Sunday, March 26, 2006

Rant: Why AMDroids suck!

If you don’t know what AMDroid is, well, they are my counterpart. I am called a FanBoy countless times, a Fanatic if you may. So while I may also suck just as bad, I know where I stand and have stood before, and will still be standing in the future.

The dawn of Intel®’s Core™ (a.k.a. Next Generation Micro Architecture a.k.a. NGMA)
This past couple of weeks, there has been quite a stir in both camps of the fence: Intel® and the other microprocessor manufacturer. Intel® showcased a working demo of a processor built on Core™ µA at the recently concluded IDF (Intel® Developer Forum) and even presented a side-by-side comparison of both its competitor’s top of the line processor. And mind you, it’s an overclocked µProcessor with the tightest timings on RAM possible, versus a supposedly entry-level µProcessor from Intel®. Now, why is it surprising?

AMD Top Dog Got the Beating = Angry Mob of AMDroids
Almost every AMDdroids post on the net scrutinized how Intel® may have configured the systems for Intel®’s favor. Wow, this is a number one sign of AMDroids not having enough brain cell power to even comprehend the breadth and width of IDF. Hey, Intel® invited in the conference some of the great minds in the industry, a mix of businessmen, journalists, enthusiasts and even tweakers and they will not treat any one of them as just dumb attendees who wouldn’t know how to look into the configuration of the demo system. Even the AMDtech top gun Anand played with the system and reported the great performance of the new Core™-based processor.
With this, it’s easy to deduce some facts: if you question the report coming from the very top gun, who have been blabbering a lot of negatives against Chipzilla, maybe AMDroids should be questioning the validity of all the reports and reviews posted on their site. Not just that particular report, but all reports, reviews, and articles. I mean hey, he’s the top gun, supposedly the best, and never even noticed that the demo units are configured in favor of Chipzilla? Hmmm, more reason to put all their reviews in the trash. A top gun who can’t even check whether a rig is configured with bias….speaks a lot about credibility.
Maybe, just maybe, AMDroids ought to look at this old news article posted sometime ago and let them remember how A* vs I* was set up before. Yes, it’s not Intel® who is duping the benchtest rig. AMDroids, you suck!

The deal about Old Technology vs New Technology
What in the crap are these AMDroids are talking about? Did anyone of the Chipzilla fanboys reacted the way they did when Netburst* is being compared with A64? They even said Netburst* is too old, not scalable, and all the negatives you can throw at it including the usage of the “antiquated” front side bus! Give it a break people, we still see 3DMark01SE on some of the benchmark tests posted on the web. It’s not the latest benchmark, but the results are still “scaled” and unbiased. Core™ still use that “antiquated” bus, and also utilized that “high-latency” DDR2 and yes, still use Netburst* technology. Come to think of it, all the three(3) items I mentioned were stomped, cursed, and looked down by the so called AMDroid experts. And they can’t really decide whether Intel® systems is antiquated or old: on one hand, AMDroids complained that Conroe is a new architecture being compared with AMD old architecture but on the other hand, they have been blabbering how the old FSB is still being used. What a bunch of nutcase, AMDroids you suck!

Clock Speed : AMDroids missed it
Yes, they have always been raving about clockspeed. Ever since that faithful days and early days of Intel® Pentium® 4, Chipzilla always put a GHz rating on their boxed processor. The other manufacturer tried to catch up, and use a so-called Performance Rating, wherein the so-called “estimated” performance is used instead of actual clock frequency since they can’t catch up with the GHz race. Anyone remembers when A* broke the 1GHz first? AMDroids were like reaching nirvana at the time because their “idol” got to it first. Who said A* wasn’t concerned about the GHz race, “AMD plans to lead in the gigahertz era” is a very clear message to me. But, as history have shown, A* got left behind with the GHz race, and switched to PR. Crap, just crap. And from then on for the longest time, it’s about PR vs GHz (yeah, I know A* PR mentioned it wasn’t supposed to be but it’s been an accepted standard, at least in the market and on the testbench).

But then A64 architecture debuted and matures, and suddenly, the comparison was shifted away from Clockspeed vs PR. Wow, there must be a massive brainwashing that must have swept the world when this became the “de facto standard” comparison from web sites. It all became A* 4xxx+ vs Intel® Pentium® 3.xGHz and AMDroids everywhere rejoiced once more for self-proclaiming A* as the performance crown holder. But if this ever happen on a Chipzilla camp, say, a Intel® Xeon® 3.6GHz versus A64 3500+ and you’ll be surprised if none of the AMDroids jumps up to point the inconsistency of the test bench even if the reviewer have clarified the reasoning for pitting the two processors together. Hmm, another reason to trash AMDtech articles.

