The Tom's review by itself invalidates this statement, with the 2.2GHz P4 outperforming both Atom offerings in a variety of benchmarks, and the Passmark scores support that assertion as well (and for what feels like the millionth time: you CAN NOT take the "overall" passmark score when comparing multi-core to single-core processors).

No it doesn't. The Atom they are comparing has a 600MHz clock deficit. That is huge.

Pentium 4 1.6GHz Passmark Single-core: 344 Pentium 4 2.4GHz Passmark Single-core: 564 Atom N270 Passmark Single-core: 240 Atom D […]
Show full quote

Pentium 4 1.6GHz Passmark Single-core: 344
Pentium 4 2.4GHz Passmark Single-core: 564
Atom N270 Passmark Single-core: 240
Atom D510 Passmark Single-core: 266

My original statement of 30% is based on the 1.6GHz P4 vs the Atom D510 - it's 29.3% faster, but it's easier to just round that off for the sake of conversation.

Where are you even getting these numbers?

http://www.cpubenchmark.net/low_end_cpus.html

Lets keep it simple by keeping dual core out of the equasion.
The Atom Z530 is single core, 1.60 GHz and has a TDP of 2W (!)
It scores 281 in passmark.

Above it a Pentium 4 2.60GHz scores 291. That is 10 more points for a whole gigahertz more of clock.

Despite the wall of text trying to defend it, again P4 really sucks it.

At 120x the power consumption. And if you are going to run DDR2 and PCI-E then why not just run Core2?

Atom D510 is 13W TDP. 120*13 = 1560W. 🤣 On it's absolute worst day even a Pentium 4 Extreme Edition will barely scratch at 1/10th of that. The 1.6GHz and 2.4GHz chips mentioned thus far are about half of that - 50-70W TDPs, versus the 100W+ of later Prescott models. As far as picking power efficiency goes, the Celeron on the linked board has a TDP of 19W, and that's not too far away from a P3. Just using my 1GHz CuMine as an example, it has a TDP of 29W. Performance wise I wouldn't expect them to be all that far apart for our purposes, and if you need the additional expansion (especially ISA) the P3 will make more sense. If we're only talking about games that need a Voodoo3 (or for which a Voodoo3 is overkill), the Celeron, P3, etc are much more efficient choices - their performance difference to a P4, Core 2, Core i7, etc whatever will be largely irrelevant.
[/quote]

The Atom Z530 (launched in 2008!) has a 2W TDP. I was wrong about 120x the power consumption but 30x is about right.

If you're wanting to step up to DX8/DX9 gaming, and are worried about power efficiency, Pentium M is a fine choice versus Athlon64 or Pentium 4. A Core 2 would also work fine, and may even become required depending on how far into the future you need to go.

+1
A Pentium M or Core 2 is both more efficient and still fast. If you need ISA, you would be better off with a P3 chip

My point in bringing up expansion options for the P4 wasn't really aiming at that though - it's more that Atom, and other integrated solutions (like the Celeron ITX linked, or the VIA ITX boards, etc) lack expansion options relative to more conventional desktop platforms, which can limit their suitability for gaming. For example in the Tom's comparison, it's no contest in 3DMark between the Pentium 4 and Atom systems, simply because the Pentium 4 systems can have a Radeon X800 installed in them, which is substantially faster than the Intel GMA. Core 2 is of course a better choice than Pentium 4, when available, and has all of the same basic advantages over a mostly integrated platform. Pentium M *can be* as good, as there are some Pentium 4 motherboards (like P4P800) that will support it, as well as motherboards designed exclusively for it (like the DFI in the TechReport article). For our purposes here, which are primarily gaming related, this is probably more pertinent than absolute power efficiency of a highly integrated ITX machine vs a conventional desktop. Of course if the ITX machine will meet all of your needs, go ahead and save the power and complexity. 😀

Which it in itself makes it a very unfair test. I don't know about what the Atom boards have available but I came across some mini-itx boards that had a PCI-E x16 slots
allowing just about any GPU to be used.

