Indian IT Industry

Singha · Post by **Singha** » 07 Sep 2009 17:35

netz & chipz are like the last two surviving dinosaurs eating up all the vegetation they feel like because rival dinosaurs have perished. they are formidable but ultimately could be evolutionary dead ends.

due to climate change or meteorite impact - the age of mammals has dawned. smaller, smarter, cheaper, self-managed is where all the future could be - not yet more iterations of tired 'big-iron' boxes that cost $1 mil in capex and need a batallion of certified experts to manage and upkeep.netz is desperately trying to wear new
clothes in cloud, virtualization, green, energy grids, flip cams, anything thats not tied to costly big-iron gear.

while nobody could fell these last great swamp beasts in a direct contest, time and environmental changes itself have felled many a great species and empire in the past.

a chill wind blows across the steppe.

manish · Post by **manish** » 07 Sep 2009 17:45

Meanwhile, Abu Dhabi keeps pumping money into semiconductors. First they took on AMD's fab unit, now its Chartered Semiconductor's turn:
Abu Dhabi to Buy Chartered for S$2.5 Billion, Challenging UMC

Sept. 7 (Bloomberg) -- Abu Dhabi agreed to buy Singapore’s state-controlled Chartered Semiconductor Manufacturing Ltd. for S$2.5 billion ($1.8 billion) to create a challenger to the world’s second-biggest maker of customized chips.

Advanced Technology Investment Co., an investment company owned by Abu Dhabi, will pay S$2.68 a share in cash, ATIC said in a statement today. The offer, backed by Chartered’s largest shareholder, Singapore’s Temasek Holdings Pte, is two Singapore cents higher than the chipmaker’s closing price on Sept. 4.

A purchase would allow Abu Dhabi to combine its Globalfoundries Inc. venture with the maker of chips that run Xbox 360 game consoles, threatening United Microelectronics Corp.’s rank as the second-largest contract manufacturer of semiconductors. Temasek is ending a 22-year investment in unprofitable Chartered, which has eliminated workers and cut overtime to reduce costs.

This is one industry where things have been pretty bad I guess? There have been a bunch of spinoffs/sales (Intel-XScale, Moto-Freescale, Philips-NXP, AMD-Globalfoundries etc) followed by costly LBOs (NXP, Freescale). Where are they headed? Will fabless be the only way forward for the non-Intel types? I read somewhere that 2008 saw the appearance of fabless cos for the first time in the Top 20 (Qualcomm I think).

IIRC sum was with FS - some gyaan on the scene there perhaps?

bart · Post by **bart** » 07 Sep 2009 19:24

Singha wrote:netz & chipz are like the last two surviving dinosaurs eating up all the vegetation they feel like because rival dinosaurs have perished. they are formidable but ultimately could be evolutionary dead ends.

due to climate change or meteorite impact - the age of mammals has dawned. smaller, smarter, cheaper, self-managed is where all the future could be - not yet more iterations of tired 'big-iron' boxes that cost $1 mil in capex and need a batallion of certified experts to manage and upkeep.netz is desperately trying to wear new
clothes in cloud, virtualization, green, energy grids, flip cams, anything thats not tied to costly big-iron gear.

while nobody could fell these last great swamp beasts in a direct contest, time and environmental changes itself have felled many a great species and empire in the past.

a chill wind blows across the steppe.

As we move from a world of large quantities of interconnected physical infrastructure to virtualized applications running on a random cloud, will Cisco's hardware still be relevant? Branch offices might just have a single large generic server box where the router, firewall, and Layer-3 switch are all images on a hypervisor.

http://vyatta.com/quiz.php

Singha · Post by **Singha** » 07 Sep 2009 19:45

Amazon allegedly runs its entire EC2 (elastic compute cloud) on freeware version of Xen hypervisor, not even the Citrix supported Xen products.

tools to provision, manage, probe and protect the cloud are quite hot with startups apparently. as is "service chaining" where multiple virtual machines deliver some service as a collective entity.

bart · Post by **bart** » 07 Sep 2009 21:02

The only thing that can't be made to go away by virtualization is storage. No matter what software you run, you need somewhere to store the data in some form or the other. In fact the more virtualized the world becomes, the more storage architectures come into the picture, for example if you run a datacenter where you want to move a virtual server from one rack of hardware to the next without service disruption, you need some kind of SAN in the background to make it work.

Hence EMC and Vmware are a perfect fit for each other. Cisco's attempt at server hardware was probably not so much an attack on HP/IBM as perceived, but rather a way to make sure it could still provide networking solutions to enterprises in a future virtualized world where Cisco might replace its current devices with hypervisor images of some kind. That and tie it all in with FCOE to make an end-to-end sever/network/storage solution.

