Palomino, WHEN?!?

Goa^uld · 29th June 2001, 8:49 AM

hi people, great to see the AMD section of the forums is pumping

I have seen Athlon4 (Palomino) laptops in the news, and i know AthlonMP is on the palomino core, but when can we expect to see desktop palomino's?

these things look really cool, and im hanging out for one, i almost upgraded my 800mhz t-bird to a 1.33ghz version, but i decided to wait for palomino, and i just want an idea of how long i will be waiting!

so, any ideas?

MonteCristo · 29th June 2001, 10:10 AM

I heard August.

chainbolt · 29th June 2001, 11:41 AM

I also think August (maybe in AUS a bit later). It will be a 1.53 Giga. But don't expect miracles. The difference will show in some benchmarks, but not in rea life. At overclockers.com they are more or less disappointed. A bit cooler, a bit bandwidth more, but less friendly in overclocking. I do not think there is any benfit for somebody with a well running, highly overclockable 1.0/1.2/1.33 AXIA TBird. I'm running 1.33@1605 and I will not buy anything which is short of bringing me over 2 Giga aircooled. Even the
later following 1.7 Palmino will have problems doing this.

Goa^uld · 29th June 2001, 11:59 AM

ok thanks guys
maybe an AYHJA is still on the agenda

seaQuest_AMD · 29th June 2001, 12:33 PM

When its ready!!!

A 1.4GHz T-Bird is the fastest processor on the planet, geeze, how many more FPS do you want in Quake III????

Show me an application that requires more than 1GHz?!?!?!!

(Except Windows XP + Office XP)

Goa^uld · 29th June 2001, 1:13 PM

haha, the XP family needs 2ghz

i wont need them, but i dont want to risk buying technology that will be outdated. but, as it was so kindly pointed out, a palomino isnt much better than a t-bird.
decisions decisions

Manaz · 29th June 2001, 1:23 PM

I believe August or September.

Though I don't think Palomino will really hit it's straps till it's reduced to a 0.13micron process - then it will consume considerably less power, and will produce considerably less heat - and so should be able to go a considerable amount faster.

If I were in charge at AMD at the moment, 0.13micron production would be my MAIN goal at the moment. Because, as good as the Palomino may be on a 0.18micron process, the P4 will scale MUCH faster, and when Northwood comes out, I'm pretty sure that Intel are really going to be pushing it with all their might. If AMD don't have something that can compete in clock speed (as much as enthusiasts realise that clock speed on it's own isn't a good way to compare chips, Joe Public think's that faster is better, end of story), then they're going to start losing market share again...

Goa^uld · 29th June 2001, 1:42 PM

but if it is on a .13 micron process, wouldnt that mean a need for a new chipset? i think it probably would anyway, but we all thought that about tualatin, but then abit released new bios'es that claim tualatin support.

maybe thats why AMD is starting out with a .18 process, so people can upgrade to it. but then again, why bother with that, if they wont be able to with a .13, it could be just easier to goto .13 straight away. blah too many things to consider!!

seaQuest_AMD · 29th June 2001, 1:43 PM

Apparently www.auspcmarket.com.au will have some Athlon MP CPU's in next week.

Starting price of $550 for 1.2GHz & Fan

Manaz · 29th June 2001, 1:50 PM

Tualatin is *not* just a Coppermine processor produced using 0.13micron production methods, as opposed to 0.18micron. There are other changes to the P3 core that make Tualatin different to Coppermine.

If AMD change Palomino from the 0.18 micron manufacturing process to the 0.13micron process, but don't make any other changes to the core, then effectively it's the same processor, just the core will (or rather could) be smaller, and it should produce less heat.

The only reason Tualatin needs a new chipset is that they made changes other than just a drop to the 0.13micron process, and those changes aren't supported by earlier chipsets.

chainbolt · 29th June 2001, 2:41 PM

Quote:

Originally posted by Manaz
I believe August or September.

Though I don't think Palomino will really hit it's straps till it's reduced to a 0.13micron process - then it will consume considerably less power, and will produce considerably less heat - and so should be able to go a considerable amount faster.

If I were in charge at AMD at the moment, 0.13micron production would be my MAIN goal at the moment. .

That exactly is the point: the current "Palamino" 1.2 Giga is nothing else than a Tbird with some additional instructions for multiprocessing, it's the same architecture and the same .18 micron technology. Only a change from the current .18 to .13 micron technology would bring some real advantages.

xtreme2k · 29th June 2001, 6:31 PM

You guys seems to have missed the point.