Another notable items here is that, AMDroids just buys whatever A* spewed out of their factory because its got better PR label and so it must be better. But my oh my, even among their ranks, people are saying why a certain PR model is better/worst than a certain PR model with the same PR label (don’t knock yourself if you didn’t get it, it’s a socket confusion galore). Take a look for example, an AXP 2800+. If you are not well versed with A*, you will not know that there’s several flavor of the same thing, OTMH, there was a T-bred and Barton core (which both suck) btw, in comparison, Chipzilla use different naming convention (such as with a letter prefix i.e. Northwood A, B, C and moved to model numbering (such as 5xx, 6xx, 7xx, 8xxx series). Sure, there was an overlap for the Intel® Pentium 4 2.4A but at least the L2 cache is clearly stated on the label anyway not unlike some equivalent tray-type A* chips.
Now, the Core™-based processor on the tested over at IDF has much lesser clockspeed, and as usual, the Chipzilla FanBoys just smiled on the sidetrack while AMDroids whines and whines. Well, it’s been a long known fact that A* and its followers have been bashing the current line up of Intel® processors but if anyone even paid attention, it’s A* 4xxx vs Intel® Pentium® GHz (and A* has the most tightest timing on the RAM while I* has the loosest possible settings on the RAM). Notice the problem here? How can one say that their product is 4xxx+ rated when there’s no Intel® equivalent part number? Maybe they overclock the Intel® processor? But that’s a “maybe” and so AMDroids sucks once more.

Socket Soup Mix up and Memory Support: AMDroids got chained in the heel
When the newer Athlon64 debuted, it actually comes in many flavor. The first incarnations were so sucky, its got problems with integrated memory controller. And wow, they can’t really make up their minds back then. When AMD64 was launch sometime 2003, there were mix of sockets, and memory support. Such a very bad sign of integrating a memory controller on the CPU, as it tied up the processor to a certain generation of memory. AMDroids were bashing Intel® for moving to LGA775 with so many chipsets, hey, no one is forcing anyone to buy a newer motherboard with newer chipset everyday so what’s the big deal? Besides, there was only Socket423 (very early Intel® Pentium® 4), Socket 478, and now LGA775 and memory support was never a problem.

If you have an old Intel® Pentium® 4 Socket478, you can still use it with DDR400 and DDR500 provided the motherboard supports the RAM. You can even using this very same processor on motherboard equipped with PCIe graphics card. The newer LGA775-based processor can still run on both DDR1 and DDR2, with both AGP and PCIe options. Wow, antiquated design eh? With early Sempron based on S754, well boohoo, you can’t put it in the S939/S940 motherboard and tied with single channel memory technology. In short, you suck! For those who bought early S939, and one of those who are touting “64bit” you suck even more since most of you can’t even make use of RAM beyond 4GB because of cost and applications. For those who got FX series based on S940 that needs ECC Memory, you suck the most, not only those memory expensive, the CPU is not compatible with S939 and A* left you to rot in hell when they decided to go S939 for their mainsteam A64 processors, forcing you to move to S939. Yes, A* is now touting “drop in upgrade, just BIOS upgrade” after they kick the nuts out of the early adaptors.

On the other hand, FanBoys who migrated to LGA775 early can still use their processor, memory and graphics card to even greater performance by moving into newer motherboard with better features. And hey, isn’t DDR2 much cheaper compared to DDR1? Isn’t PCIe graphics card much cheaper than AGP? AMDroids, you suck!

Thermal Envelop: A heater no more
The new Core™ architecture touts performance/watt and along with it, the lessening of the heat being produced by the processor. True, the Prescott generated some additional heat load relative to the older brother Northwood. But umm, as far as the number of burned processors I’ve seen on the net, Intel® Pentium® 4 near Zero while I’ve seen more than a couple of burned and dead A64. If the processor is “hot” then why is it none ever burned to death? Well AMDroid kids, heat is relative, our Sun is so hot, very very hot but it will appear cold when it is compared to NGC 2440 Nucleus. Antarctic weather is cold, but it will appear summer compared to Pluto. Now, what’s the point? Well, if your CPU is running at 30C at full load but still burns, and my CPU is running at 70C and still having a merry way, then which is hot or hotter? AMDroids, you suck so bad!

Even back in the early days of AXP vs Intel® Pentium® 4, AXP suffered massive death galore from overheating, while Intel® Pentium® processors survive even the most rigorous test of overclocking and overvolting. AMDroids sucks!