Tanaji · Post by **Tanaji** » 08 Sep 2009 14:03

Not India related, but this thread is going OT anyway

O2 and TMO are merging to create a massive entity in the UK market with 37% share. There is going to be a fallout from this.

manish · Post by **manish** » 08 Sep 2009 14:03

Tanaji wrote:The other side to the story is that it is currently not very profitable to be a wireless equipment maker (apart from base stations). You are either squeezed by players like Huawei that use slave labour to reduce their costs (or outright steal as they did with Cisco) or you have the young startups like Starent that snap at your heels in terms of features and prices. Add to that with the advent of COTS stuff like ATCA, getting margins from proprietory hardware is out. Only way to make profit is hoping for support contracts... but if you are the size of AlcaLu, that only takes you so far....

Chambers is as smart as they come, unless AlcaLu sells individual divisions they wont buy. I dont know much about what patents AlcaLu holds in 4G or related areas, but that may be the tempting offer for some. Witness how RIM is salivating and hoping to get hold of Nortel's LTE patents (Ericsson got the licenses with the purchase but not patent ownership).

Thanks for the insight Tanaji. Starent esp. was savage in poaching from other big telecom players back in 2007-08. They used to offer fairly lucrative salary+ESOPs to even 2yr workex types.

Meanwhile, more talk of consolidation...
A shot at survival for threatened vendors?

The new chief executive officer of Nokia Siemens Networks certainly has set himself a tough task if his words last week are to believed. He told Reuters that he believed the current global vendor market—as in the vendors who can genuinely project globally to win contracts—would halve in number from six to three. His self-defined goal is to ensure that NSN is one of the survivors.

Certainly others seem to believe that consolidation is in the wind. Alcatel-Lucent’s stock price surged last week on rumours that it might seek a merger with either Huawei or Nokia Siemens Networks. Imagine that: after a couple of years of expensive re-brands and traumatic restructures if the marketers and business process specialists had to work out what to do with an entity monikered Alcatel-Lucent-Nokia-Siemens (ALUNSN?).

The current live dismembering of Nortel is a sign that no one in the industry can treat their immediate future for granted.

BTW, the new NSN CEO is Indian - Rajeev Suri.
And an interesting, even if improbable possibility that we have all been hoping for...

So are things really that bleak for the vendor community and, by default, their customers? Perhaps not. There has been talk in India of creating a homegrown champion to build the nation’s networks—a driver which of course inspired the creation of the likes of Alcatel, Ericsson, Nokia et al in the past.

Well we tried, didn't we? ITI was the 'National Champion'...

Singha · Post by **Singha** » 08 Sep 2009 15:03

there is only one competitive homegrown telecom maker in India - Tejas. but they pretty much 100% into optical and metro eth only. could function as backhaul boxes for bband and wireless but they do not any telecom boxes.

vishwakarmaa · Post by **vishwakarmaa** » 08 Sep 2009 15:36

Rajeev suri is an Indian employee of a Gora company. He is paid to bring benefits for the company. Not for how India succeeds in her semiconductor dreams.

Gora knows how to suck up the talent out from developing countries. Gora grows, those countries suffer due to brain-drain.

We need a national policy to stop this brain-drain and better "national" and "domestic" business and invest climate rather than running after Gora funds to come and take brains away from soil.

Make it mandatory for passing out IITians and IISc's to spend atleast 3 years on R&D posts within India, on National basic science projects, which benefit National science and improve the home sitation than wasting brains in Bell labs whose patents are "closed" knowledge to Gora countries.

manish · Post by **manish** » 08 Sep 2009 15:58

Singha wrote:there is only one competitive homegrown telecom maker in India - Tejas. but they pretty much 100% into optical and metro eth only. could function as backhaul boxes for bband and wireless but they do not any telecom boxes.

Ah I remember Tejas. They had an office in J P Nagar IIRC. Nortel was a minority shareholder there with 10% or so shares.

vishwakarmaa wrote:Rajeev suri is an Indian employee of a Gora company. He is paid to bring benefits for the company. Not for how India succeeds in her semiconductor dreams.

vishwakarmaaji, nobody is extolling Rajeev Suri here. It is true that he is working for himself and his bosses, as he should be. The point here is that we have thrown away a golden opportunity to build a homegrown technology/telecoms powerhouse, despite having one of the world's most vibrant and competitive markets for telecom. The so called 'demand conditions' couldn't get better than this.

Despite all the setbacks they have had, Huawei had set and still maintains an internal target to make India its #2 market globally and seconf only to their home market and this year's numbers seem to be headed in the right direction for them. EU has AlcaLu, NSN, khan has Cisco, Juniper and gadzillion other startups and niche players, Japanese with Fujitsu (?) and the Chinese have Huawei/ZTE. We are the only ones missing out bigtime.

