially. The Sledgehammer should be a different story at least according to this from "Chip Architect:"
A recently issued AMD Patent (6,275,905) on the name of Dirk Meyer and Jim Keller gives a possible system solution for an 8 way Sledge Hammer multiprocessing system. Such a system would contain four SledgeHammers each with two cores. Each SledgeHammer has its own memory controller on chip and three Hyper Transport busses. Two of these link to other SledgeHammers forming a rectangle with a SledgeHammer on each corner. The third bus interfaces each processor to its local I/O slots.
AMD won't be able to scale anywhere close to the speeds that the P4 is engineered for. For them to do so would not only require changes in the CPU, but also costly and R&D intensive changes to their packaging like Intel's BBUL. Increasing the IPC is the best choice for AMD. Including the memory controller is the single most important improvement in my eyes (along w/a heat spreader). AMD has always said they don't want to get into the chipset biz, and this will help to free them of the need. They can improve the "Northbridge" by updating the MCT and improving or changing the DCT. This will allow them go from DDR SDRAM to RDRAM as the need arises. It should also allow third party chipset makers to focus on improving AGP and bandwidth performance. With Hypertransport the FSB speeds should increase dramatically from the EV6's 266 FSB as well. The only part of the design I don't really like is the specifications for a maximum of two unbuffered DIMM slots. That seems a little short to me, but that still allows for 1GB of DDR (at least at the moment). The multi-processing Sledgehammer should have support for 4X that (or more).
As far as multiple cores, I expect the Clawhammer to have a single core...at least initially. The Sledgehammer should be a different story at least according to this from "Chip Architect:"
A recently issued AMD Patent (6,275,905) on the name of Dirk Meyer and Jim Keller gives a possible system solution for an 8 way Sledge Hammer multiprocessing system. Such a system would contain four SledgeHammers each with two cores. Each SledgeHammer has its own memory controller on chip and three Hyper Transport busses. Two of these link to other SledgeHammers forming a rectangle with a SledgeHammer on each corner. The third bus interfaces each processor to its local I/O slots.
AMD won't be able to scale anywhere close to the speeds that the P4 is engineered for. For them to do so would not only require changes in the CPU, but also costly and R&D intensive changes to their packaging like Intel's BBUL. Increasing the IPC is the best choice for AMD. Including the memory controller is the single most important improvement in my eyes (along w/a heat spreader). AMD has always said they don't want to get into the chipset biz, and this will help to free them of the need. They can improve the "Northbridge" by updating the MCT and improving or changing the DCT. This will allow them go from DDR SDRAM to RDRAM as the need arises. It should also allow third party chipset makers to focus on improving AGP and bandwidth performance. With Hypertransport the FSB speeds should increase dramatically from the EV6's 266 FSB as well. The only part of the design I don't really like is the specifications for a maximum of two unbuffered DIMM slots. That seems a little short to me, but that still allows for 1GB of DDR (at least at the moment). The multi-processing Sledgehammer should have support for 4X that (or more).
MadRat
Lifer
- Oct 14, 1999
- 11,967
- 280
- 126
I see AMD scaling better than they are letting on. Sure the raw MegaHertz will not be there, but the FPU and integer performance should be nothing short of amazing. Intel won't be including the "64 bit" selling point in the Pentium 4, which may hurt their marketing.
The only part of the design I don't really like is the specifications for a maximum of two unbuffered DIMM slots.
I was also wondering about that. Do you know if it?s true?
Intel won't be including the "64 bit" selling point in the Pentium 4, which may hurt their marketing.
I wonder how that'll be marketed....
I was also wondering about that. Do you know if it?s true?
Intel won't be including the "64 bit" selling point in the Pentium 4, which may hurt their marketing.
I wonder how that'll be marketed....
64-bits? Why would you want a bloated processor? Genuine Intel processors are a lean, fast 32-bits!
That's how it'd be marketed.
Of course, we can't forget Yamhill.
That's how it'd be marketed.
Of course, we can't forget Yamhill.
64-bits? Why would you want a bloated processor? Genuine Intel processors are a lean, fast 32-bits!
LOL, yeah right. Intel would get sued for false advertising....
Of course, we can't forget Yamhill.
I think I can safely say that ClawHammer will have a good 1.5-2 years head start on Yamhill if it ever is released.
LOL, yeah right. Intel would get sued for false advertising....
Of course, we can't forget Yamhill.
I think I can safely say that ClawHammer will have a good 1.5-2 years head start on Yamhill if it ever is released.
Head start doesn't necessarily mean it'll do anything like becoming a bigger name than Intel's stuff. Didn't 3DNow! get a head start on SSE? And if Intel is already working on Yamhill, then as soon as they see that Hammer is doing well, they'll just accelerate it, and make sure information leaks about it even more, and people will still, as always, think "Well I'll see what Intel does, they've got to be better than AMD", and even though Intel will be playing catch-up in this case, they'll still think Intel will be "more compatible" even though they're copying what AMD did.
I'd be surprised if it took Intel even a year to get Yamhill out, if Clawhammer really does make a splash. Intel won't just sit around and watch AMD rack up sales while saying "well, we're doing the best we can to catch up".
I'd be surprised if it took Intel even a year to get Yamhill out, if Clawhammer really does make a splash. Intel won't just sit around and watch AMD rack up sales while saying "well, we're doing the best we can to catch up".
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 Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter