Will the real Agena please stand up!

humm that does indeed look impressive.

Another site also have GP9600 @ 2.3Ghz, but with voltage around 1.2V. I suspect that 2.3Ghz will consume roughly around Q6600, with idle voltage all the way down to 0.9V. Seems like AMD's split plane voltage is really helping them.

I would say how can you doubt the creators of the first 1GHz X86 chip, first dual core X86 server CPU and the ONLY X86-64 implementation?

There should be some systems popping up soon. From that voltage 3GHz should be no problem.

yay, I want to see some benchmarks...

As you probably forgot Baron, Intel released a 1 Ghz and 1.13 Ghz chip before AMD's Athlon 1.1 Ghz.

I'll give you that AMD introduced the first X86 dual core Server chip (yay AMD!). IBM produced the first dual core server chip in 2000, long before the AMD milestone in 2005. Intel took longer to produce a dual core chip (released in 2006) and it turned out to obliterate AMD's dual core offering. Not sure what your point is here...

You also left out that Intel produced the first quad core x86 part. You also seem to be forgetting to mention that on current roadmaps, Intel will continue to outpace AMD in adding cores for the foreseeable future.

Um, EM64T (aka Intel 64) is for all relevant purposes identical to AMD64 (you know, AMD's implementation of X86-64) and has no significant disadvantages. Intel produced IA-64 before X86-64 was developed. Maybe it's time to admit that you don't know much about compilers and the evolution of x86... And so now even though most people have 0 use today for 64 bit computing, it's available from both vendors.

You come up with interesting "arguments".

Everyone knows that K7 hit 1GHz first - unless they were just fan boys spreading FUD. The point was that all of you act like AMD just started making CPUs after NetBurst. FX60 ruled the dual core world and Intel's 64bit implementation(IA64) was so useless that MS stopped making the Wksta OS - yes I actually tested Itaniums. I don't remember mentioning compilers so I'm not sure what that's about.

As far as 64bit, games are getting the support and the ONLY reason why they don't recompile is that there are TONS and TONS of HeatBurst chips out there.

LOL, so true, everyone knows the big Slot A K7 hit 1.0GHz first. Intel's Pentium III Coppermine only got it after a month or a few months after. Then they announced NetBurst and Pentium 4 was around the corner.

Thunderbird K7 @650Mhz - 1.4GHz stood ground to compete with P4 until the Palomino XP1500+ - X2600+ arrived and later Barton with 333MHz FSB from XP2500+ to XP3200+ with 400MHz FSB to compete with P4 at some what 3.6GHz

Then off come the A64 2800+ to A64 3800+/4000+ and FX-57 to leave the Pentium 4s behind.

Then came the Pentium D dual core and only to find the A64 X2 two weeks later to re-take the lead and until 2006 to be replaced by Core 2 Duo....

Now your up to date.

I thought the discussion was about x86 CPUs...
IA64 had nothing to do with x86 architecture, and IA64 is not the first non-x86 64-bit architecture : UltraSparc CPUs were already 64-bit (sparc v.9 architecture).... and it was in 1994...

anyway, does it really matter now to know who made the first 1GHz cpu, or the first x86-64 ? those chips are now obsolete, and what matters is the performance we can get NOW...

and we gotta hope AMD will be able to take the lead again, not because AMD is good and Intel is evil, but because concurrence is good for us (look at GPUs : the all new 8800gt is more powerful than a 8800gts and is priced like a 8600gts... do you think NV would have released it if the HD2900 wasn't here ?)

the last thing we need is one company taking the lead

BM is right about the 1 GHz race. So says CNET News.

http://www.news.com/2100-1040-237712.html

Not sure how that's relevant to anything in the present, but what they hell!

All you need to do is look at the HT Link speed, that tells you if it's an HT3 (AM2+) or HT1.1(AM2). HT3 has a much higher link speed. The Sapphire mobo that was used for the WGC in Seattle was HT3 with a 2.6GHz HT link. That board may have been HT3 since it was 1600MHz rather than 1000MHz. All of the retail 790FX boards should have HT3 enabled up to 3600MHz. The FX dual sockets will have the 3600 HT.

LOL Baron. You've got me kinda excited about it. I hope it lives up to the hype. I can't wait to see:

A) Third party benchmarks
B) Availability and pricing on our favorite Newegg



Keep us updated Mr. Baron.

Wow, it's nice that people use looking through old web articles as the way to "remember" what really happened. Here's the truth. AMD may have officially "launched" 1Ghz barely before Intel, but as with most AMD "launches" availability wasn't intact until some time later (you all should be familiar with AMD paper launches). Yes, Intel was rushed as well, but they still had the supply when needed. So the question becomes, which processor could you buy first? It was Intel.

It's also true that who got there first makes little difference in terms of today's terms. Ford created the first mass produced automobile, but currently, only a moron (or "car fanboy" ) would choose an American car over a Japanese counterpart. I merely am trying to counter Baron's twisting of historical facts. Give AMD credit where credit is due - but keep it there.

This looks like the whole pot and kettle situation. Chill out man. Buying an AMD based system isn't a bad thing, it is actually a very good system and makes a lot of sense for people buying on a budget.

quadhardcore,

What CPU were you running 2 years ago? Just curious...

Don't you love trolls who talk like they have actual facts to back up their statements? *begins rant* (1OOth time I'm saying this)UNTIL BENCHMARKS ARE DONE BY A REPUTABLE SITE SUCH AS THG, WE DON'T KNOW HOW WELL THE PHENOM DOES PERFORM. UNTIL THEN, WE ALL NEED TO STOP POSTING OPINIONS FOR THAT'S ALL THEY ARE. *ends rant*

Also, quadhardcore, without AMD, we all would be running 1.4Ghz Prescotts that costs $1,000 not $300 quads. Be glad AMD is around so you can buy Intel CPU's cheap.