Again, you cannot take the "overall" Passmark score when comparing multi-core to single-core processors or single-threaded performance, because Passmark is unevenly weighted towards multi-thread execution. Nobody is arguing that in an ideal, modern, parallelized workload the dual-core Atom is at a disadvantage to a single-core Netburst (although the Tom's article does show some situations where the P4 outperforms the dual-core Atom). But that kind of performance is largely irrelevant here - none of these chips are going to be fast enough for games that require multi-core processors, and for older games (like would run on a Voodoo3, from the original post) single-threaded performance is all that you have to worry about.

I was looking at the single core Atoms so that is moot

Your own SINGLE CORE Atom N270 gets 282. The D510 has a SINGLE CORE rating of 266. Even taking the best P4 score and the worst atom one that is a clear win for the Atom.

The "best" P4 score, coming from the Pentium D 965 (it's the fastest/highest-spec NetBurst - if you have another meaning for "best" please clarify) is an overall rating of 1000, and a single-thread score of 845. The D510's single thread rating is 266, and "overall" is 667. In both cases the EE 965 is consistently higher. See above for the comparison numbers for the 1.6GHz, 2.4GHz, D510, and N270. The Mobile 1.6GHz achieves a single-core rating of 392, which is better than the desktop variant (344; to save scrolling). As a side note: mobile Pentium 4s are potentially dubious to use as benchmark sources, because many laptops with P4m will end up aggressively throttling due to heat (there's a thread in System Specs that explores this somewhat). This isn't to say P4m always performs badly, just that there seems to be more potential for variation between P4m-based machines due to throttling.
[/quote]

Its irrelevant. Of course a 3 GHz+ P4 will beat a 1.6GHz Atom. The question was per clock performance

I'm not seeing a "clear win" here - it instead looks like you're cherry-picking data to perpetuate an argument. And remember: none of this actually matters for this thread/discussion. The ITX board in question has a Core-based Celeron, the Celeron 220. That chip has a Passmark single-thread rating of 427, putting it well ahead of anything that's been mentioned here short of the EE 965 and 2.4GHz P4. And it's far more power efficient than either of those chips.

And you are not cherry picking?

I do agree on the Core2 celeron though. It will shred just about anything discussed here.

smeezekitty wrote:

And you are not cherry picking?

I do agree on the Core2 celeron though. It will shred just about anything discussed here.

According to obobskivich's crazy numbers the P4 1.6 will tie the Celeron 220 being discussed 😀

washu wrote:

Sorry but your P4 single core numbers are obviously ridiculous and are an error. If it was really 344 as you claim that puts it not only faster than a 1.6 GHz Athlon XP 1900, but also faster than the single core speed of a 1.6 Core 2. Are you seriously claiming a P4 is faster than a core 2?

They are not "my" numbers - they are the numbers published on the Passmark website (cpubenchmark.net). Feel free to look them up for yourself (and if you have issues with how Passmark does things, feel free to call them up and scream all you like). I came to all of them by simply searching "Passmark + [cpu name]" for example "Passmark + Pentium 4 1.6GHz" which goes here:
http://www.cpubenchmark.net/cpu.php?cpu=Intel … ntium+4+1.60GHz
You can find the Atom D510 and N270 as well:
http://www.cpubenchmark.net/cpu.php?cpu=Intel … +1.66GHz&id=610
http://www.cpubenchmark.net/cpu.php?cpu=Intel … 270+%40+1.60GHz

Only the embedded Celeron and mobile Pentium 4 had very low sample sizes (2 each); everything else had more than ten samples (some of the P4s have hundreds of samples).

The P4 1.6 has a score of 131, the 2.4 has a score of 224. The N270 has a score of 282. Are you seriously saying that just hyperthreading (the N270 is NOT dual-core) makes it get over double the score of the P4?