I am sure that you would agree that the telecoms and high-tech industries are 'strategic' industries for any aspiring power. Look at how jealously khan guards access to its telecom networks (Global Crossing case comes to mind). Look at the kind of concerns UK, Australia and India have been raising wrt Huawei. We need to move, and move fast.

vishwakarmaa · Post by **vishwakarmaa** » 08 Sep 2009 16:06

With a puppet ruling the GoI, we can move fast but only in downwards direction.

Recent blunders by puppet regime -
- asking the western firms to provide telecom snooping gears.
- selling national stock exchanges stakes to western speculator funds(Goldman Sach) and regulators.
- selling BSNL stake to AT&T.(lets hope marx union stalls this)

Japan's situation was much worse than India's, having lost wars but now, ours is of paramount "ignorance" that we have hopeless leaders like MMS who refuses to see any potential in Indian engineers and scientists.

Japan risen from ashes to peak. India is doing reverse. We are going from peak to ashes. We were at peak, during 1980s, now we are headed downwards since then.

vishwakarmaa · Post by **vishwakarmaa** » 08 Sep 2009 16:10

I still think Vajpayee regime was good for Indian sciences and home-grown capabilities.

MMS regime is destroying home capabilities in favour of foreign. He is committing strategic blunders for small tactical gains.

Tanaji · Post by **Tanaji** » 08 Sep 2009 17:28

Personally, I think the future belongs to more self aware networks and boxes. I don't mean SkyNet-ish type of things, but boxes that require minimal configuration and management over the life cycle. For any given carrier or telco, a tremendous amount of expense is spent simply managing the box for its health, performance and configuration. This expense is due to the skilled personnel required to make sense of the arcane commands/combos used to provision the box, take action on a fault, patch management and general upkeep.

Imagine if we have a box that takes out the chore of doing the above. If a component fails, it simply marks it out of service, sends an automated request to the ordering system to get a new one, if there is a spike in demand (say huge number of calls due to Ganesh idols drinking mills) it will automatically assign more resources, do automatic patch management (Microsoft WSUS type) etc etc. Some of this is already available, but not at the carrier level and not in a unified way. IF this happens on a large scale, network management will require a tenth of the people it does now... of course it will also mean I wont have a job, but this is something the network should morph into

arun · Post by **arun** » 12 Sep 2009 10:56

From the RBI Monthly Bulletin of September 2009:

Survey on Computer Software & Information Technology Services Exports: 2007-08

Highlights:

*Total Computer Services and ITES/BPO services exports of India during 2007-08 was estimated at Rs.1,40,200 crore (US$ 34,841 million), of which computer services exports was Rs.1,07,438 crore (US$ 26,699 million).

*Of the total software exports (Computer and ITES/BPO services), around 74.6 per cent was through offsite services while onsite software exports accounted for 25.4 per cent.

*The United States (US) was the major destination for software exports accounting for 63.2 per cent. The share of the United Kingdom (UK) was 14.0 per cent while Other European countries contributed 12.6 per cent in India’s total software exports.

*US Dollar was the major currency for invoicing software exports, with a share of 75.4 per cent. Pound Sterling and Euro accounted for 12.8 and 6.4 per cent, respectively.

Read it all here (Note: 720 KB download):

RBI Monthly Bulletin

KrishnaMu · Post by **KrishnaMu** » 24 Sep 2009 03:36

Hi folks,

What is wrong with this moronic national tabloid news paper. It is disgrace, whom should report this?

http://infotech.indiatimes.com/quickiea ... 604056.cms

This news paper published Indian map with chopped of Kashmir in the following article twice..

thanks,

PS may i am posting in wrong forum??

manish · Post by **manish** » 06 Oct 2009 12:13

X-Posted from the telecom folder:

Breaking news on CNBC-TV18: AlcaLu is reportedly looking to sell off its India R&D units. Apparently they are planning to outsource the functions in future and Wipro, Infy and Cognizant have been named as potential buyers. The units have been reportedly valued at Rs. 250cr.

No links yet.

Avinash R · Post by **Avinash R** » 06 Oct 2009 14:12

KrishnaMu wrote:Hi folks,

What is wrong with this moronic national tabloid news paper. It is disgrace, whom should report this?

http://infotech.indiatimes.com/quickiea ... 604056.cms

This news paper published Indian map with chopped of Kashmir in the following article twice..

thanks,

PS may i am posting in wrong forum??

Thanks for bringing that error to our notice. Used the feedback link to contact them about the inaccuracy. Let's wait for the response.

sinha · Post by **sinha** » 07 Oct 2009 10:10

vishwakarmaa wrote:I still think Vajpayee regime was good for Indian sciences and home-grown capabilities.
MMS regime is destroying home capabilities in favour of foreign. He is committing strategic blunders for small tactical gains.

Vishwakarmaa

Purely from Indian IT Industry perspective, this govt has been a star performer. The number of mission mode projects on now gives ample room for whole IT Industry to innovate and grow within India.

