How Did Anand Let This Get Published?

[Arachnotronic]

From the recent AMD A10 mobile APU review,

I am, however, forced to address a few elephants in the room that are getting glossed over. As consumers, we need AMD to succeed. Lack of competition is showing in a major way: desktop Haswell is a joke, Haswell's GT2 IGP is a minor improvement yet promises to be the most common one in Intel's lineup, and Intel seems to be planning to mostly coast on Haswell for two years while focusing on Atom's successor.

I'm sorry, but while I do agree that desktop Haswell is underwhelming for those looking to upgrade from Ivy Bridge, it seems like complete [travesty] that this reviewer would actually say that Intel is "coasting" on Haswell when it was very clearly a major step in the right direction...mobile.

Haswell ULT brought some serious battery life gains, and Haswell in general brought us ~10% IPC increase in legacy code + some seriously cool new instruction set extensions. Broadwell should be good for mobile/AIO devices where it will be in BGA form and probably will focus much more on IGP improvements than CPU improvements anyway, so it'd be pointless to bring it to LGA.

Thoughts?


No profanity in the tech forums, please.

Moderator jvroig

 

[tynopik]

Haswell ULT brought some serious battery life gains

desktops don't run off battery

Broadwell should be good for mobile/AIO devices where it will be in BGA form and probably will focus much more on IGP improvements than CPU improvements anyway, so it'd be pointless to bring it to LGA.

so you're agreeing that Intel is coasting on the desktop side?

 

[blackened23]

Haswell reviews across many websites made similar statements, actually. I don't agree with the assertion of Haswell being a "joke" at all - I think Haswell is a fine chip, but what's the extent of subjectivity within a review? I think all reviewers show some degree of subjectivity, even if I completely disagree with them.

It's a tough thing to discuss. On one hand, reviewers will always be somewhat subjective. You won't always agree. In this case I don't agree. I'm also interested in other opinions - I certainly don't comdemn the reviewer for stating such things because there are actually many Haswell reviewers saying similar things.

 

[Arachnotronic]

desktops don't run off battery

All-in-ones are still very sensitive to power consumption. Also, is electricity free where you live?

so you're agreeing that Intel is coasting on the desktop side?

In the all-in-one space? No. In the LGA mainstream space that the company is quietly trying to shift towards the higher end LGA platform? Sure, you could see it that way.

Again, with Haswell Intel extended its lead over AMD even further at half the power consumption. What more do people want?

 

[BallaTheFeared]

Were people, even reviewers it seems, expecting a generational leap between a tick and a tock when the tick had half the advantages of the upcoming tock?

Haswell is exactly what we expect for a Tock over the previous tock, the difference here seems to be ignorant people thinking Ivy was your typical Tick and therefore Intel is no somehow slowing down, which is all just a product of AMD being nothing short of terrible.

 

[sushiwarrior]

All-in-ones are still very sensitive to power consumption. Also, is electricity free where you live?

awwww yiss dat power savings

but oh wait it saves 6w!...

awww dammit then it uses 31 more

Were people, even reviewers it seems, expecting a generational leap between a tick and a tock when the tick had half the advantages of the upcoming tock?

Haswell is exactly what we expect for a Tock over the previous tock, the difference here seems to be ignorant people thinking Ivy was your typical Tick and therefore Intel is no somehow slowing down, which is all just a product of AMD being nothing short of terrible.

but if you tick the tock of the ticking tick on the previously tocker ticking, then the tick tocking ticktock becomes more like a tocking tickerticktack.

 

[richierich1212]

What would you like to be published instead? "Intel is god"?

 

[BallaTheFeared]

So you're comparing it to Ivy, the 22nm product, with most the power saving features on a desktop SKU?

Why not at least compare them overclocked, since you felt the need to compare the k sku's in the first place.

Look at the jump over the previous Tock, Sandy Bridge.

The only thing Intel screwed up was releasing a full lineup of Ivy Bridge sku's, certainly didn't gain them any praise even though it was one of their best, if not the best Ticks they've put out.

 

[mrmt]

Rather ironic to give that kind of criticism when writting an AMD review.

