* Posts by Zola

231 posts • joined 13 Nov 2012

Page:

Qualcomm moved its Snapdragon designers to its ARM server chip. We peek at the results

Zola
FAIL

So why bother with Kyro?

It doesn't seem like Kyro is that critical to their mobile roadmap if they can swap out their fully custom design for off-the-shelf IP without any obvious impact on the bottom line, yet it's undoubtedly a damn sight cheaper/faster to use the off the shelf ARM cores.

I always thought Kyro was a case of NIH, and never really justified the additional design cost given the marginal performance gain over stock ARM IP.

At best Kyro is just a dick swinging marketing exercise - no real gain or benefit for the end user, but look how clever we are to design our own stuff.

1
0

Nokia's comeback is on: The flagship 8 emerges

Zola

Re: No wireless charging? No sale.

Too many times I've received a call on a phone charging while tethered to a USB, picked it up and had the phone shoot out of my hands once the cable reaches its full extent. User error? Most certainly. Annoying as hell? Absolutely. USB-C might make charging ever so slightly easier but it's still no match for put down, pick up convenience.

Does wireless charging take longer than a tethered fast charge? Sure, probably. To be honest it's hard to tell as my phone is now almost always fully charged most of the time thanks to having multiple charge plates at home/office, and putting a near empty phone on my bedside cabinet at night I awake to a fully charged phone in the morning as if by magic! Fast charging becomes irrelevant when charging is something you no longer have to even think about (and if I ever did need it, I can always find a cable somewhere).

4
7
Zola
FAIL

No wireless charging? No sale.

20 years ago I bought a TiVo Series 1 (still have it, in daily use) and tried to explain to people the convenience of HDD recording, the typical response was "Why would I want that when I can just put a tape in my machine?" It only took them another 5-10 years until Sky (I'm in the UK, after stabbing TiVo in the back) "invented" HDD recording and for the penny to finally drop among the masses.

Similarly, a lot of people just don't "get" wireless charging, but it really is so much more convenient and puts an end to "battery angst". Once you've tried it, why would you want to go back to fumbling around with USB connectors?

And for that reason, any device that can't be bothered to include the £2 coil for a wireless charging receiver loses my entire business/interest. Which is a shame, as the Nokia device is quite attractive in all other respects.

A replaceable battery would also have been nice, but I can understand from a design point of view that that is a trickier proposition.

10
6

Intel bolts bonus gubbins onto Skylake cores, bungs dozens into Purley Xeon chips

Zola
Go

Early Purley Xeon vs. AMD EPYC benchmarks at Anandtech

Anandtech

Summary:

Purley is best for very large databases and vectorisable HPC code (due to the Xeons superior L3 cache latency), while AMD EPYC is simply great for everything else and whips Intels arse on price/performance.

6
0

Seagate SNAFU sees Cisco servers primed for data loss

Zola
Facepalm

Why are Seagate being fingered for this? It's not their SNAFU!

This is 100% Cisco's fsck up because it was Cisco that assumed rather than erred on the side of caution/professionalism by configuring the drives on receipt from their supplier - or on boot up, fancy that! - with the correct parameters/characteristics required to reliably perform the job they will be doing.

There is absolutely nothing wrong with what Seagate did, or did not do!

(Not a Seagate shill, I just think Cisco are entirely at fault here)

2
2

In the Epyc center: More Zen server CPU specs, prices sneak out of AMD

Zola

Re: Epyc 7351P @ $700

@analyzer: What I'm getting at - assuming these Epyc prices are accurate - is that this gives a ballpark figure for for the cost of a 16c/32t Threadripper. I can't see AMD pricing a Threadripper (at the same Epyc 7351P base frequency) higher than an Epyc 7351P, particularly as the Epyc 7351P includes a range of technologies that won't be included/enabled in Threadripper. Perhaps if the Threadripper is clocked significantly higher than the Epyc 7351P (while still maintaining sensible thermals) then it might allow AMD to charge a slightly higher premium, but not much more.

@John Pombrio: I wouldn't have thought 2.4GHz x 32 threads (Epyc 7351P or Threadripper) would be too slow for a HEDT, maybe slow for a gaming rig, but for a headless build system that spends all day compiling it could be pretty sweet!

4
0
Zola
Go

Epyc 7351P @ $700

At 16c/32t isn't this basically a Threadripper?

Presumably the Threadripper will not have the enterprise features such as memory encryption, in which case it should be cheaper than than the Epyc 7351p, which would make it a fricking steal...