Examples
a) R-APDPR - all PSU Discomms are going for standardized deployment of metering, billing, energy saving, GIS systems across whole of India (say approx 400-500 crores each)
b) state SDCs/SWANS : for connectivity and hosting (take 100-200 crores each)
c) Homeland security initiatives - CCTNS, NATGRID, Immigration systems (see MHA semi-annual report card on MHA web site) [ 2000 crores for CCTNS ]
d) Airforce systems upgrade (dont know the exact budget )
e) UID kicking off - hopefully multi-biometric based de-duplications (PAN and UID together had a budget of 2000 crores)
f) Department of Industrial Promotion - eBiz project (800 crores)
g) State Portal for eGovernance and automation (~ 80 crores each)
i) Department of Post IT Modernization (4500 crores)
j) CPC systems for CBDT (~ 300-400 crores)

Regarding whether Indian science/tech Industry is being given a boost by this - I think NISG, NeGP, CDAC have all been innovating with these sets of projects. Unfortunately you will find that Indian IT Major focus has been services to the west/Europe and they havent spent enough R&D budgets on developing technologies for India. For example, no vendor today will have a credible technology for handling billion odd dedpulication (text or biometric). Only one Indian vendor has AFIS technology homegrown. It will be the imperative of Indian IT to innovate - Govt is giving the projects.

I feel hopefully we will see some more funding from Govt purely for research in PPP Mode - some things which come to my mind are
a) advanced analytical models for counter-terrorism - copying blidnly what US does may not satisfy our needs for MAC/NATGRID
b) innovate on Biometrics - standard face recognition/ fp technologies have limit - but one can definitely cut into contactless 3-d fingerprinting esp in Civilian apps given higer accuracy and enrollment rates and lack of legacy of flattened prints.
c) Semantic services - syntactic services have limits.
etc etc

Go back to BJP led govt and count me the number of high profile mission mode/transformational projects which yielded anything for India IT.

Cheers

negi · Post by **negi** » 09 Oct 2009 21:24

Netzilla is upto something big; perhaps GD and Vina ji can throw more light on this.

Cisco Teaches Routers to Act Like Servers

Cisco Systems has found a way to accelerate its push into the server market. The company is simply turning many of the millions of routers it has already sold into servers.

Typically, a router sits in a data center or communications closet somewhere just pushing around data packets. It’s sort of like a cop being punished with traffic duty. Meanwhile, the servers — the detectives, if you will — are back in the office working on grander things.

Well, Cisco has been hammering away on something it calls the Application Extension Platform, or AXP, that it thinks can change this relationship between routers and servers.

At its core, the AXP gives software developers a way to write applications that run on an additional bit of hardware that sits alongside Cisco’s routers. In many cases, the applications written for the router can take on jobs typically handled by a small server sold by companies like I.B.M., Hewlett-Packard and Dell.

To drum up interest in AXP, Cisco held a contest for developers who could write the most compelling, router-friendly software. And, this week, Cisco officially began doling out cash prizes to the three big winners.

Tanaji · Post by **Tanaji** » 09 Oct 2009 21:30

There is the small matter that router hardware is under control of network admins who are a paranoid lot. Imagine the horror and hair pulling when they find a code junkie will put an application on top of it and will require major access. And imagine the further amusement when they find the said code junkie has inadvertently left backdoors open or other faux pas.

No doubt it is technically feasible, but it will be a brave manager that will over rule the doom and gloom predictions of his network admins...

Singha · Post by **Singha** » 09 Oct 2009 22:07

juniper also has something along those lines. kind of offering some API for third-party applications to use the operating system and other infra but not giving them full access ofcourse. these apps could be anything that can use spare cycles
or simplify things. for a long time there have been "service blades" that go into
slots and do some specific thing on all/part of traffic like run firewall / ipsec / traffic analysis or such on very high rates of traffic. here there could be a generic service blade with cpu, ram, a pair of local 500gb 2.5" HDD and backplane connection to the rest of the
chassis.

but there is the little problem that in IOS classic everything including all processes are kernel mode, there is a single flat shared address space and hence no per process address space protection or user mode.

IOS XR, linux, nxos and Junos do not suffer this issue. so most likely we will see
these third-party ideas demoed on such.

Tanaji · Post by **Tanaji** » 10 Oct 2009 22:32

Singhaji

Can you imagine the trouble tickets for such a contraption? How about patch management? Generic ticket comes up about a router dropping packets, and a massive pi**ing match breaks out on what is responsible: high usage due to the said app running on the router or whatever. Another perfect example:

App junkie: Your patch just broke my app, it will take 3 days to fix, get it out.
Network admin: No, it fixes a known IPSec vulnerability, cant be backed out.

And the war continues...