 

[Mushkins]

awwww yiss dat power savings



but if you tick the tock of the ticking tick on the previously tocker ticking, then the tick tocking ticktock becomes more like a tocking tickerticktack.

Gotta agree here. I leave my Ivy based gaming desktop on 24/7 outside of restarting for updates. My overall household power bill is like $70 a month. Upgrading to haswell would cost me about $300 for chip and motherboard. The power savings might make up for that $300 purchase... by the time I retire. Not unplugging the microwave when its not in use is costing me more money than that power draw difference. Unless its a major change or you're talking overall power consumption for businesses with 1000+ 24/7 workstations, power draw in the desktop space is kind of a moot point when it comes to processors. Hell, i'd save more power than the upgrade just by turning the damn thing off once in a while, and even then the difference in my monthly bill would be maybe $10 a month tops, who cares?

Now on the other hand, if a similar specced Haswell gave my gaming laptop an extra 20 minutes of battery life over Ivy? That's a considerable improvement, and reaches even further for lower spec mobile devices with less power hungry hardware in them. Haswell power savings are *all* about mobile, and they did a good job.

 

[sushiwarrior]

So you're comparing it to Ivy, the 22nm product, with most the power saving features on a desktop SKU?

Why not at least compare them overclocked, since you felt the need to compare the k sku's in the first place.

Look at the jump over the previous Tock, Sandy Bridge.

The only thing Intel screwed up was releasing a full lineup of Ivy Bridge sku's, certainly didn't gain them any praise even though it was one of their best, if not the best Ticks they've put out.

Ivy Bridge = good (that's why I bought Ivy Bridge)

haswell = intel pls

Nobody is calling IB bad, they're calling haswell bad for a new product. because, well, it runs the same as the old product...

 

[BallaTheFeared]

Ivy has 22nm and Tri Gate, the rest is pure uarch.

Haswell scales down better than Ivy, and it scales up better as well while having an improved instruction set.

Ivy wasn't good if Haswell isn't, sorry. Haswell is a better chip than Ivy on the same node.

 

[ShintaiDK]

Wow, and in a review of an AMD mobile chip? And no battery tests. Matching a 15W Intel CPU vs a 35W chip.

And why are so many benchmarks left out with only AMD?

Oh the irony is complete!

Seems the reviewer blames Intel for what AMD is lacking. The review is simply full of bad excuses and what is essentially slander.

 

[mrmt]

It's better get used to it:

http://semimd.com/blog/2012/10/18/deep-inside-intel/

SMD= Along those lines, does Intel see the smart phone and small mobile device market as a key direction?
Bohr= Intel is very serious about getting into the smart phone and tablet markets. We are a very different company from what we were five or six years ago. We are developing process technologies, but also products, that span a much wider range of performance and power than anywhere in our history. We’re not just after the high-performance desktop. We’re developing products that support 100-watt server chips down to sub-1 watt smart phone chip

(...)

SMD= Has the priority for what you’re designing into a chip changed? Is it still all about performance, or has power overtaken that?
Bohr= Ten or 15 years ago, performance was the main goal in developing a new process technology. That really has gone away as the No. 1 priority. We still strive to provide a performance boost with each new technology, but there’s much more emphasis on improving power or efficiency on each new generation. We do that by reducing active power for the work a chip does. That’s a much more important goal for us today. Part of the reason is that the market has shifted from desktop applications to more mobile products. The first transition was from desktops to laptops. Now the move is to put things into smart phones. Today’s consumer wants computing power he can hold in his hand in the form factor of a smart phone and a tiny battery. He wants the performance he had on his laptop only three or four years ago. That’s what we shoot for.


(...)

SMD= But wires don’t scale well. How do you deal with that?
Bohr= RC delay gets worse as you scale, compared with transistors, which tend to get faster as you scale down. The industry has had 20-plus years of struggling with that problem. One way we’ve addressed that is that we’re no longer striving for very high operating frequency, especially in the phone market where 2 or 2.5GHz would probably be sufficient. That’s one advantage. The other advantage is that the average size of the chips is smaller in these laptop and cell phone applications so you don’t have interconnects traveling a long distance across a large chip. Instead, it’s a more compact chip so the signals don’t have to go as far. But even with those chips, we still have a challenge of performance from the interconnect. We have to be clever about what pitches we choose. Some of the lower layers are dense pitch, where density is important. Some of the upper layers are coarser pitch, where performance is important. We’re also continuing to drive down interconnect capacitance by employing lower-k dielectrics.