0
0

What is this bullsh*t, Google? Nexus phones starved of security fixes after just three years

Zola
Linux

Jolla provide Sailfish updates for their Nov 2013 device

Latest update received April 2017. And that's not just security updates - that's almost 3.5 years of regular OS updates.

Sure, the hardware is a bit long in the tooth but it's working fine and I've no reason to upgrade just yet - not until there's better Jolla hardware on the market.

0
0

The Psion returns! Meet Gemini, the 21st century pocket computer

Zola

Re: Sailfish OS

Comment from the "backer" Davide Guidi to someone asking for Sailfish OS support:

Thanks for the comment, very interesting! We are at 4YFN, stand P4, at Mobile World Congress and we will go and visit the Jolla stand tomorrow. We have not considered it so far, but you never know!

2
0

Flashy Intel sees the XPoint of solid state

Zola
Meh

“And we've demonstrated on stage"

Sounds pretty desperate if he has to use a stage demo as evidence of their performance! Let's see some independent benchmarks... until then, colour me not convinced.

0
0

Ridiculously small Linux build lands with ridiculously few swears

Zola

Re: Which graphics?

The latest 375.26/340.101 nvidia drivers don't yet build with the 4.10 kernel, and need patching (as usual). Funny how the video drivers from the other vendors (AMD/Intel) avoid the pitfalls that befall nvidia with every major new kernel. It's usually the same memory management API changes every time, too. Although this time the changed HOTPLUG_CPU support clobbers them as well.

7
0

Samsung share plummets – but it’s not because of the Burning Note

Zola
Flame

What's Elop up to these days?

Pretty sure he can save them - he's got experience of burning platforms, after all...

5
0

Czech, mate: Cops cuff Russian bloke accused of LinkedIn mega-hack

Zola
Joke

Can't be long until he tries the Aspergers defence...

The post is required, and must contain letters.

1
0

Linus Torvalds says ARM just doesn't look like beating Intel

Zola

Re: Almost bought a QL

> Don't think so. There's lots of differences between 68K and VAX at a fundamental level, eg. on VAX most registers can be used for most purposes, not so easy on the 68K.

Although I'm talking about orthogonality at the assembler level - the 68K had many different opcodes for implementing the various instructions (unlike the VAX) so yes at a fundamental level the hardware implemented very different approaches but as far as the assembler programmer was concerned the 68K assembly language (ie. which is what I mean by the instruction set in this context) is very similar to that of the VAX.

1
0
Zola

Re: Almost bought a QL

Yes, the very basic windowed GUI was certainly lacking, but the underpinnings (IO channels, job control, pre-emptive multitasking) were all certainly far more impressive, particularly when it was done in 48K.

5
0
Zola

Re: Almost bought a QL

> Don't think so. There's lots of differences between 68K and VAX at a fundamental level, eg. on VAX most registers can be used for most purposes, not so easy on the 68K.

That's exactly what orthogonal means. You could use the 68K and VAX data or address registers as source or destinations (in any combination) with pretty much any relevant instruction. Try doing that with 6502 or x86.

6
0
Zola
Go

Re: Almost bought a QL

I learned to programme MC68K assembler on a Sinclair QL (and BCPL, then C). Then went into my first job where I programmed VAXes in Macro-32, and it turned out the VAX had a very similar instruction set to MC68K - both are/were wonderful (for the time, circa 1988) modern 32-bit orthogonal instruction sets. Presumably Motorola based their 68K instruction set on the VAX as they were that similar. Things would have been a lot harder for me if I'd learned the horrors of 6502 or, god forbid x86, so in that sense I don't regret buying the QL at all!

On the other hand that keyboard, those microdrives... eugh! The fully pre-emptive mutitasking QDOS operating system with SuperBASIC (all in 48K ROM, in 1984) was quite an impressive achievement though. I'd love to see a proper write up on QDOS with input from the original author Tony Tebby.

15
0

WD: Resistance is not futile

Zola
WTF?

"the technology is not that close to productisation."

So it's basically vapourware, just like those batteries in a lab that re-charge in 30 seconds or whatever. Wake me up when they're shipping a product.

2
0

Linux security backfires: Flaw lets hackers inject malware into downloads, disrupt Tor users, etc

Zola
Facepalm

Re: Patch incoming in... 3,2,1

Nexus 7 (2013) with all of the latest Marshmallow updates:

~$ uname -a

Linux localhost 3.4.0-g1fc765b #1 SMP PREEMPT Wed Jun 8 18:49:02 UTC 2016 armv7l

So no worries there - for once being on an ancient kernel is a blessing!