CalvinH · Post by **CalvinH** » 11 Oct 2009 03:23

This is not a practical idea of lot of reasons apart from operational issues Tanaji has mentioned

1. How many routers do companies own - How much more operational capacity can be increased by loading the available routers. The server/router ratio would easily exceed 100 for a normal datacenter in numbers only. If you look at the processing capabilities the ratio would be far higher. The gain in operational capacity would be negligible.

2. Sometimes telco or ISP owns the routers instead of companies, how would it help them or the company they provide service too.

3. Why would someone want to run an app outside the firewall where most of the routers are (boundary)

If this is being used to provide a firewall+router combo in a single device may be it makes sense. Firewall or some IP traffice management s/w are the only viable applications that can be run on a router.

Singha · Post by **Singha** » 11 Oct 2009 11:31

I saw it iirc in some of the smaller CPE type jnpr routers - the kind thats a gateway in some small outfit . lots of cos are in that space offering LAN + WAN ports and various apps like print servers, firewalls, spam filters in the same device and claiming thats the only box customers need to deploy.

I agree this 3rd party app thing is motherhood and apple pie some SVP and prod mgr types dreamt up on the golf course after seeing apple doing it on their ipods. brings in some 'open' 'crowdsourcing' 'building an ecosystem around the box' kind of buzz and looks good on resumes for the next job.

in future if the control plane of router/switch becomes a virtual machine there is more scope to let in these apps via their own virtual machines which are inherently sandboxed away from the gonads of the beast.

armenon · Post by **armenon** » 12 Oct 2009 09:18

^^ We can add storage also to the above list. I have seen boxes with the above feature list along with a usb port so that an external storage box can be added to it and make it a file server. Perfect idea for the SOHO area. Big ticket users are unlikely to take this route.

sinha · Post by **sinha** » 12 Oct 2009 12:42

CISCO's desperation to break into app layer is so so evident. Reactivity, AON --- all classic examples. If they would have worked hard on AON/Reactivity - they would have got something special like IBM Datapower appliance - but as usual, everything lagged and nothing cool ever came out. This will remain on fringes, I think.

Incidentally, some issues raised in previous post are covered in AXP. AXP Virtualization Manager allows segmenting applications guaranteeing CPU, memory, and disk resources. Each application runs in its own environment, allowing it to be installed, upgraded, and debugged independently of the AXP OS and the other applications.

The Linux kernel based on 2.6.+ is hardened for security and manageability. The application architecture can be configured to utilize router security features - firewall, intrusion prevention systems, ACL, NAT

It was quite interesting to read the 3rd placed winner app-
"-- Third Place: Team BugsBernie, led by Bernhard Beckmann in Germany, won US$20,000 for the Integrated Surveillance System application. With this application, Internet Protocol phones can be turned on during nonworking hours to monitor any audio signals in the offices. When abnormal audio signal patterns are detected (crossing a configurable threshold), the
application notifies external security services or devices such as mobile phones, computers and video monitoring systems. Sabotage of telephony equipment is also detected. The Integrated Surveillance System is a simple and cost-effective means to enable a security solution in branch offices by taking advantage of an existing IP-telephony network. The application improves manageability of security systems by providing an integrated security framework to an existing network."

SRoy · Post by **SRoy** » 18 Oct 2009 22:29

Some news from the trench for fellow IT-Vity abduls...

We completed our 2 year long re-organization. Many of the my buddies are gone, and some have left.

A pleasant change which I hope that some desi outfits (pure desi or desi units of MNC) might emulate...our line of business driven units have been dismantled and reorganized along portfolio driven units.

It were the management types that used to administer the line driven units, but now the control has been handed over to techie types.

Here goes...

If I'm a senior architect and I've some business idea, I'll have to follow a standard product lifecycle management process to define my case. If accepted, I'll have a delivery team and infrastructure transferred to me. I of course then will be responsible for service delivery and business acquisition. So an architect gets to be the service offering manager if he (no she's here...a$$ licking doesn't work here) is willing to take risk, innovate and has some business sense.

Imagine the plight of the project managers...

, these buggers were the demi-gods. Now the service offering managers will pick from the general pool of project managers and make them work. What's more the project managers are now out of reporting hierarchy. Even lowly the developer crowd will be controlled by their SOM and PM will just jot down his/her remarks but will not conduct appraisals.

The fun doesn't stops here...every service offering being accepted locally must be aligned with some global portfolio (i.e. a portfolio here may consist of multiple service offerings), and SOM will be visible worldwide and will be responsible for projects under his SO. Wow...this eliminates desi bootlickers that have managed sell them as "domain experts" and have landed architect positions.

sinha · Post by **sinha** » 18 Oct 2009 23:17

SRoy wrote:Imagine the plight of the project managers... , these buggers were the demi-gods. Now the service offering managers will pick from the general pool of project managers and make them work. What's more the project managers are now out of reporting hierarchy. Even lowly the developer crowd will be controlled by their SOM and PM will just jot down his/her remarks but will not conduct appraisals.