 

[Magic Carpet]

So you're comparing it to Ivy, the 22nm product, with most the power saving features on a desktop SKU?

Why not at least compare them overclocked, since you felt the need to compare the k sku's in the first place.

Look at the jump over the previous Tock, Sandy Bridge.

The only thing Intel screwed up was releasing a full lineup of Ivy Bridge sku's, certainly didn't gain them any praise even though it was one of their best, if not the best Ticks they've put out.

I'd like to see haswell's power consumtion at 5 Ghz, can anyone measure it?

 

[njdevilsfan87]

I like it when they did both... ie Q6600 to Q9550. That was a tick that by today's standards would have been a good tock. Sandy Bridge was such a tock. IPC improvement wasn't much, but reduced power and amazing overclocks were. I'm alright with small IPC improvements, as long as they bring reduced power consumption, and higher overclocks to widen the performance gap.

 

[moonbogg]

Get over it OP. Haswell sucks. So does Ivy. But SANDY is still sexy sexy SANDY!! (I just bought a 4770k for work but SHHHHHH!!!!)

 

[BallaTheFeared]

I'd like to see haswell's power consumtion at 5 Ghz, can anyone measure it?

I can, but how are we going to create a baseline?

Also it seems some people went "Bulldozer" on Intel, their i5 Haswell at 4.6GHz is faster than their 5GHz i5-2500k at half the power

 

[Binky]

I didn't read any of the OP's post, but really, how can I take your Intel-defense post when your name has Intel in it?

 

[frozentundra123456]

Wow, and in a review of an AMD mobile chip? And no battery tests. Matching a 15W Intel CPU vs a 35W chip.

And why are so many benchmarks left out with only AMD?

Oh the irony is complete!

Seems the reviewer blames Intel for what AMD is lacking. The review is simply full of bad excuses and what is essentially slander.

Yea, that was the problem I saw as well, comparing a quad core, full power mobile Richland chip to a dual core ulv intel chip. Totally irrelevant comparison. He also makes the blanket statement that Richland is faster than Haswell while totally ignoring the higher level graphic solutions for haswell, as well as the tdp. He should have at least qualified that statement that there were faster solutions for Haswell and that even with HD 4400 the results might have been different if a quad core Haswell with a higher TDP had been tested. Obviously this data is not available yet, but this just goes to show again that basically the comparison should have been against full power notebook Ivy, not ulv Haswell.

 

[Magic Carpet]

I can, but how are we going to create a baseline?

Well, I'd really want to see the delta in power consumption between stock (3.4), light OC (4.2), heavy OC (4.6) and top OC (5.0+?). That's all. If you get some time on your hands, I'd really appreciate that.
We all know, that Haswell is more efficient at idle and lighter loads but consumes a tad more power at load, but not significant, to go to war

Also, if you could run wPrime 1.55 (needs to run as Admin) with 1 thread (@ 3.0 Ghz and best OC), while at it :thumbsup:

 

[BallaTheFeared]

Based on what workload?

 

[Magic Carpet]

Based on what workload?

How about latest Handbrake (0.9.9.5530)? You can re-code this to mp4 using High Profile preset and note the times, to make it comparable to other systems.

NB: HB uses some SSE 4.2/AVX instructions, so "ancient" processors will be somewhat penalized. Keep in mind.

 

[sushiwarrior]

Ivy has 22nm and Tri Gate, the rest is pure uarch.

Haswell scales down better than Ivy, and it scales up better as well while having an improved instruction set.

Ivy wasn't good if Haswell isn't, sorry. Haswell is a better chip than Ivy on the same node.

More like "Haswell is pretty much the same chip as Ivy on the same node, one year later."

 

[BallaTheFeared]

How about latest Handbrake (0.9.9.5530)? You can re-code this to mp4 using iPad preset and note the times as well (that would be easy to compare to other systems).

NB: It uses some SSE 4.x instructions, so older processors will be somewhat "penalized". Keep in mind.

Where is the thread with those results