0
0

Get yer gnashers round 64-layer 3D NAND, beam WDC and Toshiba

Zola

Re: Layer limit?

Stacking more than 64 layers is proving troublesome due to alignment issues with the through silicon vias (TSV) - each layer in the stack has to be perfectly aligned with the layer above and below. The more layers, the greater the difficulty of achieving the required TSV alignment. One solution is to combine two 64 layer stacks, creating a psuedo 128 layer stack - this should be relatively easy to achieve and increase yields.

3
0

Brit chip biz ARM legs it to Softbank for $32bn

Zola
Unhappy

Indeed.

Unless ARM need investment I don't see what Softbank is bringing to the table. The risk is that when one of the many other Softbank business units is haemorrhaging cash and dragging down the bottom line, ARM could be sacrificed.

ARM losing its independence sounds like a very bad trade. I wouldn't be surprised if Apple end up as owners of ARM (again) longer term, and that will be very bad news for everyone. Seriously, seriously bad.

23
0

Citrix bakes up Raspberry Pi client boxes

Zola

Re: Internal PS? - what about an "internal Pi3"?

One can't help but wonder when we'll see a monitor with a built-in Pi3 board.

It would be a piece of cake if they added the SODIMM socket for the Compute Module (currently using the RPi1 SoC but pin-compatible and so upgradeable when the CM3 is launched). Monitors have pretty much all the IO ports (USB, audio) required for the Compute Module to become a full computer, so the cost would be pennies (RJ45 port for wired Ethernet and the SODIMM for CM). The CM could even be optional, as it's as easy as a stick of RAM to install.

Connect a keyboard, mouse, network (a $2 USB dongle would enable WiFi) to the monitor, select the Compute Module as an AV Input and crack on...

0
0
Zola

If this little puppy can drive a couple of decent sized displays,

I wonder if it would be possible to add additional Citrix/RPi devices for each additional HDMI display, then it would just be a software/configuration issue and you could have as many displays as you need. Keyboard and mouse would be connected to the "master" device with all the additional screens connected to the "slave" (display only) devices.

0
0

GitHub to devs: 'We ignored you but we never stopped caring'

Zola

SourceForge is still a thing?

Who knew.

Sorry, too little way, waaaayyy too late. SourceForge became a joke years ago, killing DevShare won't make a bind bit of difference. Any projects still on SourceForge tells you all you need to know about those projects - ether no longer maintained, or maintained by developers prone to making really bad development choices.

4
1

BT broadband is down: Former state monopoly goes TITSUP UK-wide

Zola

BT ADSL, South London has been fine all day

See title.

0
0

Qualcomm tries to figure out if Apple, Samsung made it anti-trusted

Zola
WTF?

Surely it doesn't matter who dobbed them in

if Qualcomm are guilty, which apparently they are, and have been found to be, on numerous occasions.

If Qualcomm don't want their competitors dobbing them in to the competition authorities because of anti-competitive licensing, then don't make your partners sign anti-competitive licences for Qualcomm products.

It's not fcking rocket science.

3
1

SpaceX launch is a go for Sunday after successful static fire completed

Zola

Re: Routine Disasters

Never mind, delayed 24 hours - current launch time: Tue, Dec 22 2015 1:34 AM GMT

2
0

Why Microsoft's .NET Core is the future of its development platform

Zola
FAIL

AMD recently dropped .NET

AMD used to use .NET for their Catalyst drivers (ie. full desktop app). They've now switched to Qt, gone from 8 second application start-up to 0.6 seconds, with the added bonus of now being fully cross-platform, even on the desktop. It's not hard to see why they made the switch, and .NET Core won't be of any help.

There are better alternatives to .NET, even if your only platform is Windows.

0
0

Yesterday: Openreach boss quits. Today: BT network goes TITSUP

Zola

More like 3 hours than 3 days, solved already by BT Service page

The problem actually started around 9pm Monday night, so took over 15 hours to fix, which is hardly impressive. Though better than 3 days, perhaps they're under promising and then over delivering.

3
0

AMD sued: Number of Bulldozer cores in its chips is a lie, allegedly

Zola

APUs and "Compute Cores" muddy the water even more

AMD have recently taken to counting the combined number of CPU and GPU cores as "Compute Cores" when describing their APUs, so for example the A10 PRO-7850B has 4 CPU cores and 8 GPU cores, or 12 "Compute Cores" in total.

