HP tested webOS on an iPad, it ran twice as fast? Not so fast... | webOS Nation

HP tested webOS on an iPad, it ran twice as fast? Not so fast... 40

by Derek Kessler Fri, 19 Aug 2011 9:52 am EDT

The Next Web published an article this morning that we find rather interesting. In it, they claim “a source close to the subject” revealed to them that HP tested webOS on an iPad (presumably an iPad 2) and found that it was able to “run the platform significantly faster than the device for which it was originally developed.”

Twice as fast, on a device with a processor built on the same ARM Cortex-A9 architecture as the Qualcomm APQ8060 in the TouchPad, but clocked 200MHz slower. Twice as fast, on a device with half the RAM of the TouchPad. Normally we’re not one to call “bull” on something, but this just doesn’t smell right. As far as hardware is concerned, the TouchPad should be more powerful. Sure, Apple’s been able to build some great efficiencies into the dual-core 1GHz A5 chip, that’s what you get to do when you make your own silicon. But enough to make a 1GHz chip run an inefficient OS twice as fast as 1.2GHz chip?

We don’t buy it. There are a few reasons why webOS seems so much slower than iOS, but the hardware isn’t one of them. webOS 3.0 still isn’t fully hardware optimized. It doesn’t take full advantage of the Adreno 220 graphics chip in the TouchPad, and despite being all HTML5 and CSS3 friendly, webOS doesn’t support things like CSS transformations. Instead it relies on more processor-intensive methods to make webOS looks slick. The problem with the performance experience of the TouchPad and webOS 3.0 is not the TouchPad’s problem – HP threw the best silicon they could at the problem, and were prepared to throw even more.

correction: The Next Web has clarified to us that they in fact tested Enyo apps on the iPad, which while possibly faster, has significantly better HTML5 implementation with Apple's mobile Safari.

Additionally, The Next Web’s source claims that HP had the TouchPad hardware ready and complete when they announced their plans to acquire Palm back in April 2010 – 16 months ago, and just a three months after Apple unveiled the original iPad. More cold water: there is no possible way HP (1) had an iPad competitor ready three months after the iPad was announced and only a few weeks after it shipped, and (2) had a tablet just sitting around with a slew of chips that hadn’t even completed development yet.

The rest of this report is perfectly believable, as it more-or-less jibes with the information we’ve already seen: HP had plans to release a 7-inch webOS tablet: the TouchPad Go. As one might have surmised from it passing through the FCC, this tablet was already in production and planned for a launch in the near future, but now we don’t know if that’s ever going to happen. According to The Next Web’s source, the TouchPad go was both “a better looking and nicer feeling device,” which HP believed would outsell the original TouchPad.

It’s also safe to assume that HP was already working on the next generation TouchPad, a device that The Next Web reports that the new tablet was going to be lighter with a higher-resolution display and metal case. This TouchPad 2 would have been an attempt to head Apple off at the pass instead of playing catch-up with the last-generation iPad. If this report rings true, it makes us mourn even more HP’s decision to can webOS hardware development. We had little doubt that HP was finally putting all of their development muscle into making good hardware that would really appeal to customers, and the fact that we’re not likely to see it cuts us deeply.

Source: The Next Web


I think they meant they tested Enyo apps on the TouchPad and on the iPad.
And it's true I think, Enyo apps run much faster on the iPad & Co. than on the TouchPad.

I agree - mobile Safari is a fantastic browser that is tightly integrated with the GPU, which makes all the difference. Having tested HTML5 apps on iOS, Android and webOS, I saw by far the best performance in iOS.

The fact that after three years of development that Palm couldn't get their browser fully accelerated on the GPU is probably the main cause of their downfall.

Not that I believe that engineers had less important things to do than spend weeks on end recompiling WebOS3 and optimizing WebOS on iPad2 without (presumably) any help from the hardware supplier, but the article does suggest both scenarios were tried: Running the whole environment resulting in a 2x speed improvement, and running the emulator in the iPad browser was "many times faster".

This is probably what was meant.

Good job, Derek.

The original article on Engadget was botched from the getgo.. they even originally declared that the TP had a single core chip and then recanted that.

And, Mobile Nations should have let you speak more.. I don't think you ever got to really give your full prognostications about the future of WebOS...

Umm...the APQ8060 isn't a Cortex-A9.. it's still an A8 chip for one. That's why the extra 200mhz is needed to try and match the other current 1Ghz A9's already out.