I remember those days well. Back then, the thought was, "Who would ever need a CPU that is that fast?", and "Wow, a 30 gig hard drive, only a business could ever use that, cause I'd never fill it in my wildest dreams". And of course there was that $5,995 price. If we had to pay prices like that, how many of us could actually afford to buy the baseline computers,, much less high tech video cards? In many ways, I'm glad we're not living in the good old days.

The point of the post was to figure out where you are coming from. I wasn't implying one should buy and AMD based chip because they were the best a year or two ago. You post came across as fanboyish, so I thought I would inquire as to if you were buying the inferior chips back then or the fastest.

If you hate this topic, don't participate. It's OK. You don't have to agree. I can't say there won't be one since they have been demoing it for a few months now and yes it was hands-on. Showing it playing StrangleHold at the highest settings smoothly is all they really needed to show.

That implies that Phenom is a match for the SW. That's what it needs to be. Where were you when Intel pushed those TERRIBLY crappy NetBurst chips on everyone? I still have to use that crap at work. I work in a huge multinational corporation and I have yet to see a Merom OR Cointreau anywhere. If things were all even, there would be plenty of much better X2s around.

So basically, just go away.

Additional Phenom X4 information leaked again

http://myevilprocessor.blogspot.co [...] -6400.html

This time its GP-7000(correction, GP-9000), Phenom X4 running at 2.0Ghz. This time though, more well rounded testing suites are used.

This is the system specification

Looks like HT3 is used this time. Let's see if it will significantly help Phenom, like others claimed.

EDIT: The author of the article made a mistake. The 7000 series should be Tolliman, which is a tricore, not Phenom.
EDIT 2: The author later changed the model to 9000, so it was indeed a X4.

Seems like Phenom 2.0Ghz will be around E6550, @ 2.33Ghz (5988). Its a little disappointing, since PCMark 05 does take advantage of multicore. E6600 scored 6150, when Q6600 scored 7753. By using linear regression, a 2.4Ghz Phenom X4 should score 6996. Still falls short of Q6600 in this test though.

I'm not quite sure which score the author used to test for CPU, since there are two: ALU and MFlops. For comparison, Q6600 scored 43863 in ALU, and 29823 in MFlops. By comparing the 6400 score, the test should likely be ALU.

For multimedia test, it seems like the author used the Integer test, since 6000+ scored 62531 in Floating point. Q6600 scored 265322 in Integer, and 144367 in Floating Point.

The SisSandra also take advantage of multicore, with near perfect scaling (quad vs. dual = 2x performance).

Summary of site owner's remark: Phenom is like K8 when it first came out: high performance, but low clockspeed, low manufacturing capacity, low yield, and very expensive. It would take them until next year to become competitive against Core 2, but when that time comes, Intel's Nehalem will be launched. Therefore, I [the author] do not have high hopes for AMD in the next three years.

My own remark: The preliminary results look good, but unfortunately, like the author said, not good enough. Phenom will be plagued by low yield, low volume, and expensiveness (to manufacture). There's still no words on the actual price, but I wouldn't expect them to be cheaper than their Core 2 counterparts (unless Hector continues to stick to his "market share at all cost" ).

Again, as time goes on, things will definitely change. I wouldn't expect an improvement in IPC, but certainly in TDP and clockspeed. If AMD can push out 3.0Ghz by Q1 of next year, things can look a lot better for them.

Will it be a good drop in for current AM2 users? Sure. I would recommend Phenom to those who wants to upgrade from their X2s. But will it be a processor with enough punch to convince you to switch to AMD? Not likely. It is certainly a good processor to use in HTPC though, due to its reduced power consumption.

Intel - First to 2ghz? 3ghz? and the First to scrap a 4ghz? only to have a quad core overclock to 4ghz on air? True 64 bit itanium?

Is this supposed to be cool or something?

My own remark: The preliminary results look good, but unfortunately, like the author said, not good enough. Phenom will be plagued by low yield, low volume, and expensiveness (to manufacture). There's still no words on the actual price, but I wouldn't expect them to be cheaper than their Core 2 counterparts (unless Hector continues to stick to his "market share at all cost" ).


Again, as time goes on, things will definitely change. I wouldn't expect an improvement in IPC, but certainly in TDP and clockspeed. If AMD can push out 3.0Ghz by Q1 of next year, things can look a lot better for them.

And yet again you claim to know yields AND volume. First the designation X4 GP7000 is incorrect because the 7 series now TriCore not quad and I saw nothing that said what rev the engr sample was. Run CPU-Z and it will have a rev number below the stepping.

My X2 is JH-E6.

The fact that it is keeping up with a chip clocked 1.2GHz higher on tests shows that it will be competitive with C2Q and a great upgrade for AM2 owners.

I just hope they offer it in the Blackbird so I can get water-cooling when I upgrade.

Heh, I guess for outright quality and general driving experience, you're absolutely right. However, most of the US is not so fortunate to be so financially blessed to own an "ultimate driving machine". You have to admit though, for what you get in terms of use of car for cost of car + maint + parts, really Toyota is just about the best you can do and buying a Ford/Chev/Dodge is kind of foolish (remember - this is about cars - trucks are a whole different story).

Oh ya...computers and stuff!

Well I'm no expert but the low volts might actually be the reason that k10 is being released at lower speeds.

I agree. We'll see.