Although I'm a little uncomfortable with this marketing-motivated move I do understand the distinction but I'm not entirely sure it's necessary or helpful (which is not to suggest that AMD try to hide the number of actual CPU cores, as they don't). However our clueless, dickhead plaintiff would no doubt sue on the basis that he thought he was buying 12 *CPU* cores - after all, he did overhear someone speaking about CPU cores once upon a time.

1
2
Zola

Re: Frivolous legal case, should be tossed out

> But what about memory bandwidth?

At the time of purchase (about 3 years ago) I considered over 14GB/s of DRAM bandwidth to be perfectly adequate, and considering it consistently outperforms Intel i7 quad-core systems of a similar vintage the AMD memory bandwidth (or shared FPU) hasn't proved to be a handicap.

6
0
Zola

Re: So like Intel's Hyper-Threading bullshit

Unlikely because Bulldozer does actually have the physical cores (although as many as half of them may not always be fed with data/instructions, depending on the workload, and depending on who you believe, plaintiff or AMD) whereas hyperthreaded cores are entirely virtual, all of the time.

6
1
Zola

Indeed, AMD do need to significantly improve their IPC. This is what Zen promises, so let's hope they deliver (and you eat your shorts) as a completely dominant Intel in the x86 space doesn't bear thinking about.

10
1
Zola
FAIL

Frivolous legal case, should be tossed out

Can't see this case succeeding, nor should it. There is no doubt that Bulldozer has the AMD stated number of cores, the fact that some aspects of the design is shared between paired cores is well known, add to that if your workload is heavily FPU based you'd have to be an idiot (or a cheapskate) to choose AMD. I selected an 8-core/4-module FX-8350 specifically for kernel and OS builds, mainly because there is so little FPU action (and there is no doubt it has 8 cores).

Unfortunately the guy bringing this case failed to do his homework and is now able to bring a frivolous legal action - I hope he loses and I'd like to think it will cost him a fortune (but it probably won't, which might be part of the problem).

21
5

No 4King way: Dolby snuggles its high-def TV tech into MStar SoCs

Zola

Re: Yeah...

> Does anyone really care that much about high-dynamic range?

Apparently HDR really is the mutts nuts, offering far more noticeable picture quality improvements than even the jump from 1080p to 4K. From all that I've read about HDR, written by people that have seen HDR content with their own eyes, it really is going to be a major step-change, and far more so than regular 4K.

2
1

Google flubs patch for Stagefright security bug in 950 million Androids

Zola

Re: Monthly security updates will soon become a major PITA

Correction: Nexus 7 (2013), not Nexus 5...

0
0
Zola
Mushroom

Monthly security updates will soon become a major PITA

It's all very well Google promising to push out monthly security updates, but the design of the current Android platform ensures that frequent updates will become a major PITA and something I'm sure users will grow weary of pretty quickly.

The problem is that the Android platform takes over 20 minutes - tested on a quad-core Nexus 5 (2013) - to apply even the smallest update. Every application on the device (and I haven't installed many myself, maybe only a dozen, but the number of apps on the device still runs to about 120) has to be (re-)"optimised" - thanks to ART - every time the system is updated. And optimisation is a very, very slow process (I actually wonder if it's only running on a single core, it's _that_ frickin' slow).

1MB update? Boom, 20+ fudging minutes to apply the update.

10MB update? Another 20+ fudging minutes to apply.

200MB update? You get the picture. The size of the update doesn't matter, it's always going to be dwarfed by the colossal time it takes for ART to get it's shit together.

It's a horribly flawed process that is going to become a major burden for users if small security updates are pushed out frequently. I can see myself skipping updates just to avoid the inconvenience of the slow update process (although at least they're unlikely to be as bad as Twitter, who seem able/willing to publish new builds of their app on an almost daily basis with no hint of a changelog - it does make you wonder how crap their developers are).

3
0

Apple tries to patent facial recognition

Zola

Umm... pretty sure my Nokia N9/N950

did this in 2011 - it would analyse pictures in the gallery and recognise the faces of your contacts.

0
0

BT broadband in broad-based brownout and TITSUP incidents

Zola

Now seems to be fixed

Sometime between midnight and 1am.

There never was a public acknowledgment of the service issue in the 48 hours it lasted, resulting in BT customers up and down the country continuing to contact India for "help" only to be told there isn't a problem and it must be their router/landline etc., wasting an hour or more of each customers valuable time.

I genuinely wonder if the lack of public acknowledgment is because of the new regulation that allows in-contract customers to walk away without paying a penny if they fail to receive an adequate service. I guess we'll know the next time there's a prolonged outage - will BT once again lie to and dick their paying customers around by treating them with contempt, or will they behave like a reputable business? I do hope Ofcom are watching...