Price/Performance, Same Old Same Old Same Old
AMDroids used to depend on how A* back in the AXP days where bang-for-the-back and should be used if PC are to be used for games since “it’s just games” even world wide benches shows Northwood C were kicking the said processor dead on every corners. They justified to still purchase AXP because performance versus comparable Intel® Pentium® 4 C is negligible, and about a good 10% to as high as 30% cheaper. But heck, look right now, the very same AMDroids will not purchase the much more cheaper say Intel® Pentium® D 805 because the A64 X2 3800+ can play games faster and that the former has good applications only on desktop, multimedia, and multi-tasking: in short, the A64 X2 3800+ has become a very expensive gaming console, which before, as per the AMDroids, is a cost-effective gaming PC. Wow, know where you’re standing so you won’t lose your path. AMDroids never really cared for the price, they just want to suck even more and pimp A*.

Some AMDroids even claim then can discern the difference in actual gameplay between 100FPS and 105FPS and 95FPS. You clearly suck AMDroids. Even at 100FPS vs 150FPS, feeling the difference is near impossible, if it’s between 25FPS and 50FPS, then yes it’s noticeable. So, AMDroids justify that you need to have a good video card and that if it’s just a mid-end card then better stick with A64. Oh man, high-end gaming is most of the time GPU bound. I doubt there will be a jump from 25FPS to even 70FPS by just changing CPU speed. You suck!

Innovation is the name of the game
Funny how many AMDroids will say Intel® has been playing catch up. What is it that is going to be caught up? The only thing Intel® should be catching up with is regaining the lost 2 or 3 points of marketshare and not about innovation. Why not innovation? Because AMDroids ought to know that the PC they are using based on A64 is very much based in Intel® technology. If you are enjoying the PCI and PCI devices integrated on your board, guess who pushed for this technology? If you are still using AGP, guess who invented it? If you like your USB MP3 player so much, guess who pushed it so hard in the market? WiFi, guess who makes this technology so common the world over? Stop mentioning about Intel® not innovating anything, you’re just showing how dumb you are.

The case of FX vs Extreme Edition
This is a very sucky subject indeed. AMDroids ought to know where their FX series is standing on the equivalent A64 4xxx+ line up before they even try to say they own the crown. Sure, price is being compared, price/performance hullabaloo but hey, read the top part regarding same old comparison, which, unfortunately, applies very much here as well. In any case, what many AMDroids failed to see is that these benchmarks, even if its an unfair comparison (remember I*’s EE never hits past 4GHz while A*’s FX lines up pretty much at their A64 4xxx+ processors) never really reflects user performance. Heck, my Dothan can score a lot higher than my Prescott and yet the smoothness with gameplay and multi-tasking can be felt: Dothan rules at pure benchmark numbers but a bit slow on multi-tasking which is where my Prescott with HT comes in. The EE, specially the newer ones, can make multi-tasking a lot lot easier. It helps process physics calculation and off-loads tasks from the background applications, freeing a core (logical or physical) and dedicating it in game processing. The result, a lot smoother multi-tasking and smoother, less sudden dropped frames.

On another thought, why is it that no one is complaining why BMW* and Kia* that can both reach a destination equally well, that they are priced so far apart? Why Levi’s* has different pricing than Jag*? If you are playing smarty, no, it’s not just the name recognition, it’s the reputation that they have built through time. The brand name shouldn’t be discounted, brainless AMDroids spews crap like “you only buy the brand if you buy Intel®” and they fail to see that that consumers are “buying THE brand” if they buy Intel® products”. If you don’t care about your name, well, don’t think Intel® doesn’t care for theirs as well, they have built it and developed it thru service, commitment, and quality products. It’s not “just” a name, IT IS a name.

Warranty and Support, a not so well known fact!
Yep, AMDroids always say A* is so much cheaper (back then) and so it must be better. They’d buy a cheap motherboard, cheap processor, and then go whine in the forum if something goes wrong. Oh freaking crap! And they just won’t stop, they’ll post and post crappy help me with this and that. If you buy Intel®, rest assure that you will have a number you can call on. You will not have to scour the forum and spam each and every message board you found asking for help and enlightenment. Sure, answers can be found that way, but isn’t it more elegant and convenient if there’s a number you can call and vent your frustration?

And if things don’t go well and you got a defective part, isn’t it nice that if you buy your motherboard in Hong Kong that you can still have it RMAed in the Philippines? Yep, Intel® will pick up the item from your location and have it delivered to you free of charge if it is found defective and within warranty period. Now now now, isn’t that great? Beat that with a competing product. You can only get this kind of service if you buy a pricey OEM-build PC, but not with Intel®, you will be treated equally.

In Closing
I guess in the end, I am the suckiest being too much of a fanboy and have been blinded as well. Well, I guess that's a day in a life of a fanboy. Till next rant.

No comments: