flat assembler
Message board for the users of flat assembler.

Index > Heap > Intel Vs. AMD

Goto page 1, 2  Next
Author
Thread Post new topic Reply to topic
AlexP



Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.
AlexP
Okay, there's somebody here who thinks that AMD is better than Intel, what do u think???
Post 18 Mar 2008, 14:03
View user's profile Send private message Visit poster's website Reply with quote
edfed



Joined: 20 Feb 2006
Posts: 4237
Location: 2018
edfed
µP vs µP
I think that silicon is better than vaccum tube. Wink
Post 18 Mar 2008, 14:06
View user's profile Send private message Visit poster's website Reply with quote
AlexP



Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.
AlexP
lol I wish I had a vacuum tube processor...
Post 18 Mar 2008, 14:18
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
As the market is currently, Intel have the superior processors.

Maybe AMD will return again, they did have an advantage with AMD64 vs. the Pentium4... but imho things look grim for AMD right now, which is a shame.
Post 18 Mar 2008, 15:08
View user's profile Send private message Visit poster's website Reply with quote
asmfan



Joined: 11 Aug 2006
Posts: 392
Location: Russian
asmfan
This concurrent on the market makes fast progress more possible. Some time AMD is better other time Intel. This kind of situation makes opponents move and think faster because no one client wants to pay money for slow outdated things.
AMD is releasing new CPUs with new stepping with TLB problems being fixed. The picture is changing.
Post 18 Mar 2008, 16:31
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
asmfan: the CPUs AMD released (with the cache bug...) came quite a while after the core2, and were supposed to be a kind of response to those. But they're slower clock-for-clock, and come at lower frequencies as well. Fixing the cache isn't going to help (that was an additional penalty). And intel are getting ready to unleash 6- and 8-core CPUs...

I really do home that AMD will have a comeback, even if it wasn't on the performance side (ie., if they could deliver 75% the performance of an intel CPU with 50% power consumption and half the price - that would be interesting).
Post 18 Mar 2008, 23:47
View user's profile Send private message Visit poster's website Reply with quote
asmhack



Joined: 01 Feb 2008
Posts: 431
asmhack
btw can someone explain me plz why celeron processors are worse than pentium ? are they for cheap solutions ? Confused
Post 19 Mar 2008, 00:39
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Celeron has always been the "cheap brand", usually cutting on things like FSB speed and cache size. Can't remember when the celerons were first introduced, but I think it was p2 - and there were p3 and p4 celerons, and there's now also core2 celerons. They're dirt cheap and still fine for boxes that don't need to be super powerful, and they tend to consume less power than their faster brothers.
Post 19 Mar 2008, 00:43
View user's profile Send private message Visit poster's website Reply with quote
asmhack



Joined: 01 Feb 2008
Posts: 431
asmhack
ah ok.. thx
Post 19 Mar 2008, 00:49
View user's profile Send private message Reply with quote
edfed



Joined: 20 Feb 2006
Posts: 4237
Location: 2018
edfed
definitelly inel.
because they are the x86 mum.

they are abandonning the speed parameter, and try to increase the
multithreading.

then, it will be judicious to code in an object oriented way in asm x86.
each object will developp a thread in some cores.
and a virtual core (the kernel) will arrange the datas and the threads.

you'll see the power of asm soon.
Post 19 Mar 2008, 00:54
View user's profile Send private message Visit poster's website Reply with quote
daniel.lewis



Joined: 28 Jan 2008
Posts: 92
daniel.lewis
Rolling Eyes AMD has been offering 8-processor Opterons for ages. They don't outperform yet for most stuff because almost all programs are currently designed for single-processor CPUs, and most users only have 1-2 programs open at a time that actually deserve CPU time at all.

AMD Pacifica is better than Intel-VT, which means virtualization is faster. That will be a major issue in real life applications for companies, but won't show up in benchmarks.

AMD Hypertransport is better than Intel's FSB system, memory access is significantly faster. Intel's been working on an equivalent and it's due out in late 2009. It essentially makes all memory roughly equivalent to Intel's level 3 cache.

Intel has better clockspeeds, and tends to get the smaller nm processes faster. The motherboards for it are also better here in Japan.

I use x86-64 though, so I use AMD. Long mode FTW.