God bless ur company. Long overdue in most company.

Tanaji · Post by **Tanaji** » 19 Oct 2009 13:45

^^^

Wow a re-org that actually made sense! Best wishes!

Regarding BugsBernie:

I wonder how this works? Is it neutral to the IP phone manufacturer? That would be something if they managed to pull it off that way. Its more likely that it is Asterisk oriented.

Speaking of IP phones, here is an old video, but one that is really well made:

http://www.youtube.com/watch?v=aHKHVsAVVw8

Its about the Microsoft OCS soft switch. It helps if you have seen the movie "Devil Wears Prada" though.

Muppalla · Post by **Muppalla** » 19 Oct 2009 14:50

SRoy wrote:If I'm a senior architect and I've some business idea, I'll have to follow a standard product lifecycle management process to define my case. If accepted, I'll have a delivery team and infrastructure transferred to me. I of course then will be responsible for service delivery and business acquisition. So an architect gets to be the service offering manager if he (no she's here...a$$ licking doesn't work here) is willing to take risk, innovate and has some business sense.

This is not new. In massa land, it has become the norm during the last couple of years. The only interesting thing is the delivery portion to the architects.

Even in the jobs board you will see the terms " Technical Manager " and " Engagement Manager " etc. as opposed to Project Manager. The qualifications for TM and EM will have technical skills as opposed to just PMP or CMMI etc.

This model has its drawbacks from a personal level of the architects who are in these roles. The delivery and management portions of the work is very involved and time consuming. Over a period of time the architects may lose their edge in the skills that they originally possess.

sinha · Post by **sinha** » 19 Oct 2009 17:01

Muppalla wrote:This model has its drawbacks from a personal level of the architects who are in these roles. The delivery and management portions of the work is very involved and time consuming. Over a period of time the architects may lose their edge in the skills that they originally possess.

But then as an architect you can go and hire from this pool of project managers and man managers SRoy was talking about to do ucharts/run charts and predictive defect analysis, client MoM, deal with leaves and schedule planning type of crap.
There is a danger if you are the framework junkie type of architect (and not the MS-WORD,PPT, UML model variant) that you will try and do everything yourself and shoot yourself, but most smarter architects get help when they recognize it is not something they wanna do.
In the model SRoy is talking about, it is expected that Architects should be versatalist - not generalist and not specialist. If need be - get down and dirty in all types of aspects.

manish · Post by **manish** » 19 Oct 2009 17:14

sinha wrote:
Muppalla wrote:This model has its drawbacks from a personal level of the architects who are in these roles. The delivery and management portions of the work is very involved and time consuming. Over a period of time the architects may lose their edge in the skills that they originally possess.
But then as an architect you can go and hire from this pool of project managers and man managers SRoy was talking about to do ucharts/run charts and predictive defect analysis, client MoM, deal with leaves and schedule planning type of crap.
There is a danger if you are the framework junkie type of architect (and not the MS-WORD,PPT, UML model variant) that you will try and do everything yourself and shoot yourself, but most smarter architects get help when they recognize it is not something they wanna do.
In the model SRoy is talking about, it is expected that Architects should be versatalist - not generalist and not specialist. If need be - get down and dirty in all types of aspects.

Sounds good on paper, but will there be enough such SuperMen in a single organization in the desired positions? Might be a tough (and risky) call on the part of the company.

I hope that it doesn't lead to a situation wherein we simply end up with the new SuperMen taking over the much reviled positions earlier held by the PMs. Then the problems would still remain, only the actors would be different. Worse still, it may actually end up depleting the Tech Corps if not handled carefully, by overburdening the architect junta. Esp since quite a few of them seem disinclined to get into such managerial roles.

JMT.

sinha · Post by **sinha** » 19 Oct 2009 17:38

Tanaji wrote: Regarding BugsBernie:
I wonder how this works? Is it neutral to the IP phone manufacturer? That would be something if they managed to pull it off that way. Its more likely that it is Asterisk oriented.

Cisco contest - cannot be Asterisk - will all be based around CISCO callmanager and CISCO IP Phone APIs (not my area, but can hazard a guess here)

SRoy · Post by **SRoy** » 19 Oct 2009 19:54

sinha wrote:
Muppalla wrote:This model has its drawbacks from a personal level of the architects who are in these roles. The delivery and management portions of the work is very involved and time consuming. Over a period of time the architects may lose their edge in the skills that they originally possess.
But then as an architect you can go and hire from this pool of project managers and man managers SRoy was talking about to do ucharts/run charts and predictive defect analysis, client MoM, deal with leaves and schedule planning type of crap.
There is a danger if you are the framework junkie type of architect (and not the MS-WORD,PPT, UML model variant) that you will try and do everything yourself and shoot yourself, but most smarter architects get help when they recognize it is not something they wanna do.
In the model SRoy is talking about, it is expected that Architects should be versatalist - not generalist and not specialist. If need be - get down and dirty in all types of aspects.