2
0
Zola

Re: 18 months and still not fixed

Only one question - why are you still with them?

1
0
Zola

> 50p says that a firmware update has caused this. It's been getting gradually worse as the update is pushed out to more routers.

I use my own router/modem (Netgear DGND4000) and have never connected the supplied HomeHub to my line, so I highly doubt this is a HomeHub-specific issue - the problems are all upstream.

0
0
Zola

Re: Bt have finally admitted there is a problem

> Hmm. So your deicsion to not download Windows 10 must have reduced the amount of W10 downloads going on in the UK today by, what, 30%, 40%?

Obviously not, but Windows 10 background downloading is being claimed as a potential cause and I'm just saying that's not the case with me. Other users downloading Windows 10 could be a factor, as in contention issues, but not at 3-4am and besides bandwidth isn't even a issue, you could have 30Mb+ bandwidth but still can't connect to a website. It's entirely down to packet loss/routing issues within the BT network.

Oh, and BT have now removed the status announcement from their status page even though the problem is ongoing.

1
0
Zola

BT forum thread - currently 19 pages and growing rapidly.

1
0
Zola

Bt have finally admitted there is a problem

"Wed 29/07/2015 at 10:26

BT customers are having trouble loading webpages"

All other service incidents prior to this are unrelated.

And it's not Windows 10 - I turned that shit off, have not seen any unexpected downloads on my SNMP graphs, and bandwidth is not the issue. It's packet loss or a routing issue (secure/encrypted connections seem to fair the worst - https, rsync over ssh etc. both failing).

2
0

So just WHO ARE the 15 per cent of Americans still not online?

Zola
Trollface

15% of Americans not online?

Who'd have guessed BT would have cornered 15% of the US ISP market.

17
0

Slashdot, SourceForge looking for new owners after parent dumps them on the web's doorstep

Zola

SourceForge is so dated

I'm surprised any active projects continue to use it in preference to a superior and adware-free alternative such as Github. I can understand projects that are no longer maintained parking their code there, but for active projects to continue using SourceForge just seems bizarre given how crusty it is, and now all the adware nonsense.

3
0

Help! Our Virgin Media TiVo boxes are stuck in a loop! Help! Our Virgin..

Zola

Rebooting used to be TiVo SOP

If the latest TiVo boxes are anything like the original Series 1 TiVos (mine still going strong after 15 years) it's standard procedure to reboot in the event of an upstream failure - in the case of the S1 it will reboot if it's not receiving a TV signal (the assumption being the encoder, decoder or some other part of the chain has crashed/locked up). It's an extreme, but pragmatic, solution for a problem that happens very rarely.

If some part of the Virgin network or cable system is down this could easily explain the behaviour of the box which will keep rebooting until normal service is restored.

3
0

Wi-Fi was MEANT to be this way: Antennas and standards, 802.11 style

Zola

MU-MIMO - phones will still be as shit as they are now

Can someone explain to me how MU-MIMO "will solve the poor phone performance problem" as it seems to me the phone is still stuck using only one antenna, although the AP will now better utilise it's multiple antennas allowing the AP to communicate concurrently with devices other than the phone.

However the phone itself will still perform as poorly as it would have done without the AP having MU-MIMO, so how exactly does MU-MIMO "solve the poor phone performance problem"? It doesn't do anything of the sort, at least not based on the description in this article... unless the article meant to say the phone is a performance lead weight for AP throughput?

2
0

Hordes spaff cash on Chip titchyputer to rival Pi (maybe)

Zola
Linux

software and hardware are "totally open source."

This claim seems somewhat unlikely considering it's based on AllWinner hardware - just Google for "AllWinner GPL Violations" to see what kind of attitude AllWinner have towards "open source".

Anyone that ships AllWinner hardware and releases software is, by definition, going to be in breach of the GPL too as AllWinner deliver binary blobs to their hardware customers, so it's impossible to release "all software" and be GPL compliant.

I'm aware of hardware companies that have had to move their projects away from AllWinner (to Amlogic, Freescale, HiSilicon etc.) just to avoid the GPL stench that pervades AllWinner.

8
0

Hackable media box based on the Raspberry Pi compute module: Five Ninjas Slice

Zola

Although not yet available, a Pi2 Compute Module should be a (relatively cheap, sub-£30) drop in replacement although you'll need to reconfigure Kodi and rescan your library, etc. - a minor inconvenience considering the significant performance boost from the quad-core ARMv7 SoC and extra memory.

1
0

Page:

Forums

Biting the hand that feeds IT © 1998–2017