A Tualatin is EXACTLY the same as a Coppermine instead, using a 0.13 micron as opposed to a 0.18 micron on the Cumine. The reason why it wouldnt work on the currect mobo is not an architectural decision but more of a Marketing decision. Intel wants the SELL the P4, thats why the make it HARDER for us to 'upgrade' to the tualatin. Since if it is too easy, no one will bother with the P4 due to totally screwed performance most of the time. It has been shown in benchmarks that performance wise at the same MHz, a Tualatin and Coppermine is VERY VERY close, ie, identical. No improvement other than a die shrink. One could argue the Tualatin-512 (Pentium 3-S) is an 'improvement' but all it is is that 512KB cache is used instead of 256K.

On the other hand, Athlon 4/MP (palomino) is way different compared to Athlon Thunderbird. Using the SAME process (0.18) it is already using lower power (about 20% less). This shows the die layout has been optimised. Architecturally also. Larger "Translation Look-aside Buffer" from 24 to 32 is an improvement. Hardware Prefetch is an improvement. SSE is an improvement. You will fine NONE of these architectural improvement moving from CUMINE to Tualatin.

Draffa · 30th June 2001, 2:47 AM

Any day now. Aust. Distributor have already got them in the pipeline.

Goa^uld · 30th June 2001, 10:15 AM

awesome
but i spose i will have to look at some benchmarks before i decide to get one

if they are alot more expensive, and dont perform too much better, or o/c as high, or i dont know there is nothing crash hot about them, an AYHJA is on the books then

thanks for your thoughts guys

Tarrasque · 30th June 2001, 2:46 PM

Quote:

Originally posted by seaQuest_AMD

Show me an application that requires more than 1GHz?!?!?!!

show me an application (not games / pro gfx) which needs more than a 200mhz

even games run fine on a 500ish (maybe low res) but pro gfx faster the better.. but yeah..

Tarrasque

29th June 2001, 8:49 AM	#1
Goa^uld Member Join Date: Jun 2001 Location: Melbourne Posts: 161	Palomino, WHEN?!? hi people, great to see the AMD section of the forums is pumping I have seen Athlon4 (Palomino) laptops in the news, and i know AthlonMP is on the palomino core, but when can we expect to see desktop palomino's? these things look really cool, and im hanging out for one, i almost upgraded my 800mhz t-bird to a 1.33ghz version, but i decided to wait for palomino, and i just want an idea of how long i will be waiting! so, any ideas?

29th June 2001, 12:33 PM	#5
seaQuest_AMD Member Join Date: Jun 2001 Location: Wollongong Posts: 1,233	When its ready!!! A 1.4GHz T-Bird is the fastest processor on the planet, geeze, how many more FPS do you want in Quake III???? Show me an application that requires more than 1GHz?!?!?!! (Except Windows XP + Office XP) __________________ Awesome...awesome to the max.

29th June 2001, 1:43 PM	#9
seaQuest_AMD Member Join Date: Jun 2001 Location: Wollongong Posts: 1,233	Apparently www.auspcmarket.com.au will have some Athlon MP CPU's in next week. Starting price of $550 for 1.2GHz & Fan __________________ Awesome...awesome to the max.

30th June 2001, 2:47 AM	#13
Draffa Member Join Date: Jun 2001 Location: Sluggy Prime Posts: 7,781	Any day now. Aust. Distributor have already got them in the pipeline. __________________ \| This post was brought to you by the Australian DSD, and Americas' NSA. If you think no one's paying any attention to you, we're listening. \| Worship the Comic. \| I aim to misbehave. \| Reduce, Reuse, Repair, Recycle. \| 2010 HoR Candidate for Longman.

29th June 2001, 10:10 AM	#2
MonteCristo Member Join Date: Jun 2001 Location: Brisbane, QLD, Australia. Posts: 437	I heard August.

29th June 2001, 11:41 AM	#3
chainbolt Little member Join Date: Jun 2001 Location: Tokyo Posts: 76,965	I also think August (maybe in AUS a bit later). It will be a 1.53 Giga. But don't expect miracles. The difference will show in some benchmarks, but not in rea life. At overclockers.com they are more or less disappointed. A bit cooler, a bit bandwidth more, but less friendly in overclocking. I do not think there is any benfit for somebody with a well running, highly overclockable 1.0/1.2/1.33 AXIA TBird. I'm running 1.33@1605 and I will not buy anything which is short of bringing me over 2 Giga aircooled. Even the later following 1.7 Palmino will have problems doing this.