I'm taking the single-core numbers for all processors in question, because we're comparing single and multi-core processors. You cannot (I think this is the seventh time in this thread I've said this now) compare the "overall passmark score" (the "big number" in red) between single-core chips (like Pentium 4, Athlon XP, etc) and multi-core chips (like Atom D510, Core 2, etc) because the Passmark benchmark is weighted towards multi-threading (which is not "wrong" in and of itself - most modern benchmarks are, because modern processors and application loads are increasingly multi-threaded, but that isn't really applicable when we're talking about games that predate consumer SMP, and when comparing to CPUs from that era; single-threaded performance is all we really care about for those purposes). What Passmark is showing, and what Tom's review showed to a large extent, is what I've been saying all along: the Pentium 4 has higher single-threaded IPC than the Atom, and higher overall single-threaded performance due to generally being clocked higher to boot.

I have not, and am not, saying anything about HyperThreading. However since you mention it, yes it would appear that HyperThreading is contributing to the N270's higher score in the overall Passmark suite. SSE3 and other instruction set enhancements are also likely factors as well. Without more detail it's hard to say specifically if any single feature is mediating the higher score. For example, I know that in a lot of versions of PC Wizard, SSE3-equipped CPUs will score consistently higher than non-SSE3 chips, even if their x87 and ALU performance is lower than a non-SSE3 CPU (e.g. AthlonXP vs Prescott at similar clocks). Something similar may be happening in Passmark, although not explicitly with SSE3 (it may be HyperThreading, it may be some other instruction set enhancement, etc).

smeezekitty wrote:

No it doesn't. The Atom they are comparing has a 600MHz clock deficit. That is huge.

You claimed only a 2.4GHz P4 can begin to outperform the Atom, Tom's showed a 2.2GHz chip able to perform faster in a variety of tests. Where is the disconnect?

Where are you even getting these numbers?

For the eighth time, and please read: I am comparing the single-threaded Passmark scores, which is the correct way to evaluate these chips side by side, ESPECIALLY for our purposes at Vogons. Passmark themselves even acknowledge this, which is why both numbers are provided.

The Atom Z530 is single core, 1.60 GHz and has a TDP of 2W (!)
It scores 281 in passmark.

Above it a Pentium 4 2.60GHz scores 291. That is 10 more points for a whole gigahertz more of clock.

Atom Z530 has a single-thread score of 248. Pentium 4 2.6GHz has a single-thread score of 615. Per the above on HT/SSE3, it's likely that the Atom does so well in the overall (and the P4 so inefficiently) because of the presence/lack of SSE3 (since you've removed dual-core from the equation and the Atom still ends up "on par" in the overall, despite lower single-thread performance, which reminds me quite a bit of the AthlonXP<->Prescott comparison).

Despite the wall of text trying to defend it, again P4 really sucks it.

I'm not trying to "defend" anything - I'm not really "invested" in anything here, beyond being frustrated with having to repeat myself. In a lot of cases I would agree with you that P4 is a "bad choice" for builds (despite that almost all of my retro builds are NetBurst in one way or another 🤣 ) - they use comparatively a lot of power versus hardware on either "side" of them (P3 and Core 2), which makes them tough to recommend beyond their seemingly infinite availability.

The Atom Z530 (launched in 2008!) has a 2W TDP. I was wrong about 120x the power consumption but 30x is about right.

I'm not disputing that any of the embedded hardware mentioned here is much more power efficient, just that 120x seemed a little wild. I mean, P4 uses some serious power, but 1500W is a little bit of a stretch. 🤣 Tom's had power consumption #s (remember though, the P4 systems are "tainted" by that X800 card - why they couldn'tve just gone with a GMA-equipped board for the P4s...) that showed around 5-6x higher draw for the P4. The Atom itself is very power efficient, but it's usually the rest of the platform that brings the draw up (especially if 945 chipset). Pentium M can have the same "problem."