_________________
dd 0x90909090 ; problem solved.
Post 19 Mar 2008, 04:35
View user's profile Send private message Reply with quote
asmfan



Joined: 11 Aug 2006
Posts: 392
Location: Russian
asmfan
edfed but you forget those who proposed x86-64 (I mean AMD64 indeed) and if to remember that first 64bit Intel CPUs were suxx they weren't compatible with AMD64 and later Intel64 implementation (lahf/sahf Recent implementations) also who was the first implementer of SIMD with floating point?Smile Of cource AMD with 3dnow (1998) and only later Intel made its SSE (1999).
This derby moves progress.
Post 19 Mar 2008, 05:30
View user's profile Send private message Reply with quote
bitRAKE



Joined: 21 Jul 2003
Posts: 2906
Location: [RSP+8*5]
bitRAKE
AMD has had the better memory interface for SO long, but Intel will again catch up. I find it difficult to believe AMD doesn't have some surprises around the corner. Many of Intel's design improvements are having great impact on performance right now, but it is easy to make a big mistake riding out front.

It's a win-win for the consumer all the way around. I hear people upgrading to smaller power supplies and faster performance. Which is the way it should be - more costly to operate older hardware from many perspectives.

_________________
¯\(°_o)/¯ unlicense.org
Post 19 Mar 2008, 05:38
View user's profile Send private message Visit poster's website Reply with quote
sinsi



Joined: 10 Aug 2007
Posts: 692
Location: Adelaide
sinsi
I've had 9 computers since the early '90s and it's split intel/amd 5/4. I just got a q6600 because it seemed to be the best at the time.

It's a bit like ati/nvidia - sometimes one is ahead of the other.

Special instructions for certain cpu's (3dnow,sse3 etc) are nice, but a real PITA for an asm programmer.
Post 19 Mar 2008, 05:54
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
daniel.lewis wrote:
AMD has been offering 8-processor Opterons for ages.
I think you're confusing # of CPU sockets with number of cores on a CPU. Currently, I don't see any Opterons with more than 4 cores on a CPU mentioned at AMD's site, and here Denmark I can only find dualcore Opterons for sale... there's the quadcore Phenoms (with buggy initial revisions Smile), but those != Opteron.

daniel.lewis wrote:
They don't outperform yet for most stuff because almost all programs are currently designed for single-processor CPUs, and most users only have 1-2 programs open at a time that actually deserve CPU time at all.
Obviously you wouldn't stick an octa-core CPU in a workstation unless you have very specific needs - even quadcore processors aren't going to be utilized by normal people. Did you have a point?

daniel.lewis wrote:
AMD Pacifica is better than Intel-VT, which means virtualization is faster.
Please quantify this.

daniel.lewis wrote:
AMD Hypertransport is better than Intel's FSB system, memory access is significantly faster.
I haven't seen any realistic benchmark where HyperTransport turned out to give AMD that much of an advtange. I'd love to see one though, because in theory it is supposed to be better.

daniel.lewis wrote:
Intel's been working on an equivalent and it's due out in late 2009. It essentially makes all memory roughly equivalent to Intel's level 3 cache.
It's going to be interesting to see what happens when intel moves the memory controller on-cpu... They already eat AMD CPUs alive without Smile

daniel.lewis wrote:
Intel has better clockspeeds, and tends to get the smaller nm processes faster. The motherboards for it are also better here in Japan.
...and with Core2, it also performs better clock-for-clock than what AMD currently has to offer. Then add that the fastest stock Core2Quad are 3.2GHz (and that a standard 2.4GHz can be pushed beyond 3.0GHz easily), while the fastest Phenom Quad is 2.3GHz. (I'm talking CPUs I could walk down and purchase right now, not paper launches).

daniel.lewis wrote:
I use x86-64 though, so I use AMD. Long mode FTW.
And I use x64 as well, but on intel Smile

bitRAKE wrote:
AMD has had the better memory interface for SO long
But has it helped them outperform the Core2 architecture?

Btw., intel is going to re-introduce HyperThreading, and say that their implementation is going to be better this time. I know a lot of people bashed HT on the P4, but many of those were running on an OS where the scheduler treated the HT cores as any other core, which was stupid. It's going to be interesting to see what happens this time - an octacore CPU with HT would mean a total of 16 logical CPUs...

sinsi wrote:
I've had 9 computers since the early '90s and it's split intel/amd 5/4. I just got a q6600 because it seemed to be the best at the time.
And that's the way to go - be pragmatic, and get what has the best performance/price ratio at the time. I've been through a lot of AMD and Intel CPUs as well, and also settled for a Q6600 (Wanted a Q9450, but they weren't in stock and I needed a new box).