Gents,

Let me explain. These type of roles are not for every architect loafing around. Its not about burdening someone with workload.

I'd say its other way round. It is a belated recognition of those techie seniors (well mostly at architects roles, some subject matter experts in niche areas well) that have been doing the following unsung for many years:

1. First rate pre-sales work
2. Tapping customer contacts to generate business
3. In touch with market trends and technology changes
4. Architects/Tech leads are first line of contact with developers...some of these seniors have done a yeoman job in mentoring and retaining talents.
5. Risk taking types...that have delivered strategic projects
6. Take requisite interest in finance topics...i.e. understand billability and receivable recovery issues.

So, these "Supermen" have been the doers and the "Project Managers" have been earning the kudos.

Alas, professional Project Managers have poor adaptability in fast changing technology market. And being process driven these people take a long time to change course. Last but not the least they don't understand technology and make a a$$ of themselves in front of the customers.

Also note, most senior techies / architects have led projects themselves once or twice and many are PMP certified. Do we see ITAC certified project managers?

SRoy · Post by **SRoy** » 19 Oct 2009 20:06

manish wrote:Sounds good on paper, but will there be enough such SuperMen in a single organization in the desired positions? Might be a tough (and risky) call on the part of the company.

I hope that it doesn't lead to a situation wherein we simply end up with the new SuperMen taking over the much reviled positions earlier held by the PMs .

Valid concerns.

manish wrote:Then the problems would still remain, only the actors would be different. Worse still, it may actually end up depleting the Tech Corps if not handled carefully, by overburdening the architect junta. Esp since quite a few of them seem disinclined to get into such managerial roles.

JMT.

OTOH, we may have just addressed an oft repeated complain that there is no career path for techies, unless they move to management.

Here you get some management responsibilities since you have been allocated manpower, funds, sales channels and business targets.

Earlier the goal was to command a bunch of hapless developers, "Managing" was the goal by itself.

The portfolio responsibility implicitly means what you offer as service may be obsolete tomorrow, unless you keep in touch with market demand and technology. If your service offering gets dropped from the organizational portfolio your are back with lowly abduls. Plain and simple.

There is qualitative difference between this and the project management driven units.

manish · Post by **manish** » 19 Oct 2009 20:58

SRoyji, thanks for the reply. I am not really what you could call as an old timer, so please bear with me if I make some stupid observations or comments. My comments derive from my relatively short career in the tech industry

SRoy wrote: OTOH, we may have just addressed an oft repeated complain that there is no career path for techies, unless they move to management.

Agree. In my previous organization, they followed a twin-track policy where the initially common career path would split past the PL level for techies and 'pure' manager types. People would often be asked their preferences early on regarding their choice.

There were equivalent positions for both tracks and finally they would merge at the top, by which time the techie could have climbed up to become an 'Engineering Manager' (IIRC). But I must also add that I was working in a product devpt. environment, which would have obviously had a bearing on the existence of such a hierarchy.

SRoy wrote: 1. First rate pre-sales work
2. Tapping customer contacts to generate business
3. In touch with market trends and technology changes
4. Architects/Tech leads are first line of contact with developers...some of these seniors have done a yeoman job in mentoring and retaining talents.
5. Risk taking types...that have delivered strategic projects
6. Take requisite interest in finance topics...i.e. understand billability and receivable recovery issues.

SRoyji, I agree with your points here.I did observe the above SuperMen types among the technical specialists, and these guys were all somehow heroes in the eyes of us young mujahids who fought in the trenches. These guys would negotiate with the customer reps, get the functional reqs, do high level design work and then guide nanna mujahids during implementation. They would 'own' complete products and projects, but in conjunction with a PM. During the boom years leading upto 2008 when people quit and joined jobs on a whim, these people invariably tended to be long-termers who were in it for the long haul, quite unlike the rest.

The org did not employ MBAs as BA/BC for such roles unlike most of the IT service firms today. However, there was a separate mktg org.

But the my concerns were due to the fact that in almost all of the people I encountered, #6 was lacking. They had little or no interest in the financial side of things. And the fact that most of these people tended to look down upon PMs as MS-Word/Excel coders wasn't helping either

So I am sure that there would be such qualified people, but they must be really rare! I think your organization has taken a novel approach and it would be interesting to see the outcome.

CalvinH · Post by **CalvinH** » 21 Oct 2009 08:36

This issue exist in every tech company whether IT services or IT product and even in IT departments of non consulting companies. It solely depends on the objectives of the company, delivery model, industry vertical et al which decides how much importance is to given to one over the other and I think we should stop generalizing it without giving specific references to the business/delivery model.