29th June 2001, 11:59 AM	#4
Goa^uld Member Join Date: Jun 2001 Location: Melbourne Posts: 161	ok thanks guys maybe an AYHJA is still on the agenda

29th June 2001, 1:13 PM	#6
Goa^uld Member Join Date: Jun 2001 Location: Melbourne Posts: 161	haha, the XP family needs 2ghz i wont need them, but i dont want to risk buying technology that will be outdated. but, as it was so kindly pointed out, a palomino isnt much better than a t-bird. decisions decisions

29th June 2001, 1:23 PM	#7
Manaz (Damned or Belated) Join Date: Jun 2001 Location: Sydney, Australia Posts: 4,797	I believe August or September. Though I don't think Palomino will really hit it's straps till it's reduced to a 0.13micron process - then it will consume considerably less power, and will produce considerably less heat - and so should be able to go a considerable amount faster. If I were in charge at AMD at the moment, 0.13micron production would be my MAIN goal at the moment. Because, as good as the Palomino may be on a 0.18micron process, the P4 will scale MUCH faster, and when Northwood comes out, I'm pretty sure that Intel are really going to be pushing it with all their might. If AMD don't have something that can compete in clock speed (as much as enthusiasts realise that clock speed on it's own isn't a good way to compare chips, Joe Public think's that faster is better, end of story), then they're going to start losing market share again...

29th June 2001, 1:42 PM	#8
Goa^uld Member Join Date: Jun 2001 Location: Melbourne Posts: 161	but if it is on a .13 micron process, wouldnt that mean a need for a new chipset? i think it probably would anyway, but we all thought that about tualatin, but then abit released new bios'es that claim tualatin support. maybe thats why AMD is starting out with a .18 process, so people can upgrade to it. but then again, why bother with that, if they wont be able to with a .13, it could be just easier to goto .13 straight away. blah too many things to consider!!

29th June 2001, 1:50 PM	#10
Manaz (Damned or Belated) Join Date: Jun 2001 Location: Sydney, Australia Posts: 4,797	Tualatin is not just a Coppermine processor produced using 0.13micron production methods, as opposed to 0.18micron. There are other changes to the P3 core that make Tualatin different to Coppermine. If AMD change Palomino from the 0.18 micron manufacturing process to the 0.13micron process, but don't make any other changes to the core, then effectively it's the same processor, just the core will (or rather could) be smaller, and it should produce less heat. The only reason Tualatin needs a new chipset is that they made changes other than just a drop to the 0.13micron process, and those changes aren't supported by earlier chipsets.

29th June 2001, 6:31 PM	#12
xtreme2k Member Join Date: Jun 2001 Location: Sydney Posts: 3,225	You guys seems to have missed the point. A Tualatin is EXACTLY the same as a Coppermine instead, using a 0.13 micron as opposed to a 0.18 micron on the Cumine. The reason why it wouldnt work on the currect mobo is not an architectural decision but more of a Marketing decision. Intel wants the SELL the P4, thats why the make it HARDER for us to 'upgrade' to the tualatin. Since if it is too easy, no one will bother with the P4 due to totally screwed performance most of the time. It has been shown in benchmarks that performance wise at the same MHz, a Tualatin and Coppermine is VERY VERY close, ie, identical. No improvement other than a die shrink. One could argue the Tualatin-512 (Pentium 3-S) is an 'improvement' but all it is is that 512KB cache is used instead of 256K. On the other hand, Athlon 4/MP (palomino) is way different compared to Athlon Thunderbird. Using the SAME process (0.18) it is already using lower power (about 20% less). This shows the die layout has been optimised. Architecturally also. Larger "Translation Look-aside Buffer" from 24 to 32 is an improvement. Hardware Prefetch is an improvement. SSE is an improvement. You will fine NONE of these architectural improvement moving from CUMINE to Tualatin.

30th June 2001, 10:15 AM	#14
Goa^uld Member Join Date: Jun 2001 Location: Melbourne Posts: 161	awesome but i spose i will have to look at some benchmarks before i decide to get one if they are alot more expensive, and dont perform too much better, or o/c as high, or i dont know there is nothing crash hot about them, an AYHJA is on the books then thanks for your thoughts guys