As for ATi/nVidia, I'm almost solely sticking to nv because of very bad driver experiences with ATI...
Post 19 Mar 2008, 09:51
View user's profile Send private message Visit poster's website Reply with quote
AlexP



Joined: 14 Nov 2007
Posts: 561
Location: Out the window. Yes, that one.
AlexP
I used to have an ATI, same with me I had bad experiences with it. Had to download extra software, get help from others, but finally it worked...
Post 19 Mar 2008, 12:52
View user's profile Send private message Visit poster's website Reply with quote
r22



Joined: 27 Dec 2004
Posts: 805
r22
Quote:
...and with Core2, it also performs better clock-for-clock than what AMD currently has

re: f0dder

For clarification by "roughly equivalent" I will mean within 10% of each other on most benchmarks with respective winners trading off.

An AMD x2 6400+ (3.2GHz) is roughly equivalent to a Core2 E660 (2.4GHz)
Old generation AMD vs next gen Intel. The GHz gap is pretty impressive.

But

An AMD Phenom 9700 (2.4GHz) is roughly equivalent to a Core2 Q6600 (2.4GHz) *** but this has an asterisk.

It seems when you compare Phenom to Core2 clock for clock, the AMD solution has better ALU crunching while the Intel solution has better SIMD.

K10 vs Core2
ALU performance: AMD > Intel
SIMD performance: AMD < Intel
Memory (it's a toss up but it looks like Intel thanks to tons more cache)

I used the Tom's Hardware Guide cpu charts to make this objective comparison.
http://www23.tomshardware.com/cpu_2007.html

Shouldn't this thread be on Heap?
Post 19 Mar 2008, 13:14
View user's profile Send private message AIM Address Yahoo Messenger Reply with quote
DustWolf



Joined: 26 Jan 2006
Posts: 373
Location: Ljubljana, Slovenia
DustWolf
f0dder wrote:
daniel.lewis wrote:
AMD has been offering 8-processor Opterons for ages.
I think you're confusing # of CPU sockets with number of cores on a CPU. Currently, I don't see any Opterons with more than 4 cores on a CPU mentioned at AMD's site, and here Denmark I can only find dualcore Opterons for sale... there's the quadcore Phenoms (with buggy initial revisions Smile), but those != Opteron.


This has to be one of those recursive jokes. He did say 8-processor. You didn't. Who is confused here?

Even a multicpu systems need special cpu design to work. That is why Intel opted for the cheapass multicore solution.

f0dder wrote:
daniel.lewis wrote:
They don't outperform yet for most stuff because almost all programs are currently designed for single-processor CPUs, and most users only have 1-2 programs open at a time that actually deserve CPU time at all.
Obviously you wouldn't stick an octa-core CPU in a workstation unless you have very specific needs - even quadcore processors aren't going to be utilized by normal people. Did you have a point?


How fast is your ethernet connection? I see gigabit on laptops these days. That needs a multicpu solution. And I'm sure there is more. Antivirus anyone?

f0dder wrote:
daniel.lewis wrote:
AMD Hypertransport is better than Intel's FSB system, memory access is significantly faster.
I haven't seen any realistic benchmark where HyperTransport turned out to give AMD that much of an advtange. I'd love to see one though, because in theory it is supposed to be better.


Then I suggest you stop looking up gamers for your benchmarks. Those people couldn't do a proper system setup if their life depended on it. Intel is more limited, thus to make it "fair", they set up benchmarks where both CPUs use the minimal bandwidth peripherials (RAM, motherboard, etc). And then you're suprized why AMD's advantage doesn't show?

f0dder wrote:
daniel.lewis wrote:
Intel's been working on an equivalent and it's due out in late 2009. It essentially makes all memory roughly equivalent to Intel's level 3 cache.
It's going to be interesting to see what happens when intel moves the memory controller on-cpu... They already eat AMD CPUs alive without Smile


Intel is big on few things and it's usually whatever sells. In P4 it was GHz, now it L2 and L3 cache. They won't do fast memory access the same reason they never made proper L1 cache.

f0dder wrote:
Btw., intel is going to re-introduce HyperThreading, and say that their implementation is going to be better this time. I know a lot of people bashed HT on the P4, but many of those were running on an OS where the scheduler treated the HT cores as any other core, which was stupid. It's going to be interesting to see what happens this time - an octacore CPU with HT would mean a total of 16 logical CPUs...