Its amazing when someone here quotes that Architects

. Take requisite interest in finance topics...i.e. understand billability and receivable recovery issues.

In all my experience in IT product and services companies I never came across one who actually took interest in this beyond what is required. I found tech architect hating this mostly. May be this is higher then a Tech Architect position, something like Tech evangelist, or Lead Architect, Sr product managers etc which again is more of a management role then a technical one.

A tech architect leading a service portfolio sounds like any management job to me rather then technical one.

sinha · Post by **sinha** » 21 Oct 2009 10:08

CalvinH wrote: In all my experience in IT product and services companies I never came across one who actually took interest in this beyond what is required. I found tech architect hating this mostly. May be this is higher then a Tech Architect position, something like Tech evangelist, or Lead Architect, Sr product managers etc which again is more of a management role then a technical one.

I suppose, CalvinH, that may be you need to reconsider and talk to some "real" software architects, then.

Lets start with what the Gurus have said, if you did comp sci in College you would have read the classic book "Software architecture in Practice" - Bass/Klements et al. The first thing they talked about the ABC (architecture business cycle of influence). How on earth can any one make technology/technical decision without considering financial viability and prospect of a product line, time/ease/cost of deployment/development (Internal stakeholder view) or the political considerations in Client domain.

Also check the book which Martin Fowler wrote "Beyond Software architecture" which focused on Marketarchitecture of the product lines. No software architect (technical, infrastructure, application) will ignore commercial realities and pressures which developing the architecture. Even the newer literature you will find e.g. "Software Systems Architecture: Working With Stakeholders Using Viewpoints and Perspectives" beat around the same theme.

In case of enterprise architecture, the linkages get even more pronouned (consider the Zachman's first row and some aspects of row-2). Or even TOGAF. Or Portfolio rationalization methods used where technology just seems to be one dimension.

Staying in tech cocoon is a recipe for disaster if all round force fields are not considered.
At the end, the real job of architect is to balance the contradictory forces at play and come up with something which makes sense overall - technically, commercially, functionaly. Your definition of technical architect sounds too much like a code monkey to me. (not that I support tech architects who cannot code..) and ideal for a tech arch who just wants to specialize but doesnt really care about the business value of IT. My assumption here is technical architect refers to a role wherein there is freedom to make decisions on technology choices for a domain & implementation architecture for that domain.

Cheers

SRoy · Post by **SRoy** » 21 Oct 2009 19:59

CalvinH wrote: Its amazing when someone here quotes that Architects

. Take requisite interest in finance topics...i.e. understand billability and receivable recovery issues.
In all my experience in IT product and services companies I never came across one who actually took interest in this beyond what is required. I found tech architect hating this mostly. May be this is higher then a Tech Architect position, something like Tech evangelist, or Lead Architect, Sr product managers etc which again is more of a management role then a technical one.

A tech architect leading a service portfolio sounds like any management job to me rather then technical one.

Yes leading a service portfolio is a management job. Professionals coming from project management experience fail to deliver this. Architects are required to sign-off on project sizing to the very least and also on FTE allocation in usual cases.

Who undertakes product roadmap management? Architect or Project manager? In fact a project manager not even qualified to do that.

Maybe you are confusing designers or S/W architects with solution architects. They are different roles. If you look up at the curriculum for TOGAF or ITAC certifications for solutions architects. You'll get the idea.

CalvinH · Post by **CalvinH** » 22 Oct 2009 01:24

sinha wrote:I suppose, CalvinH, that may be you need to reconsider and talk to some "real" software architects, then.

Staying in tech cocoon is a recipe for disaster if all round force fields are not considered.
At the end, the real job of architect is to balance the contradictory forces at play and come up with something which makes sense overall - technically, commercially, functionaly. Your definition of technical architect sounds too much like a code monkey to me. (not that I support tech architects who cannot code..) and ideal for a tech arch who just wants to specialize but doesnt really care about the business value of IT. My assumption here is technical architect refers to a role wherein there is freedom to make decisions on technology choices for a domain & implementation architecture for that domain.

Cheers

Sinha Ji, I would admit here that I have not worked with the "Software architects" that you have mentioned in the Post. I have seen and talked to them in some frontline tech companies in bay area and yes they do have all round view of the technical/commercial/functional part of the product/product based service. I have not met a single one having the same insight in Indian (india based MNC included) IT services or product companies. Do we have a Bill gates Steve jobs or Sergey brin in India, someone who can think beyond normal code to the other aspects of a successful product. May be, I havent met them.

Architects are required to sign-off on project sizing to the very least and also on FTE allocation in usual cases.

Agree that sizing/scoping is always done by someone who knows it. My question was when it is a management type of job why cant a smart project manager do it with Architects helping him with their expertise/Product or Industry knowledge rather then other way around.