Someone Correct me if I'm wrong...

The article does make sense in a certain way. It says they tested the JavaScript and CSS parts in the iPad Browser. This certainly is achievable. The result was it ran faster on the iPad than on the TouchPad.
Apple has aggressively developed its JavaScript engine to be very fast during the last few years. webOS development never focused on getting the most performance out of JavaScript. E.g. the iPad browser uses hardware acceleration for CSS transformations. That's something webOS doesn't even do in software.

The article says "The hardware reportedly stopped the team from innovating beyond certain points because it was slow and imposed constraints, which was highlighted when webOS was loaded on to Apple’s iPad device and found to run the platform significantly faster than the device for which it was originally developed."

Nothing about loading just apps.

> With a focus on web technologies, webOS could be deployed in the iPad’s Mobile Safari browser as a web-app; this produced similar results, with it running many times faster in the browser than it did on the TouchPad.

This makes it pretty obvious. The author of that article didn't draw the correct conclusion. It's not the hardware as he thought, it's the software used to display and run web technology.

That new TouchPads were already quite in the development stage supports my theory that HP didn't shut down webOS because they specifically don't believe in it but rather because they're getting rid of all their consumer hardware which includes among other things tablets and phones.

Yes, that's why their announcement included the PSG.

plus though maybe not twice as fast, I believe significantly faster. Apple is know to use the best memory and buses.... I think that plays a huge affect on performance...

Well the iPhone 4 received a double speedup of JavaScript going from iOS 4.2 to 4.3.
So this is possible.

Derek ? why would you not believe ? I think the issues were valid. HP was working on a device with tablet form for quite sometime. Do you remember Ballmer showing the HP Windows tablet ? They won't appear magically. HP wanted to be in tablets at that time because they know Apple will take over them eventually. They wanted to do magic with webOS without throwing more money. They lied and lied and lied , remember "Marathon" not "Sprint" comments from every top executive. At the end of the day it was bulky hardware married to unfinished software which did not sell well. Now HP needs quick profit so they went for Autonomy (10 billions is a lot). They just want to get away from the low margins.

Long story short my .02 , if Palm released a slab with 4 inch screen back in 2009, we wouldn't be talking about sliders,oreos, lags and most importantly returned phones.

The HP Slate 500 ran an Intel chip and Windows 7. It was a completely different device in literally every single respect than the TouchPad.

I was talking about prototypes. As a technology company with laptops business , HP's R&D must have tried few combinations was my point. I believe that they must of have tried Ti, Marvell and Qualcom chips to test waters.

However you are correct on the inefficient OS. That is a 100% fact.

It's easy to say that about slabs now.

It was easy then as well. What wasn't obvious about that even in 2009?

seriously whats the point, Someone needs to switch the Webos life support machine off.

You don't have to come here you know.

Derek, wasn't there an article in P¦C suggesting the TP hardware design had been complete and that it was waiting the software to catch up? Before the Feb HP/Palm show? I suppose if we assume that the TouchPad was designed as a slate without WebOS in mind, and with a different processor and memory combination at first, then the article does make some sense. It would have been a matter of upgrading the device with a better processor and other components rather than designing a device completely from scratch.

(And given the dissection reports of the TP, it is certainly possible to upgrade it on a partial basis as you would a PC)

They had CSS transforms. What they lacked was hardware acceleration for those transforms.

That last paragraph tells the tale of HP's failure. They were finally getting around to putting their muscle into the hardware but it was too late.

What they should have done the day they took over was scrap the Veer and TouchPad hardware and actually "double down" on new hardware. The Touchpad could have been lighter and thinner, the Veer should have never existed but make a 4" slate instead. The Pre3 could have been released in April or May if HP would have prioritized it and it would have held the wolves at bay until they could release the slate phone and a slimmer TouchPad in September.

It's mind boggling to see people that are supposed to be smart make these awful decisions. All the marketing resources in the world and they couldn't figure out what the consumer wanted. I'm still in shock.

While some people can say that hindsight is 20/20, I agree with you and I believe that these ideas were what most Palm fans were hoping for back then.

Yep, but Precentral banned or would give infractions to most of the ones who kept posting about things like this.

Hindsight is always 20/20, but in this case everything I mentioned is foresight. I don't think anyone IN THE ENTIRE WORLD thought to themselves "Man, that Pre is awesome, I just wish it was SMALLER" XD As webOS's biggest apologist, even I said "wtf" when they announced the Veer. I'm not saying there isn't a market for that form factor, but it wasn't right for webOS at this time since they didn't have a hold in the marketplace.