Don't you ever see when you're being fed bullshit by Intel propaganda? HT doesn't DO anything. It's a byproduct of Intel's innability to do a proper memory interface and yes I know how HT works. if HT show's you a fake CPU where there is none, this doesn't mean the new best thing since sliced bread just because you can SEE one in windows. If AMD ever went down this road, you'd see them say Athlon XP is a tripple core CPU and it'd still be more real than your hyperfake 16 core.
Post 19 Mar 2008, 23:24
View user's profile Send private message AIM Address Yahoo Messenger MSN Messenger Reply with quote
bitRAKE



Joined: 21 Jul 2003
Posts: 2906
Location: [RSP+8*5]
bitRAKE
f0dder wrote:
bitRAKE wrote:
AMD has had the better memory interface for SO long
But has it helped them outperform the Core2 architecture?
My point was that they have had positive innovation in the past which propelled them forward to a position on par with Intel, and that is amazing for both AMD and the general public. Their current position is not on par with Intel, but the cheap price we are paying for Intel chips is something AMD deserves thanks for! I find it hard to thank them with my dollars, though. Confused

Quite a powerful machine:
http://www.youtube.com/watch?v=7Lm8UGsS14A

_________________
¯\(°_o)/¯ unlicense.org


Last edited by bitRAKE on 20 Mar 2008, 01:02; edited 1 time in total
Post 20 Mar 2008, 00:19
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
DustWolf wrote:

This has to be one of those recursive jokes. He did say 8-processor. You didn't. Who is confused here?

I saw the "Rolling Eyes AMD has been offering 8-processor Opterons for ages." remark as a reply to my statement about intel getting ready for 6- and 8-core systems, and wanted to make clear the distinction between cores and sockets.


DustWolf wrote:

Even a multicpu systems need special cpu design to work. That is why Intel opted for the cheapass multicore solution.

Cheap-ass multicore solution? There's a limit to how many sockets you can fit onto a motherboard of given dimensions. More cores per socket = more total CPUs. And obviously with the added benefit of cores being able to share cache, less travelling through the memory bus, et cetera.

DustWolf wrote:

How fast is your ethernet connection? I see gigabit on laptops these days. That needs a multicpu solution. And I'm sure there is more. Antivirus anyone?

Gigabit, but that hardly requires a multicore system. With a decent I/O subsystem, you could keep a gbit connection saturated even on singlecore machine. But sure, multi-core helps, I'm just saying that as things are now, regular people aren't going to have much advantage of more than dual cores.

PS: anybody who needs to keep a gigabit connect saturated would use a NIC with jumbo frames and task offloading anyway, reducing CPU strain.

DustWolf wrote:

Then I suggest you stop looking up gamers for your benchmarks. Those people couldn't do a proper system setup if their life depended on it. Intel is more limited, thus to make it "fair", they set up benchmarks where both CPUs use the minimal bandwidth peripherials (RAM, motherboard, etc). And then you're suprized why AMD's advantage doesn't show?

Then please show me a benchmark where AMD has advantage. I haven't been looking at game benchmarks btw., since those tend to be GPU limited. The benches I've looked at have been the typical real-world usage scenarios like 3D rendering, video compression, general file compression, etc.

DustWolf wrote:

Intel is big on few things and it's usually whatever sells. In P4 it was GHz, now it L2 and L3 cache. They won't do fast memory access the same reason they never made proper L1 cache.

"almost all programming can be viewed as an exercise in caching" Smile. Isn't the important thing to be pragmatic and pick what actually works well in real-life?

DustWolf wrote:

Don't you ever see when you're being fed bullshit by Intel propaganda? HT doesn't DO anything. It's a byproduct of Intel's innability to do a proper memory interface and yes I know how HT works. if HT show's you a fake CPU where there is none, this doesn't mean the new best thing since sliced bread just because you can SEE one in windows. If AMD ever went down this road, you'd see them say Athlon XP is a tripple core CPU and it'd still be more real than your hyperfake 16 core.

I've used Hyper-Threading P4's, so I know what they're good at and what they're bad at. Imho the idea of utilizing unused execution units is good, but obviously you need better topology knowledge than just the count of CPUs, and you need to split your workload intelligently. While hyperthreading isn't a 100% increase in power (duh!), it did mean single-core P4-HT CPUs were more comfortable to work with than single-single core CPUs. So I still think it's going to be interesting to see how well the re-implementation of HT is going to work.

Please understand that I'm pragmatic and have been switching between AMD and Intel systems based on which one had the leading advantage. I'd love to see AMD come up with something good again, as I believe competition is healthy - for innovation as well as price. And currently I'm seeing an AMD that simply can't keep up, and defensive unrealistic AMD fanboiz jumping up and down and saying "not true! unfair! intel cheats!"
Post 20 Mar 2008, 00:52
View user's profile Send private message Visit poster's website Reply with quote
Display posts from previous:
Post new topic Reply to topic

Jump to:  
Goto page 1, 2  Next

< Last Thread | Next Thread >
Forum Rules:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum


Copyright © 1999-2020, Tomasz Grysztar.

Powered by rwasa.