+1
A Pentium M or Core 2 is both more efficient and still fast. If you need ISA, you would be better off with a P3 chip

Aye. General curiosity: Are there even P-M or C2 boards with proper DMA ISA? I thought P4 was the end of that... 😊

Which it in itself makes it a very unfair test. I don't know about what the Atom boards have available but I came across some mini-itx boards that had a PCI-E x16 slots
allowing just about any GPU to be used.

Yeah. They should've used a 945-based board with GMA graphics for the P4s, instead of throwing that X800 in there. It would've made graphics/gaming comparisons possible, and made the power consumption numbers more "fair." (Atom would still ofc have much lower consumption, but it'd be more comparable numbers).

Alternately, going with a more expandable Atom platform would've solved that problem. That was part of my rationale behind linking the TR article with the Atom + GeForce laptop. Sure that IGP isn't great, but it beats the Intel GMA.

I was looking at the single core Atoms so that is moot

Not quite. If it were, the single-core scores would match the overall scores (for any chip), they do not. The "aggregate" or "overall" score is weighting towards both multi-threading (and a lot of the "single core" Atoms have HyperThreading), as well as enhanced instruction sets (like SSE3) that the P4s thus far mentioned do not have. If you look at the Pentium D (which has many of the same modernizations), it scores higher overall and in single-core, which is what would be expected based on that kind of weighting because it's a more "equal" comparison (in that it can complete more of the same tests).

Its irrelevant. Of course a 3 GHz+ P4 will beat a 1.6GHz Atom. The question was per clock performance

Indeed. I have no idea why "best P4" came up, but the data is available...

And you are not cherry picking?

I'm trying to be as objective as possible - I'm comparing the best (as in best quality, not "highest possible" or "best for my opinion") numbers, per Passmark's own description of their benchmark. This kind of back-and-forth over CPU performance has come up a few times recently when Passmark is mentioned, and I think most of the confusion is because it's easy to just pull up whatever Passmark aggregate and go "wowie, it's 10x faster!!! old stuff sux!!!" when in reality it's more complicated than that (which Passmark does outline if you spend some time on their website, but that isn't out-and-out on every page, and the tables are even worse for that). Changes in instruction set extensions and SMP are the biggest obfuscating factors here - all of that stuff absolutely matters if you're trying to run modern applications, but we aren't. Things like SSE3 and multi-core don't even come into the equation if you're talking about Doom 95 or Morrowind or Quake 3 under Windows 9x (which can't even run with SMP enabled).

I do agree on the Core2 celeron though. It will shred just about anything discussed here.

Aye. And the ones it can't, it's still much more power efficient than. 😀

smeezekitty wrote:

h time: you CAN NOT take the "overall" passmark score when comparing multi-core to single-core processors).
No it doesn't. The Atom they are comparing has a 600MHz clock deficit. That is huge.

That was the point I tried to make earlier. It's not a fair test because of the clock speed difference. If you scale the clocks to be the same, the Atom 230 is faster in over 80% of the benchmarks listed in the Tom's article.

obobskivich wrote:

I'm taking the single-core numbers for all processors in question, because we're comparing single and multi-core processors. You cannot (I think this is the seventh time in this thread I've said this now) compare the "overall passmark score" (the "big number" in red) between single-core chips (like Pentium 4, Athlon XP, etc) and multi-core chips (like Atom D510, Core 2, etc) because the Passmark benchmark is weighted towards multi-threading (which is not "wrong" in and of itself - most modern benchmarks are, because modern processors and application loads are increasingly multi-threaded, but that isn't really applicable when we're talking about games that predate consumer SMP, and when comparing to CPUs from that era; single-threaded performance is all we really care about for those purposes). What Passmark is showing, and what Tom's review showed to a large extent, is what I've been saying all along: the Pentium 4 has higher single-threaded IPC than the Atom, and higher overall single-threaded performance due to generally being clocked higher to boot.