Now if Apple releases the Veer with iOS tomorrow they sell 20 million by the end of the day. If and Android version was released it too would probably do pretty well. But with an OS and an ecosystem that hasn't matured yet? It's a waste of time. Everyone knew this except HP. They totally misread the market.

Since when has research ever stopped a rumor site from spreading unreliable news as facts.

Hardware aside, there are fundamental problems with how WebOS is coded. From day one with the Pre- on. They never figured out how to do code for gpu acceleration. Sure we heard they'd hire this hot shot gpu encoder to then crickets. Then HP takes over and with all they vaulted programmers there both the TP and Veer still didn't take advantage of the chip. Never mind that it took a day and a half for the devices to boot up. Mojo may have needed a rewrite and that's where Enyo came in. But they still couldn't get it to do it right.

I hope that now if Samy or HTC can get it (HTC EVO W/ BEATS AND WEBOS PLEASE PLEASE!!!). It can be made to properly shine. Saying that I won't be holding my breath.

Man, if HP released a higher resolution TouchPad or a smaller TouchPad go, I'd be all over that like a one legged **** in an **** kicking contest.

The graphical prowess from the iPad 1 to the iPad 2 increased significantly. This news comes as no surprise to me.

Yes, the TouchPad might have a fast chip in it, but you begin to see the cost-cutting ways that HP managed to 'recycle' some of Palm's work. Didn't spend the adequate R&D hours improving upon what Palm originally brought in. Basically built it just like a little computer, slapped HP on the back, and then said in February that _this_ is the result of our hard work. They have learned the hard way what this meant when you have juggernauts like Apple bringing the latest and greatest. You cannot defeat Goliath with a shiny river stone, it must be something much better that even Apple would be jealous of.

Same goes with the Veer and Pre3. Rubenstein must be to blame for all that nonsense, ignoring market demand and going in the complete opposite direction hoping to start something new. HP played it cheap, wanting to turn the biggest profit because of low R&D, and said "Hey Palm! Great work. These are interesting concepts, and it might fare well with all these big phones everywhere! Since you've already done most of the work, let's get this very small phone, this other phone that is supposedly the flagship, and this plastic tablet to market as soon as we can! Can you do it? Can you do it? I have some HP engineers that I can lend you. They are not real keen on deadlines but they sure are handsome! Let's get this done so we can present at Feburary, our shareholders are counting on us to present something new to justify this new big purchase of ours!"

I'm not sure why you pin all this on Ruby. It took a team effort at HP to **** up this badly.

True, but for whatever reason HP followed Ruby's failed vision. He killed Palm and then finished webOS off inside HP. Blame can be spread across the board but the guy that should shoulder the most blame is definitely Ruby.

Wow the writer of this article is in complete denial. Face it HP was never in the process of making a new touch pad. HP strung everyone along and lied the whole time. HP is a company with no credibility anymore.

Time to wake up and smell the coffee. The last four phones from Palm/HP were all the same, it was clear from the original Pre that it was a dog design and Palm/HP should have designed a slab type phone with great specs and power.

The only reason HP even made the touch pad was because they all ready had the design done but then ported WebOS to it. It was going to fail no matter what. HP never put any thought or effort to try and make a new WebOS phone or tablet. It was all left over junk.

The definition of insanity is try the same thing over and over expecting difference results.

HP will never again have a product in my home when I buy things. I have a HP Blue Ray player that HP abandoned the exact same way they have done with the phones and the touch pad. HP as a serious company cannot be trusted.

I don't agree about form factor. If Palm had released the Pre3 in June/09 instead of the original Pre then I think their story is much different. Would they have knocked off iOS and kept Android down? Doubtful but they would still be in business.

The iPad2 has a cortexA9....the Touchpad has a scorpion core which is a variant of CortexA8. That is why the iPad2 is faster. End of story. Please fix your broken article.

Kind of like kicking a dead horse, don't you think. If I can find my receipt My TP is going back today!! (13 days old )

Who cares?

The CPU/GPU combo on the iPad 2 is fairly superior (especially the GPU), it isn't twice as fast, but it's faster (the iPad 2 has the best internal hardware of any tablet in the market).

The Cortex A9 on the iPad, for starters, is out-of-order, while the Snapdragon on the Touchpad is in-order, it's more like Cortex A8 than A9.