And for the n-time, we are not comparing Dual core processors! We are comparing the SINGLE CORE N270 to a SINGLE CORE pentium 4. That means the overall passmark scores are comparable. I even listed a hyperthreaded P4 so the core arrangement is exactly the same. It takes an extra 1 GHz (2.6P4 HT vs 1.6 N270) to have similar scores. You can handwave the "single thread performance" all you want, it's so hilariously wrong for the P4s that it should not even be considered.

I have not, and am not, saying anything about HyperThreading. However since you mention it, yes it would appear that HyperThreading is contributing to the N270's higher score in the overall Passmark suite. SSE3 and other instruction set enhancements are also likely factors as well. Without more detail it's hard to say specifically if any single feature is mediating the higher score. For example, I know that in a lot of versions of PC Wizard, SSE3-equipped CPUs will score consistently higher than non-SSE3 chips, even if their x87 and ALU performance is lower than a non-SSE3 CPU (e.g. AthlonXP vs Prescott at similar clocks). Something similar may be happening in Passmark, although not explicitly with SSE3 (it may be HyperThreading, it may be some other instruction set enhancement, etc).

Again this is wrong. The proper overall passmark score shows that the Athlons that don't even have SSE2 let alone SSE3 handily beat the P4. And again, this is SINGLE CORE vs SINGLE CORE so it is a valid comparison.

Pentium 4 2.6GHz has a single-thread score of 615

You are saying a P4 2.6 is faster single threaded than a several core2s and Athlon 64 x2s which absolutly destroy it in the real overall benchmark. Yeah, that's believable.

You are cherry picking a obviously invalid number to support your point. How could a single core CPU get over double the score on the "per core" benchmark and cause it to score much better than other single core CPUs that are unquestionably better at the same clock like athons or athlon64s? It's obviously an error.

Again, if you have issues with how Passmark works, take it up with them. As it stands I'm using their data as they prescribe it; dismissing it because it disagrees with your sensibilities is illogical, but if this "crazy and obvious error" is really there, perhaps they'll listen to you.

.
Was this similar comparison already considered?:
Tom's Hardware: Intel Pentium 4 Vs. Atom: A Battle Of The Generations

Look at the Atom 230 vs both models of Pentium 4. I suppose the 230 is the small desktop variant of the N270, both from 2008.

Attached the benchmarks, minus the encryption specific ones. They speak for themselves.

In my experience I could run heavier software on a Pentium 4 Northwood 2,8Ghz compared to an Atom N280. The latter being comparable to a Fast pentium III. But the Pentium 4 is a waste of electricity, just like most of the Athlon XP systems. In 2008 the Atom was interesting, but for some reason intel refused to address its meager single core performance for many years after. They even stated "It is fast enough"!? I don't know the current status.
.

gerwin wrote:

. Was this similar comparison already considered?: Tom's Hardware: Intel Pentium 4 Vs. Atom: A Battle Of The Generations […]
Show full quote

.
Was this similar comparison already considered?:
Tom's Hardware: Intel Pentium 4 Vs. Atom: A Battle Of The Generations

Look at the Atom 230 vs both models of Pentium 4. I suppose the 230 is the small desktop variant of the N270, both from 2008.

Attached the benchmarks, minus the encryption specific ones. They speak for themselves.

In my experience I could run heavier software on a Pentium 4 Northwood 2,8Ghz compared to an Atom N280. The latter being comparable to a Fast pentium III. But the Pentium 4 is a waste of electricity, just like most of the Athlon XP systems. In 2008 the Atom was interesting, but for some reason intel refused to address its meager single core performance for many years after. They even stated "It is fast enough"!? I don't know the current status.
.

Did you notice the clocks? The P4 is clocked 37.5% faster than the Atom. If it wasn't for that, they would match or equal.

Oh...and in some tasks, a P3 beats a P4 so "comparable to a Fast pentium III" certainly doesn't except it from beating a Pee4 😉